Clinical Assessment of Child and Adolescent Personality and ...

Clinical Assessmentof Child and AdolescentPersonality and Behavior

Clinical Assessmentof Child and AdolescentPersonality and BehaviorThird EditionPaul J. FrickUniversity of New Orleans, New Orleans, LA, USAChristopher T. BarryUniversity of Southern Mississippi, Hattiesburg, MS, USARandy W. KamphausGeorgia State University, Atlanta, GA, USA

Paul J. FrickUniversity of New OrleansNew Orleans, LAUSApfrick@uno.eduChristopher T. BarryUniversity of Southern MississippiHattiesburg, MSUSAchristopher.barry@usm.eduRandy W. KamphausGeorgia State UniversityAtlanta, GAUSArkamphaus@gsu.eduISBN 978-0-387-89642-7 e-ISBN 978-1-4419-0641-0DOI 10.1007/978-1-4419-0641-0Springer New York Dordrecht Heidelberg LondonLibrary of Congress Control Number: 2009926176© Springer Science+Business Media, LLC 20101st edition: © Allyn & Bacon, 19962nd edition: © Springer Science+Business Media, LLC 2005All rights reserved. This work may not be translated or copied in whole or in part withoutthe written permission of the publisher (Springer Science+Business Media, LLC, 233 SpringStreet, New York, NY 10013, USA), except for brief excerpts in connection with reviews orscholarly analysis. Use in connection with any form of information storage and retrieval,electronic adaptation, computer software, or by similar or dissimilar methodology nowknown or hereafter developed is forbidden.The use in this publication of trade names, trademarks, service marks, and similar terms,even if they are not identified as such, is not to be taken as an expression of opinion as towhether or not they are subject to proprietary rights.While the advice and information in this book are believed to be true and accurate at the dateof going to press, neither the authors nor the editors nor the publisher can accept any legalresponsibility for any errors or omissions that may be made. The publisher makes no warranty,express or implied, with respect to the material contained herein.Printed on acid-free paperSpringer is part of Springer Science+Business Media (www.springer.com)

To my inspiration, Vicki, Josh, Jordan, andJacob (PJF)To my “home team,” Tammy and Andersen (CTB)To the memory of my parents,Richard and Nancy Kamphaus (RWK)

C o n t e n t sPrefacexiPart IBasic IssuesChapter 1Historical Trends 3Chapter Questions 3Definitions of Terms in PersonalityAssessment 4Early History of PersonalityAssessment 6Projective Techniques 7The Diagnostic and Statistical Manualof Mental DisordersDiagnostic Systems 16IDEA and Special Education 17Future Trends 19Chapter Summary 19Chapter 2Measurement Issues 21Chapter Questions 21Defining Personality Tests 22Scores, Norms, and Distributions 25Reliability 34Construct Validity 37Utility 44Conclusions 45Chapter Summary 45Chapter 3Classification andDevelopmentalPsychopathology 47Chapter Questions 47Science and Assessment 47Classification 48Developmental Psychopathology 58Conclusions 64Chapter Summary 65Chapter 4Standards and Fairness 67Chapter Questions 67APA Ethical Principles of Psychologists 68Test Standards 69Conclusions 79Chapter Summary 79Chapter 5Planning the Evaluationand Rapport Building 81Chapter Questions 81Non-specifics in Clinical Assessment 81Clarifying the Referral Question 82Designing the Evaluation 85To Test or Not to Test 88Rapport Building 89Conclusions 96Chapter Summary 97vii

viiicontentsPart IIAssessment MethodsChapter 6Self-Report Inventories 101Chapter Questions 101Omnibus Personality Inventories 101Single Construct PersonalityInventories 138Conclusions 138Chapter Summary 139Chapter 7Parent and Teacher RatingScales 141Chapter Questions 141Evaluating Children via ParentRatings 141Evaluating Children via TeacherRatings 143Overview of Omnibus Parent and TeacherRating Scales 144Behavior Assessment System for Children,2nd Edition (BASC-2) 145Achenbach System of Empirically BasedAssessment (Achenbach & Resorla,2000, 2001) 156Child Symptom Inventory-4(CSI-4) 166Conners, 3rd Edition (Conners-3) 170Personality Inventory for Children-2(PIC-2); Student Behavior Survey(SBS) 174Sample Impairment-Oriented Scales 183Conclusions 185Chapter Summary 187Chapter 8Behavioral Observations 189Chapter Questions 189Basics of Observational Systems 192Examples of Observational Systems 200Conclusions 207Chapter Summary 208Chapter 9Peer-ReferencedAssessment 211Chapter Questions 211Ethics of Peer-ReferencedStrategies 212Types of Peer-Referenced Techniques 214Other Peer-Referenced Techniques 223Conclusions 223Chapter Summary 223Chapter 10Projective Techniques 225Chapter Questions 225The Controversy Surrounding ProjectiveTechniques 225Inkblot Techniques 231Thematic (Storytelling) Techniques 236Sentence Completion Techniques 243Drawing Techniques 245Conclusions 250Chapter Summary 250Chapter 11Structured DiagnosticInterviews 253Chapter Questions 253History 253Overview 254Evaluation of Diagnostic Interviews 259Recommendations for Use of StructuredInterviews 261Focus on the NIMH DiagnosticInterview Schedule for Children(DISC-iv) 265Conclusions 269Chapter Summary 269

contentsixChapter 12Assessing FamilyContext 271Chapter Questions 271Introduction 271Assessing Family Functioning:General Issues 276General Considerations in AssessingFamily Functioning 277Parenting Styles and Practices 280Parenting Stress 288Marital Conflict 291Parental Adjustment 293Conclusions 296Chapter Summary 296Chapter 13History Taking 299Chapter Questions 299Content 307Formats 307Conclusions 313Chapter Summary 314Chapter 14Adaptive BehaviorScales 315Chapter Questions 315History of the Construct 315Characteristics of Adaptive BehaviorScales 318Omnibus Adaptive Behavior Scales 322General Recommendations 327Measuring Social Skills 332Conclusions 334Chapter Summary 334Part IIIAdvanced TopicsChapter 15Integrating and InterpretingAssessment Information 339Chapter Questions 339Introduction 339Integrating Information acrossInformants 342A Multistep Strategy for IntegratingInformation 347Step1: Document all Clinically SignificantFindings Regarding the Child’sAdjustment 349Step2: Look for Convergent FindingsAcross Sources and Methods 349Step3: Try to Explain Discrepancies 353Step4: Develop a Profile and Hierarchyof Strengths and Weaknesses 353Step5: Determine Critical Informationto Place in the Report 354Conclusions 355Chapter Summary 355Chapter 16Report Writing 357Chapter Questions 357Reporting Problems and Challenges 357Report Writing as Part of Evidence-BasedAssessment 359Pitfalls of Report Writing 360Suggested Practices 365Adapting Reports to Audienceand Setting 368The Sections of the PsychologicalReport 369Communicating Results Orally 373Conclusions 375Chapter Summary 376

xcontentsChapter 17Assessment of AttentionDeficit Hyperactivityand DisruptiveBehavior Disorders 377Chapter Questions 377Introduction 377Attention-Deficit HyperactivityDisorder 378Conduct Problems 398Conclusions 410Chapter Summary 412Chapter 18Assessment of Depressionand Anxiety 413Chapter Questions 413Internalizing Disorders 413Childhood Depression 414An Assessment Strategy for Depression 422Anxiety Disorders of Childhood 425An Assessment Strategyfor Anxiety 431Conclusions 434Chapter Summary 434Chapter 19Assessment of Autism SpectrumDisorders 437Chapter Questions 437Definitional Issues 437Specialized Measures of Autism 439An Assessment Strategy for Autism 445A Sample Case of Autism in a Childwith Neurological Impairment 447Conclusions 452Chapter Summary 452References 453Author Index 491Subject Index 503

P r e f a c ePsychologists offer an increasing varietyof serv ices to the public. Among theseservices, psychological assessment of personalityand behavior continues to be acentral activity. One main rea son is thatother mental health professionals oftendo not possess a high level of competencein this area. When one views psychologistswho serve children and adolescents,psychological, assessment seems to take onan even greater role. It follows, then, thatcomprehensive and enlightened graduatelevelinstruction in assess ment should be ahigh priority for educators of psychologistswho are destined to work with youth.This book is an outgrowth of our effortsto improve our own instruction of childand adoles cent assessment skills. We foundthat existing textbooks were not servingus well. Most of them were encyclopedic,edited volumes that were (1) uneven inthe quality across chapters and/or (2) notgeared either in format or level of presentationfor beginning graduate instruction.The few single- or co-authored volumesavailable tended to lack the breadth of coveragewe deemed necessary. Some focusedlargely on the oretical issues related to psychologicaltesting, with minimal discussionof practical applications and use of specifictests. Others focused solely on summariesof individual tests, without reviewingthe theoretical or empirical contextwithin which to use the tests appropriately.Hence, this volume reflects our desire toprovide a more helpful tool for instruction- one that provides a scien tific contextwithin which to understand psychologicaltesting with children and adolescents andthat translates this scientific context intopracti cal guidelines for using individualtests in clinical practice.Among our specific objectives for thisvolume are the following:• To focus on measures specificallydesigned to assess the emotional, behavioral,and social functioning of childrenand adolescents• To provide current research findingsthat enable students to draw heavilyon science as the basis for their clinicalpractice• To help in the translation of researchinto practice by providing specific andpractical guidelines for clinical practice• To include a broad coverage of assessmentmethods from a variety oftheo retical, practical, and empirical traditions• To systematically compare tests andassess ment methods using researchfindings, reviews, and our own synthesisof positions• To provide a readable volume that woulden hance the interest and retention ofstudents through the use of numerousxi

xiicase examples, tables, research notes, andother “boxes” containing practical adviceIn writing a readable text that goes beyonda cursory survey of available instruments,we were faced with the difficult task ofdetermining what specific instruments toinclude in the book. As we struggled withthis decision, we eventually chose severalselection criteria. Our main crite rion forinclusion was a test’s ability to serve as anexemplar of some specific type of assessmentin strument or theoretical approachto assessment. In many cases we lookedfor “prototypes” that we thought wouldhighlight some key points to the readerthat could then be used in evaluatingother assessment instruments not coveredspecif ically in the text. Other criteriaincluded a test’s popularity or our estimateof a new test’s poten tial impact on the field.Granted, this decision- making was highlysubjective, but we strove to be analytic andto limit to the extent possible our personalfeelings and biases. Fortunately, severalexternal anonymous reviewers of earlierdrafts of the text helped us to be moreobjective.A final point that should be clearly outlinedis our basic orientation to psychologicalassess ment. We feel that the goalof psychological as sessment is the measurementof psychological constructs and,for clinical practice, measurement of psychologicalconstructs that have importantclinical implications, such as documentingthe need for treatment or the type of interventionthat is most appropriate. For anindividual child the constructs that need tobe assessed will vary from case to case anddepend on the referral question. But whatis important from this con ceptualization isthat our view of psychological assessmentis not test-driven but construct-driven.Without exception, assessment techniqueswill measure some constructs welland other psychological dimensions lesswell. Another important implication fromprefacethis view of testing is that it is critical thatassessors become familiar with and maintainfamiliarity with research on the constructsthey are trying to assess. In short,the most critical component in choosing amethod of assessment and in interpretingassessment data is understanding what oneis trying to measure.In this volume, we have focused on themeas urement of psychological constructsthat em phasize a child’s emotional, behavioral,and social functioning. There is stilla great debate over definitional issues inthis arena of psychological functioning interms of what should be called “personality,”“temperament,” “behav ior,” or “emotions,”with distinctions usually de terminedby the level of analysis (e.g., overt behaviorvs. unconscious motivational processes),assumed etiology (e.g., biologically determinedvs. learned), or proven stability(e.g., transient problems vs. a stable patternof functioning) (see Frick, 2004; Martin,1988). Unfortunately, people often usethe same terms with very different connotations.In writing this text, we tried toavoid this debate by maintaining a broadfocus on “psychological constructs,” whichoften vary on the most ap propriate level ofanalysis, assumed etiol ogy, or stability. Thisdefinitional variability adds a level of complexityto the assessment process becauseassessors must always consider what theyare attempting to measure in a partic ularcase and what research suggests aboutthe best way of conceptualizing this constructbefore they can select the best wayof measuring it. It would be much easier ifone could develop ex pertise with a singlefavorite instrument or a sin gle assessmentmodality that could be used in all evaluationsto measure all constructs. Becausethis is not the case, psychologists mustdevelop broad-based assessment expertise.Hence, our overriding objective in writingthis volume is to provide the breadthof coverage that we feel is needed by thepsychologist in training.

prefacexiiiWe have organized the text into threesec tions consistent with our approach toteaching. Part I provides students with anintroduction to the psychological knowledgebase necessary for modern assessmentpractice including historical perspectives;meas urement science; research in developmentalpsychopathology; ethical, legal, andcultural issues; and the basics of beginningthe assessment process (e.g., planning theevaluation, rapport building). Part II givesstudents a broad review of the specificassess ment methods used by psychologists,accompa nied by specific advice regardingthe use and the strengths and weaknesses ofeach method. In Part III we help studentsperform some of the most sophisticated ofassessment practices: inte grating and communicatingassessment results and infusingassessment practice with knowledge ofchild development and psychopathology.We think that, on completion of this volume,and a similar one covering aspects ofcognitive assess ment (Kamphaus, 2001 andin press), that the student psy chologist hasthe background necessary for supervisedpracticum experiences in the assessment ofchil dren and adolescents.Paul J. FrickChristopher T. BarryRandy W. Kamphaus

P a r t IBasic Issues

C h a p t e r 1Historical TrendsChapter Questionsl Who were the major innovators in thefield of personality assessment?l How were these innovations extended tothe assessment of children and adolescents?l What is meant by the terms personalityand behavior?l What is meant by the terms objectiveand projective personality assessment?l Who conducted the seminal researchand coined the terms internalizing andexterna lizing behavior problems?Personality assessment is a process thatmost individuals engage in throughouttheir lives (Martin, 1988). Mothers labeltheir children as happy, cranky, or similarlyshortly after birth, and often in utero(e.g., active). The musings of Alfred Binetabout the personality of his two daughtersare typical of observations made by parents.He described Madeleine as silent, cool, andconcentrated, while Alice was gay, thoughtless,giddy, and turbulent (Wolf, 1966).Adolescents are keenly aware of personalityevaluation as they carefully considerfeedback from their peers to perform theirown self-assessments. Personality assessmentis also prized by the business community,in which human resources personnelconsult with managers and others to gaugethe effects of their personality on coworkersand productivity.Early personality assessment emphasizedthe assessment of enduring traits that werethought to underlie behavior or, in modernterminology, latent traits. Kleinmuntz (1967)described personality as a unique organizationof factors (i.e., traits) that characterizesan individual and determines his or her patternof interaction with the environment.Thus, personality structure is commonlythought to be a result of multiple individualtraits interacting with one another, and withthe person’s environment.3P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_1, © Springer Science+Business Media, LLC 2010

4CHAPTER 1 HISTORICAL TRENDSDefinitions of Termsin Personality AssessmentTraitsA trait is often conceptualized as a relativelystable disposition to engage in particular actsor ways of thinking (Kamphaus, 2001 and inpress). A child, for example, may be describedby her parents as either shy or extroverted.The shy (introverted in psychological terms)child may tend to cope with stressful situationsby withdrawing from social contact,whereas the extrovert readily approachessocial situations. For parents and psychologistsalike, these traits are often thought tohave value for predicting human behavior,because of the presumption of trait stabilityacross time and, in many cases, environments.In fact, because of trait stability, parents maytake special precautions to ensure that theshy child adapts well to the social aspectsof attending a new school by asking one oftheir child’s friends who attends the sameschool to accompany the child on the firstday. Similarly, a stable tendency to be shy orintroverted should manifest itself in numeroussocial situations such as interactions inthe neighborhood, at church, and in balletclass. Personality traits, then, are characterizedby longitudinal and situational stability,not unlike other enduring characteristics ofa person such as intelligence, height, andactivity level.The Big Five PersonalityTraits (Factors)In 1961, Tupes and Christal discoveredfive factors of personality that appeared inthe reanalysis of numerous data sets fromscales of bipolar personality descriptors.These central personality traits have subsequentlybecome the focus of an extensiveresearch effort, including the developmentof tests designed to assess the constructs.One of the well known scales used to identifythe “big five” in adults is the NEOPersonality Inventory (NEO-PI; Costa &McCrae, 1985).Although commonly referred to as “factors”because of their origins in factor analysis,they are prototypical examples of traitswith the requisite characteristic of presumedstability. The big five factors are typicallyidentified by bipolar comparisons that aresummarized in Table 1.1. These factors areoften assessed using forced-choice item formatsin which adjectives are used as personalitydescriptors. This item format is in directcontrast to the more commonplace true/falseitem format that is typical of many psychologicaltests.Table 1.1 Early Descriptions of the BigFive Personality Dimensions (Goldberg,1992)Factor I – Surgency (or introversion–extroversion)Unenergetic vs. energeticSilent vs. talkativeTimid vs. boldFactor II – Agreeableness (or pleasantness)Cold vs. warmUnkind vs. kindUncooperative vs. cooperativeFactor III – Conscientiousness (or dependability)Disorganized vs. organizedIrresponsible vs. responsibleNegligent vs. conscientiousFactor IV – Emotional stability (vs. neuroticism)Tense vs. relaxedNervous vs. at easeFactor V – Culture, intellect, openness,or sophisticationUnintelligent vs. intelligentUnanalytical vs. analyticalUnreflective vs. reflective

CHAPTER 1 HISTORICAL TRENDS5Commercially available instrumentssuch as the NEO-PI have provided newopportunities to study and refine theseconstructs. Given the amount of researchand development in this area, the big fivepersonality factors could eventually have asubstantial impact on the field of child andadolescent personality assessment. Withsome noteworthy exceptions (e.g., Lynamet al., 2005) however, big five research haslargely been focused on adult populations.TemperamentA concept related to personality is temperament,which also emphasizes the measurementof specific traits that are hypothesizedas underlying behavior across settings. Inthis regard, Goldsmith and Rieser-Danner(1990) observed, “most researchers considertemperament to be the behavioralmanifestation of biologically influencedprocesses determining both the infant’scharacteristic response to the environmentand his or her style of initiating behavior”(p. 250). Therefore, some researchers distinguishtemperament from personalitybased on the presumed biological basisof temperament, whereas personality isthought to be formed by a dynamic interplayof biological and social factors overdevelopment (Frick, 2004; Martin, 1988).Predictably, much of the research on temperamentis conducted with infants andyoung children. In this conceptualization,personality may be viewed as beingsuperimposed on a person’s temperamentalfoundation. This distinction betweentemperament and personality, however, isnot universally agreed upon.BehaviorIn contrast to temperament and personalitytrait assessment, the assessment of behaviorfocuses on the measurement of observablebehaviors, although recently the definitionhas been broadened to include cognitionsas a type of behavior. For most purposes,Martin (1988) provides a useful definitionof behavior.When applied psychologists speak of behavior,they are usually referring to that range ofhuman responses that are observable withthe naked eye during a relatively brief periodof time in a particular environment.This conception of behavior rules out biochemicaland neurological events, for example,because they are not observable bythe unaided eye. Behavior is differentiatedfrom traits or dispositions because the lattermay only be seen if behavior is aggregatedover relatively long periods of time and ina number of environmental contexts. Classicalexamples of observed behaviors of interestto child psychologists include tantrumbehavior among young children, aggressiveinteractions with peers, attempts at conversationinitiation, and so forth (p. 13).There are, therefore, several distinguishingfeatures of behavioral assessment methodsthat differentiate them from trait assessmentmeasures. First, behavioral assessmentmethods have a different theoretical foundationand associated set of premises. Behavioralassessment methods draw heavily onthe theory and research tradition of operantconditioning as exemplified by the work ofB. F. Skinner (Skinner, 1963). This researchtradition also emanates primarily from laboratoryresearch, as opposed to clinical practice;thus, it is often considered to be moreempirically based.Second, behavioral assessment methodsare distinguished from medical modelsof assessment more than are trait-basedmethods. The medical model assumesthat symptoms are caused by underlyingconditions, and it is the medical conditionthat must be measured, diagnosed,and treated to remove the symptoms (seeChap. 3 for a more extended discussionof the medical model). In direct contrast,behavioral assessment emphasizes themeasurement and treatment of the symptoms

6CHAPTER 1 HISTORICAL TRENDSor behavior itself, and makes no assumptionsregarding underlying cause.Third, behavioral assessment places apremium on the assessment of discretebehaviors. For example, behavioral assessmentmay emphasize the measurement offinger tapping while completing seatworkin school, as opposed to aggregating severalbehaviors to form a test or scale with severalitems that measure “motor activity” inthe classroom. This situation is changing,however, with the work of Thomas Achenbach,Keith Conners, Cecil Reynolds, andothers (see Chap. 7), all of whom begancombining behaviors into dimensions ofbehavior that may or may not differ fromtrait-based assessment methods.As we suggested in the preface, we thinkthat it is premature to reify any of theseapproaches as the ultimate method forassessing children’s psychological adjustment.We merely seek progress in our methodsand theories. This cautious approachseems warranted as the distinctions betweenthe methods have become blurred as thescience of assessment emerges. Furthermore,we think that each approach maybe more or less helpful for answering particularassessment questions. Clearly, someof the questions directed at psychologistsare trait-based, whereas others require themeasurement of distinct behaviors. Forexample, a parent who asks, “Will my childever become more outgoing like his sister?”is asking for trait assessment, but the parentwho queries, “How can I get him to stopwetting the bed?” may require behavioralassessment expertise.Early History of PersonalityAssessmentFormal personality measures emerged asa logical outgrowth of other efforts to measureindividual differences, most notably theexperimental methods of Wundt, Galton, andothers (Chandler, 1990). Of the early assessmentluminaries, Sir Francis Galton is oneof the most notable. Although well knownfor his intelligence measurement contributions,he also studied the measurement of“character.” In order to introduce the utilityof personality measurement, Galton (1884)recounted the personality test invented byBenjamin Franklin as a crude form of personalitymeasure. The scale was described inthe tale of “The Handsome and DeformedLeg,” in which Franklin recounts how hisfriend tested people so as to avoid those who,“being discontented themselves, sour thepleasures of society, offend many people,and make themselves everywhere disagreeable”(p. 9). This friend sought to diagnosesuch pessimistic individuals by showingthem an attractive leg and a malformedone. If the stranger showed more interestin the ugly leg, the friend became suspiciousand subsequently avoided this person. Franklinastutely identified this “test” as a grotesquebut, nevertheless, an effective personalityassessment device. Galton concluded: “Theother chief point that I wish to impress is,that a practice of deliberately and methodicallytesting the character of others and ofourselves is not wholly fanciful, but de-servesconsideration and experiment” (p. 10).Intelligence tests, acknowledged as thefirst fruits of the psychometric movement,reached prominence early in the twentiethcentury with the introduction of the originalBinet and Simon scale and numerous variants(Kamphaus, 2001 and in press). A lesserknown fact is that Alfred Binet developedsome intelligence test items that resembledstimuli used 30 years later in apperceptivetechniques for assessing personality(DuBois, 1970). Test development activityalso received a boost from the World WarI effort, when ability testing became widespread(Kamphaus, 2001 and in press).Thus, it is no coincidence that the first formaland widely used measures of personalitywere developed about this same time.

CHAPTER 1 HISTORICAL TRENDS7Robert S. WoodworthThe Woodworth Personal Data Sheet waspublished in 1918 as a result of the surgeof interest in testing potential soldiers.Woodworth developed a list of 116 questionsabout daydreaming, worry, and otherproblems. Some sample items from theWoodworth Personal Data Sheet (Woodworth,1918) include:Do people find fault with you much?Are you happy most of the time?Do you suffer from headaches and dizziness?Do you sometimes wish that you had neverbeen born?Is your speech free from stutter or stammer?Have you failed to get a square deal in life?The examinee responded to each questionwith “yes” or “no” (French & Hale, 1990).According to French and Hale (1990),the Woodworth Personal Data Sheetserved as the foundation for the developmentof the Thurstone Personality Scaleand the Allport Ascendance-SubmissionTest, among others. DuBois (1970)described the Personal Data Sheet as “thelineal ancestor of all subsequent personalityinventories, schedules and questionnaires”(p. 94).The Personal Data Sheet was an importantpractical innovation because, prior tothis time, all military recruits suspected ofhaving mental health disorders, stress disordersin particular, had to be identifiedby being interviewed by trained interviewers.The Personal Data Sheet allowed forthe screening of large numbers of recruitswithout the time and expense of using hugecadres of interviewers (Kleinmuntz, 1967).Thus, it was not basic researchon personality or employee selection thatled to the eventual popularity of personalitytesting. Instead, it was the need for diagnosiscreated by World War I which providedconsiderable evidence of the need for personalitytests. The successful World War Iand then later World War II applications ofpsychological testing proved that psychologycould make practical contributions tosociety by identifying, accurately and timeefficiently,those in need of mental healthservices.After World War II the mental healthneeds of citizens, veterans in particular,were the focus of greater attention. In thepostwar years, the U.S. Veteran’s Administrationbegan to hire psychologists inlarge numbers to diagnose and treat veteranssuffering from significant emotionaldisturbance. Psychologists brought theirpsychometric expertise to bear again bycontributing new methods to the diagnosticprocess. The increased need for postwarmental health services, therefore, createdthe fertile ground in which personalitytesting flourished. As Kleinmuntz (1967)noted, “The most popular personality testsof the past 30 years grew out of the needto diagnose or detect individuals whosebehavior patterns were psychopathological”(p. 10). The use of personality tests afterthe first and second world wars expandedbeyond diagnosis into many areas includingcounseling, personnel selection, andpersonality research (Kleinmuntz, 1967).Projective TechniquesThe central assumption underlying projectivetesting is that the use of less welldefinedstimuli that are prone to a varietyof interpretations will encourage clientsto reveal information that they otherwisewould not share in response to direct questioning(Chandler, 1990). Given that teststimuli (e.g., ink blots) or questions werenot clearly linked to known personalitytraits, projective testing depended heavilyon the explicit or implicit personality theorythat was favored by the test developer.Theory was necessary to determine theunderlying nature or cause of the projectedthoughts, emotions, or behaviors.

8CHAPTER 1 HISTORICAL TRENDSThe most popular theory of the post-World War I era was psychodynamic personalitytheory as espoused by SigmundFreud (1936) and others. Psychodynamicpersonality theory provided a useful theoreticalframework for the development ofprojective assessment measures due to theconcepts of repression, projection, andother constructs that are entirely consistentwith the use of atypical test stimulifor identifying personality traits (seeChap. 10 for a more extended discussionof the basic assumptions of the projectivetechnique).Association TechniquesThe use of association techniques, such asword association methods, for assessmentpurposes can be traced as far back as thework of Aristotle (DuBois, 1970). In relativelyrecent history, Sir Francis Galtonbegan studying association techniques asearly as 1879. Galton’s contribution to thestudy of association was his introduction ofscientific rigor to the enterprise. He usedexperimental methods to study associationmethods, including quantitative scaling ofthe results (DuBois, 1970).Subsequently, Kraepelin, Wundt, Cattell,Kent, and Rosanoff studied the associationsof patients and research participantsto word lists, recording such variablesas response time and type of association.The latter names, Kent and Rosanoff, maybe least familiar to many readers becausethe other names are linked with the illustrioushistory of intellectual assessment.Kent and Rosanoff made their contributionsolely to the study of associations bydeveloping a list of 100 stimulus words andsystematically recording the associationsof 1,000 normal subjects (DuBois, 1970).This effort represents an important initialattempt at developing norms to whichresearchers and clinicians could comparethe responses of clinical subjects.The renowned psychoanalyst CarlJung made extensive use of associationtechniques for the study of personality. Inan address at Clark University in 1909, hedescribed his research efforts in detail andprovided some insight into the types ofinterpretations commonly made of thesemeasures. Jung described his associationword list as a formulary. His list consistedof 54 words including head, to dance, ink,new, foolish, and white. According to Jung(1910), normality could be distinguishedfrom psychopathology with this formularyusing variables such as reaction timeand response content. In his speech, heprovided a transcript of the responses ofa normal individual and of a “hysteric.”A sampling of their associations to theformulary follows:Stimulus Normal HystericalTo sin Much This is totallyunfamiliar tome, I do notrecognize itTo pay Bills MoneyBread Good To eatWindow Room BigRich Nice Good,convenientFriendly Children A manAs noted previously for other measures,reaction time to the stimulus words wasalso interpreted by Jung. He gave a glimpseof one such interpretation in the followingquote:The test person waives any reaction; for themoment he totally fails to obey the originalinstructions, and shows himself incapable ofadapting himself to the experimenter. If thisphenomenon occurs frequently in an experiment,it signifies a higher degree of disturbancein adjustment (p. 27).

CHAPTER 1 HISTORICAL TRENDS9These early word association methods setthe stage for the development of otherassociation (projective) techniques, suchas the Thematic Apperception Techniqueand Rorschach’s test, both of which usedpictures in lieu of word lists to elicit associations.Thematic Apperception TestThe Thematic Apperception Test (TAT) ofHenry A. Murray constitutes a prototypicalexample of a projective device. Charles E.Thompson summarized the central tenetof the projective approach in the followingquote taken from his 1949 adaptation ofMurray’s TAT. 1If the pictures are presented as a test of imagination,the subject’s interest, together withhis need for approval, can be so involved inthe task that he forgets his sensitive self andthe necessity of defending it against the probingof the examiner, and, before he knows it,he has said things about an invented characterthat apply to himself, things which he wouldhave been reluctant to confess in response to adirect question (p. 5).The TAT is unique among projective measuresin that it has traditionally beeninterpreted qualitatively, even though quantitativescoring methods are available(Kleinmuntz, 1967). Murray’s originalapproach to TAT scoring was entirelyqualitative and psychoanalytically based.He proposed the following categories foranalyzing the characteristics of the storiesgiven by the subject (Murray, 1943).1Thompson’s modification is identical to the originalTAT with the exception that African Americanfigures are used as characters on the stimulus cards.Thompson found that African Americans did notrespond optimally to the original TAT pictures. Infact, one of his patients asked if he could imagine thatthe people in the pictures were “colored,” and if hecould make up some stories about “colored people.”1. The Hero. This is the person with whomthe subject seems to identify. The heromay share characteristics such as age,gender, occupation, or other featureswith the subject that aid identification.The hero’s traits should be evaluatedto determine the self-perceptions ofthe subject including superiority, intelligence,leadership, belongingness, solitariness,and quarrelsomeness.2. Needs of the Hero. Needs may includethose for order, achievement, and nurturance.3. Environmental Forces. Factors that affectthe hero and these are also referred to aspress. An example would involve scoringaggression if the hero’s property or possessionswere destroyed in a story.4. Outcomes. The success of the hero andthe hero’s competencies are assessed byevaluating the outcomes of stories.5. Themas. Themas assess the interplay ofneeds and presses, and they reveal theprimary concerns of the hero.6. Interests, Sentiments, and Relationships.For this aspect of scoring the examinerrecords the hero’s preferences for topics.Murray’s qualitative scoring system forthe TAT is a classic example of systemsthat dominated the early interpretation ofprojective devices. Numerous quantitativescoring systems followed Murray’s originalwork as exemplified by scoring systems eventuallydeveloped for Rorschach’s test.The RorschachHermann Rorschach (1884–1922) was amajor figure in Swiss psychiatric researchwho began his work as a physician in 1910.He married a Russian colleague who becamehis comrade and collaborator (Morgenthaler,1954). He served as a physician in a hospitalin Herisau until his death from complicationsof appendicitis in 1922. His death was

10CHAPTER 1 HISTORICAL TRENDSdescribed as a critical blow to Swiss psychiatry.In a 1954 eulogy to Dr. Rorschach,published in the English translation of hisoriginal work, Morgenthaler attempted todescribe Rorschach for future generations.Flexibility of character, rapid adaptability, fineacumen, and a sense for the practical werecombined in Hermann Rorschach with a talentfor introspection and synthesis. It was thiscombination which made him outstanding. Inaddition to this rare nature, which temperedpersonal emotional experience with practicalknowledge, he possessed sound traits ofcharacter most valuable in a psychiatrist. Mostimportant of these were an unerring tendencyto search for the truth, a strict critical facultywhich he did not hesitate to apply to himself,and a warmth of feeling and kindness. (p. 9)Rorschach’s approach to personality assessmentwas novel in many respects. The teststimuli used were inkblots placed on paperthat was then folded in half. Rorschachwas not, however, interested in the contentof the subject’s response to the inkblots.Rather, he was interested in the form ofthe response (or its function). Some functionsof interest included the number ofresponses, perception of color or movement,and perception of the whole vs. theparts. These and other characteristics ofRorschach responses continue as part ofmodern scoring systems (see Chap. 10).Rorschach first offered his methodas an experiment. His original sample isdescribed in Table 1.2. He expressed adesire for larger sample sizes but notedthat the number of experiments was limitedbecause the stimulus figures were damagedby passing through hundreds of hands.Rorschach’s legacy, his original inkblots,and many of the associated scoring criteriaremain influential as the test continues toenjoy popularity. Several scoring systemshave been offered for the Rorschach, withthe Exner Comprehensive System (Exner& Weiner, 1982) contributing most to thecontinuing usage of the instrument.Table 1.2 Rorschach’s Original ResearchSampleSampleNormal, educated 55Normal, uneducated 62Psychopathic personality 20Alcoholic cases 8Morons, imbeciles 12Schizophrenics 188Manic-depressives 14Epileptics 20Paretics 8Senile dements 10Arteriosclerotic dements 5Korsakoff and similar states 3Source: Adapted from Rorschach (1951).Sentence CompletionTechniquesSentence completion techniques are venerablepersonality assessment methods ofthe association tradition that can tracetheir roots to Payne (1928). The sentencecompletion method, however, obtained asubstantial boost in popularity because ofits use by the U.S. Office of Strategic Services(OSS), the forerunner of the CentralIntelligence Agency (CIA). Henry Murraywas the coordinator of a sophisticatedOSS assessment effort. About 60 militaryassessment stations, staffed by Americanpsychologists, were situated in the USAand abroad to screen recruits for sensitiveand dangerous assignments. Some of themethods used in this ambitious programare described in the following quote fromDuBois (1970):In one of the stations near Washington, recruitsin fatigue uniforms assumed a falseidentity and developed a cover story, which thestaff members during the three-day stay endeavoredto break. The procedures describedin a comprehensive report (OSS, AssessmentStaff, 1948) were varied: casual conversations,N

CHAPTER 1 HISTORICAL TRENDS11searching interviews, the sentence completiontest, questionnaires about health and workingconditions and personal history, conventionalaptitude tests such as map memory and mechanicalcomprehension, and a number of situationaltests (p. 111).After World War II, the sentence completiontechnique continued to enjoy somefavor among psychologists. The wellknownRotter Incomplete Sentences Blankwas published in 1950 (Rotter & Rafferty,1950). Sentence completion methods havea lengthy history of use with children andadolescents as well. They enjoyed worldwideuse in countries such as Finland, Germany,Denmark, India, Japan, and Taiwan(Haak, 1990).Projective Techniques forChildrenThe use of projective techniques withchildren dates back to the early part ofthe twentieth century when FlorenceGoodenough began to study children’shuman figure drawings (DuBois, 1970).Goodenough noted, as did others, thatchildren’s drawings were affected bytheir emotionality. The typical paradigmfor drawing techniques has been tohave the child draw a picture of a person.Traditionally, the content of the drawingshas been interpreted as a measure ofchild adjustment and personality. Someaspects of content that were extensivelystudied for adults included (Swensen,1968):Size of the person depictedPlacement on the page (bottom, top, corner,etc.)Stance (vertical, horizontal, balanced, etc.)Line quality (heavy, light, etc.)ShadingErasuresOmissions (missing body parts)Distortion (poor proportion of body parts)Various interpretations have been associatedwith these and other content variablesover the years. Heavy lines, for example,have been associated with assertive andaggressive individuals and light lines havebeen viewed as being indicative of passiveindividuals (Koppitz, 1968). Swensen(1968) found such interpretations to behighly unreliable. The most reliable andvalid interpretive approach involved makinggeneral judgments about the mentalhealth status of the individual based on theoverall quality of the drawing, rather thanspecific content interpretations (Cummings,1986).The TAT, among other projectivemethods, has also been adapted by manyfor use with children and adolescents. Oneof the most well-known TAT adaptationsis the Children’s Apperception Test (CAT)(Bellak & Bellak, 1949b), designed forages 3–10. The CAT consists of ten pictureswith animals as stimuli in contrast tothe TAT’s depictions primarily of people.The Rorschach has also been widely usedwith children, and several compendiumsof child responses have been published toaid interpretation (e.g., Ames, Metraux,Rodell, & Walker, 1974).The proper interpretation of children’sprojective responses remains a topic of debate.Indeed, the degree to which children obeythe projective hypothesis has been questioned.Chandler (1990) elucidates the nature of theprojective hypothesis as follows:Projection, in common usage, means to castforward. In this sense, projection implies adirect extension of psychological characteristicsonto the outer world. But projectionalso has a specific meaning within psychoanalytictheory. Freud (1936) used the termto refer to the process that occurs when theego, faced with unacceptable wishes or ideas,thrusts them out onto the external world asa means of defense. In projection the individualattributes his or her own thoughtsand actions to someone else. Thus, if one’sown faults or feelings are unacceptable to

12CHAPTER 1 HISTORICAL TRENDSthe ego, they may be seen as belonging tosomeone else; in the process, the materialmay become distorted or remain partiallyrepressed. From such a perspective, projectivematerial would not be seen as directrepresentation of aspects of the personality,certainly not with the sort of one-to-onecorrespondence that the first meaning ofprojection implies (p. 57).For adults as well as children, the processof projection still rests primarily on a theoreticalrather than empirical foundation. Inthe absence of data to support the projectivehypothesis, psychologists have focusedon the use of psychometric methods toassess the reliability of obtained scoresand the validity of score inferences. Thisshift to the accumulation of psychometricevidence for measures is reflected bestin the work of Exner. In the 1960s, JohnExner began a research program designedto take the best of the Rorschach scoringsystems and incorporate their features intoa comprehensive system (Exner & Weiner,1982). Further, a standard method for scoringresponses on the test has led to scoresthat have proven to be reliable and, as aresult, has set the stage for direct tests ofthe validity of various interpretations thatcan be made from them. The applicationof psychometric standards to projectivemeasures is a clear departure from a longhistory of qualitative analysis and interpretation.The efforts of Exner and othershave set a new course for projective measuresin that they are increasingly held tothe same standard as tests of intelligence,adaptive behavior, and “objective” personalityassessment methods.Objective TestsAlthough we acknowledge that the distinctionbetween projective and objectivetesting is an oversimplification, it is neverthelessuseful for pedagogical purposes.Objective methods can be differentiatedfrom projective tests in several ways. First,objective methods are often consideredto be atheoretical and/or empirical. Asopposed to requiring the examiner to usetheory to interpret results, the results oftenderive their meaning from empirical procedures,such as matching a person’s resultsto those of a clinical sample. Second, objectivemethods are not likely to be based onpsychodynamic theory. Hence, the resultsof objective measures are often consideredto be less useful for providing insight intothe dynamics of an individual’s interactionswith the world. Third, objective methodstake greater advantage of measurementscience for the development of tests. Issuesof item selection, reliability, and validityare often emphasized in the test manuals.Minnesota MultiphasicPersonality Inventory (MMPI)Until the advent of the MMPI, projectivetechniques reigned supreme. In a 1961survey of tests used by psychologists in theUSA, the MMPI was the only nonprojectivemeasure mentioned among the topten most used tests. Of the top ten tests,five were intelligence tests and four wereprojective measures (Sundberg, 1961). Aconfluence of circumstances, includingthe expansion of clinical psychology practiceduring and after World War II, andthe emergence of an extensive researchbase led to almost immediate acceptanceof this self-report personality inventory(Kleinmuntz, 1967). Further, the MMPIwas one of the first tests to gain popularitywith others outside of the mental healthprofessions (see Box 1.1). However, thispopularity led to significant friction anddisagreements over the relative meritsof the MMPI, and its objective methods,compared with the popular projectivetechniques. This tension is reflected in thecomments of Paul Meehl that are summarizedin Box 1.2.The MMPI (Hathaway & McKinley,1942) differed from its predecessors (such

CHAPTER 1 HISTORICAL TRENDS13Box 1.1Sample Items from 1960s MMPISpoofsPersonality testing eventually became popularenough to warrant derision by membersof Congress, well-known humorists such asArt Buchwald, and others. Some of thesealternate MMPI items were published in a1965 issue of American Psychologist (p. 990)to poke fun at this method of personalityassessment.When I was younger I used to tease vegetables.I think beavers work too hard.I use shoe polish to excess.When I was a child I was an imaginary playmate.Box 1.2Meehl on Science and TechnicsPaul Meehl is considered one of the foundersof modern personality assessment and diagnosticpractice. His 1973 collection of selectedpapers published by the University of MinnesotaPress provides a unique glimpse of thegenius of an astute clinician. In the followingquote, Dr. Meehl discusses the tension betweenscience and practice in psychology and takes astance against theoretical dogmatism:Doubtless every applied science (orwould-be science) presents aspects of thisproblem to those working at the interfacebetween science and technics, as isapparent when one listens to practicingattorneys talking about law professors,practitioners of medicine complainingabout medical school teaching, real engineersin industry poking ambivalent funat academic physicists, and the like. So Ido not suggest that the existential predicamentof the clinical psychologist is uniquein this respect, which it certainly is not.But I strongly suspect that there arefew if any fields of applied semiscientificknowledge in which the practitionerwith scientific interests and trainingis presented daily with this problem inone guise or another, or in which its poignancy,urgency, and cognitive tensionsare so acute. I am aware that there aresome clinical psychologists who do notexperience this conflict, but I have met,read, or listened to very few such duringthe thirty years since I first began workingwith patients as a clinical psychologytrainee. Further, these rare exceptionshave seemed to me in every case to beeither lacking in perceptiveness and imaginationor, more often, deficient in scientifictraining and critical habits of mind.When I encounter a hard-nosedbehaviorist clinician who knows (for sure)that Freud’s theory of dreams is 100 percenthogwash and is not worth five hoursof his serious attention; or, toward theother end of the continuum, when I conversewith a devoted Rorschacher whoknows (for sure) that the magic inkblotsare highly valid no matter what the publishedresearch data indicate-I find bothof these attitudes hard to understand orsympathize with (p. viii).as the Personal Data Sheet) in at least onefundamental way. It was one of the firsttests to use an empirical approach to objectivepersonality test development. Mosttests of the day used a priori or rational–theoretical approaches (Martin, 1988).Rational approaches, as the name implies,depend heavily on the test author’s theoryof personality for many aspects of testconstruction, including item developmentand scoring methods. On the other hand,empirical approaches make greater use ofempirical data to make such decisions (seeChap. 2 for a more detailed discussion ofthis distinction).The MMPI used an item selection methodcalled empirical criterion keying (Anastasi &

14CHAPTER 1 HISTORICAL TRENDSUrbina, 1998). Simply stated, this methodinvolved selecting items that meet an empiricalcriterion. In the case of the MMPI, itemswere selected if they were able to routinelydifferentiate clinical groups from samples of“normal” subjects, and distinguish clinicalgroups from one another. For example, itemsfor the Psychasthenia scale (a scale designedto assess anxiety-related problems such asobsessions and fears) were selected basedon a clinical group of 20 cases, the results ofwhich were compared with “normals” andother clinical groups to identify items thatbest differentiated the target clinical groupfrom the others.The original version of the MMPI consistedof 550 statements printed on separatecards. The cards were separated by thepatient into three categories: true, false,and cannot say. The first MMPI clinicalscales were linked to the major diagnosticnosology of the day (Kleinmuntz, 1967),which is another factor that contributedto its popularity. The ten clinical scales ofthe original version included are providedin Table 1.3. The MMPI has undergoneTable 1.3 The Original Scales from theMMPIClinical scalesHypochondriasisDepressionHysteriaPsychopathic deviateMasculinity–FemininityParanoiaPsychastheniaSchizophreniaHypomaniaSocial IntroversionValidity scalesQuestion scaleLie scaleF scaleCorrection scaleSource: Kleinmuntz, 1967.many changes since its inception, with themost recent edition entitled the MMPI-2. Infact, some of the scale names (e.g., Psychasthenia)had fallen into disuse at about the timeof original publication (Kleinmuntz, 1967). Achronology of MMPI developments is listednext and a thorough discussion of this importantmeasure is provided in Chapter 6.MMPI VersionMMPI 1942MMPI-2 1989MMPI-A (Adolescent) 1992Publication DateThe “Children’s MMPI”Not surprisingly, the MMPI profoundlyinfluenced child assessment practiceincluding the development of the PersonalityInventory for Children (PIC) inthe 1950s. The PIC was based on a poolof 600 items; hence, it was comparable inlength to the MMPI. A central differencebetween the MMPI and the PIC was theinformant. The PIC was not a self-reportmeasure. Instead, a parent rated the child’sbehavior. Lachar (1990) gave the followingrationale for this decision:Selection of the parents as the source of PICtest responses helps overcome two of themajor obstacles posed by requesting the referredchild or adolescent to respond to numerousself-report descriptions in order toobtain a multiple-scale objective evaluation.The majority of children seen by mentalhealth professionals in a variety of settingsappear for such an evaluation because oftheir noncompliant behaviors and/or documentedproblems in academic achievement,most notably in the development of readingskills. Therefore, it seems unlikely thata technique requiring such children to readand respond to a large set of self-descriptionswill find broad acceptance in routine clinicalpractice (p. 299).

CHAPTER 1 HISTORICAL TRENDS15The scales of the PIC were derived usingfactor-analytic methods. Thus, the PIC,like the MMPI, was developed with aheavy emphasis on empirical methods (seeChap. 6). In the 1960s, empirical methodsof test development were also applied tothe development of other types of childassessment devices.Rating ScalesParent and teacher ratings of children’sbehavior and emotions trace their rootsto the assessment of adult psychopathologyin hospital settings. Conceptualizedas one type of observational method, ratingscales were developed in the 1950sfor use by nurses and other caretakerswho worked closely with patients forextended periods of time. One of the firstsuch measures was the Wittenborn PsychiatricRating Scales (1955). Accordingto Lorr (1965), the scales were designedfor recording currently observable behaviorand symptoms in hospitalized mentalpatients. The Wittenborn could be completedby a social worker, psychologist,psychiatrist, nurse, attendant, or otherindividual familiar with the patient’s dayto-daybehavior. The original scale consistedof 52 symptoms that were combinedto yield 9 scores for acute anxiety, conversionhysteria, manic state, depressedstate, schizophrenic excitement, paranoidcondition, paranoid schizophrenic,hebephrenic schizophrenic, and phobiccompulsive. An item assessing withdrawalincluded the following options:No evidence of social withdrawalDoes not appear to seek out the company ofother peopleDefinitely avoids peopleThe Wittenborn was used for diagnosticpurposes as well as for the design and evaluationof treatment (Kleinmuntz, 1967).Reviewers of the day found many reasonsto recommend the Wittenborn, includinga thorough research base (Eysenck, 1965)and easy administration and scoring (Lorr,1965). There was considerable concern,however, about overlapping scales. Thehebephrenic schizophrenic and schizophrenicexcitement scales correlated at .88and the paranoid condition and paranoidschizophrenic scales correlated at .79. Onthe basis of these data, Eysenck (1965) andLorr (1965) recommended that these scalesbe combined to reflect this overlap.Other rating scales of adult psychopathologyfor use in inpatient settings included theHospital Adjustment Scale (McReynolds,Ballachey, & Ferguson, 1952) and the InpatientMultidimensional Psychiatric Scale(Lorr, 1965), a rating of symptomatologycompleted by the clinician after a diagnosticinterview. Such measures probably fell intodecline for many reasons, one of the mostprominent being the deinstitutionalizationmovement of the 1970s. These instrumentsdid, however, clearly demonstrate the utilityof ratings of behavior as practical and usefulassessment tools. These scales set the stagefor the development of parent and teacherrating scales of child behavior.Internalizing and ExternalizingDimensionsResearch into the diagnosis of child psychopathologyled to increased attention tothe use of rating scales for child diagnosis.In a 1978 article in Psychological Bulletin,Thomas Achenbach and Craig Edelbrockintroduced many clinicians to the termsinternalizing and externalizing psychologicaldisorders of childhood. These dimensions,or types of child psychopathology,were based on an extensive empirical analysis(typically using factor analysis) of parentand teacher behavior problem ratingscales. Children experiencing adjustmentdifficulties of the internalizing variety havealso been described as over-controlled,with problems of inhibition, anxiety, and,perhaps, shyness (Edelbrock, 1979). On

16CHAPTER 1 HISTORICAL TRENDSthe other hand, children with externalizingproblems have been described asundercontrolled with difficulties such asaggression, conduct problems, and actingoutbehavior (Edelbrock, 1979).These two dimensions of child psychopathologytrace their roots to the workof Peterson (1961), who labeled the syndromesas conduct problem (externalizing)and personality problem (internalizing). Theveracity of the broad internalizing andexternalizing categorizations of child psychopathologyis supported by many factor-analyticinvestigations of both parentand teacher rating scales alike (Edelbrock,1979). The utility of these behavioral distinctionswas also demonstrated in an earlystudy of 163 consecutive referrals to a childpsychiatry outpatient department (Cohen,Gotlieb, Kershner, & Wehrspann, 1985).Children were classified as externalizersand internalizers based on the AchenbachChild Behavior Checklist (CBCL; a parentreport form) and the Teacher Report Form(TRF) (see Chap. 7). The resulting analysesuncovered distinct differences betweenthe two groups, particularly on the TeacherForm. Internalizers were found to be moreintelligent, better readers, less egocentric,and they used more adaptive means of copingwith stressful situations. Internalizerswere also generally rated as being less disruptivethan externalizers.Numerous independent researchstudies, many of which have been conductedinternationally (Ivanova et al.,2007), have demonstrated strong factoranalyticsupport for these two types ofchild behavioral adjustment. This preponderanceof evidence, supported byother types of validity evidence, hasresulted in these factors serving as thefoundation for the development of manyteacher and parent rating scales, mostnotably the Achenbach Child BehaviorChecklist (Achenbach, 1991b) and theBehavior Assessment System for Children(BASC-2; Reynolds & Kamphaus,2004). Furthermore, the terms internalizingand externalizing are now a part ofpsychologists’ everyday parlance whendiscussing child behavior problems.The Diagnostic andStatistical Manualof Mental DisordersDiagnostic SystemsDiagnostic systems have had a profoundimpact on child assessment by definingsymptoms and other diagnostic indicesthat have subsequently been incorporatedinto various assessment methods. The mostobvious link exists between the various editionsof the Diagnostic and Statistical Manualof Mental Disorders (currently DSM-IV-TR;American Psychiatric Association, 2000)and structured interview methods designedto assess symptomatology associated withvarious DSM diagnostic categories (seeChap. 11). Given this interdependence, athorough knowledge of the nature of theDSM and its variants is prerequisite to thestudy of child assessment.As mental disorders became recognizedas conditions worthy of medical treatment,the need for diagnostic systems becamemore pressing. Consistent diagnosis wasnecessary for communication among cliniciansand for the conduct of epidemiologicalresearch and other scientific investigations.The American Medico-PsychologicalAssociation (now the American PsychiatricAssociation) began efforts to standardizediagnostic procedures in 1917. The firstdiagnostic manual, a classification of mentaldisease, was produced by the AmericanPsychiatric Association in conjunction withthe U.S. Census Bureau (Widiger, Frances,Pincus, Davis, & First, 1991).The first edition of the DSM (Diagnosticand Statistical Manual) appeared in 1952.Part of the impetus for the creation and

CHAPTER 1 HISTORICAL TRENDS17frequent updating of the DSM has beenprovided by the International StatisticalClassification of Diseases, Injuries, and Causesof Death (ICD). The ICD, currently ICD-11,is published by the World Health Organization.The DSM has been revised both tocoordinate with the ICD and to add criteriafor conditions that are of concern to USclinicians, and delete conditions that arenot apparent in the USA (Widiger et al.,1991). The DSM has also been revisedbecause of a desire to make the diagnosticcategories more evidence-based. Priorto the development of the DSM-III, thesystem was based primarily on the expertjudgment of a relatively small number ofclinicians. The DSM-II, for example, wasfinalized after review by 120 psychiatristsin February of 1967 (Widiger et al., 1991).The DSM-IV (APA, 1994) was based ona more comprehensive research base thanany of its predecessors. According to Widigeret al. (1991), three research methodshave formed the empirical cornerstone forthe development of DSM-IV.1. Literature reviews: Comprehensivereviews of the research were completedto advise the various committeescharged with proposing diagnostic criteriafor conditions. These reviews wereseen as a way to mitigate against biaseson the part of some committees (Widigeret al., 1991).2. Data reanalyses: Existing data sets weremade available to the DSM-IV committeessupported by funding from theJohn D. and Catherine T. MacArthurFoundation. According to Widigeret al. (1991), these data set reanalysesallowed the committees to evaluatethe validity of current diagnostic algorithmsand pilot-test new proposals formaking diagnoses.3. Field trials: These studies were particularlyuseful for testing the reliability andvalidity of diagnostic categories (Widigeret al., 1991).Table 1.4 Chronology of Diagnostic SystemsDeveloped Under the Auspices of theAmerican Psychiatric Association1917 Classification of Mental Disease1933 Standard Classified Nomenclatureof Disease1952 Diagnostic and Statistical Manualof Mental Disorders I (DSM-I)1968 Diagnostic and Statistical Manualof Mental Disorders II (DSM-II)1980 Diagnostic and Statistical Manualof Mental Disorders III (DSM-III)1987 Diagnostic and Statistical Manualof Mental Disorders III-Revised(DSM-III-R)1994 Diagnostic and Statistical Manualof Mental Disorders IV (DSM-IV)2000 Diagnostic and Statistical Manualof Mental Disorders IV-Text Revision(DSM-IV-TR)The DSM-IV-TR, because of its greaterreliance on empirical methods, has had aneven more substantial impact on the personalityassessment process (see Chap. 3).The chronology of the DSM is provided inTable 1.4.IDEA and SpecialEducationThe 1974 Education of Handicapped Children’sAct, better known as Public Law94–142 (IDEA), and its reauthorization,the Individuals with Disabilities EducationImprovement Act (IDEIA), mandatedspecial education and related services forchildren classified as having an emotionaldisturbance. As a result, some method hadto be developed to define child problemsand determine children’s eligibility for specialeducation services. Under IDEA, andthe subsequent IDEIA with few substantive

18CHAPTER 1 HISTORICAL TRENDSchanges, the classification of severe emotionaldisturbance was defined as follows:The term means a condition exhibiting oneor more of the following characteristics overa long period of time and to a marked degreewhich adversely affects school performance:(a) an inability to learn which cannot be explainedby intellectual, sensory, or healthfactors; (b) an inability to build or maintainsatisfactory relationships with peers andteachers; (c) inappropriate types of behavioror feelings under normal circumstances; (d)a general pervasive mood of unhappinessor depression; or (e) a tendency to developphysical symptoms or fears associated withpersonal or school problems (Federal Register,1999).With its passage in the 1970s, this laweffectively mandated US public schoolsto identify and serve children withbehavioral or emotional problems, manyof which had previously been educated ina variety of settings, including residentialtreatment programs or state mental hospitals.Consequently, these laws expandedschool-based diagnostic practices toinclude evaluation for the presence of“emotional disturbance,” just as had beenmore commonly done for developmentaland learning disorders. These federalmandates also enhanced the popularityof rating scales (particularly teacher ratings)as assessment methods of choice inmany school systems.The IDEA nosology of emotional disturbance(the word severe has now beenremoved) has long been the target of criticismsthat it is invalid, restrictive, or otherwiseflawed (Forness & Kritzer, 1992).Bower (1982), the recognized developerof the conceptual basis of the IDEA diagnosticcategories, raised similar questionsabout the system. He noted that:Section ii [which excludes the socially maladjustedfrom the IDEA act] is, one wouldguess, a codicil to reassure traditional psychopathologistsand budget personnel thatschizophrenia and autism are indeed seriousemotional disturbances on the one hand, andthat just plain bad boys and girls, predelinquents,and sociopaths will not skyrocket thecosts on the other hand. It is clear what thesemodifications and additions were intendedto do. It is perhaps not clear what such publicpolicy and fiscal modifications do to theconceptual integrity of the definition and thenature and design of its goals (p. 56).Despite such controversy, the IDEA classificationsystem remains as the “gold standard”nosology when determining childeligibility for oft-costly special educationand related services.Constructs (Dimensions)vs. CategoriesThe measurement problems associatedwith categorical diagnostic systems suchas the DSM are well known (Kamphaus& Campbell, 2006), and the advantagesof using dimensional methods are recognizedas well. Achenbach and McConaughy(1996), for example, noted thatthe yes/no nature of categorical methodsdoes not necessarily account for childrenwhose problems vary in degree orseverity. As a result, the nexus betweennormality and psychopathology cannotbe well understood with categoricalmethods, since most high prevalenceproblem behaviors of childhood, suchas inattention and hyperactivity, arenot classifiable when below diagnosticthreshold levels. Substantial evidence isemerging to suggest that child behaviorproblems such as inattention, hyperactivity,depression, and conduct problems,in fact, fall along continua in thepopulation (Hudziak et al., 1998; Scahillet al., 1999).As a result, dimensional classificationmethods have demonstrated their usefulness

CHAPTER 1 HISTORICAL TRENDS19in the study of psychopathology. Forexample, dimensional approaches havedemonstrated more predictive validitythan categorical approaches (Fergusson& Horwood, 1995), as well as statisticalreliability (Cantwell, 1996). Such methodsalso minimize the need for clinicaljudgment and inference (Haynes &O’Brien, 1988), provide greater sensitivityto the presence of comorbid conditions(Caron & Rutter, 1991), and havethe ability to depict multiple symptompatterns in a given individual simultaneously(Cantwell, 1996). Furthermore,the use of dimensional, person-orientedapproaches to identify subtypes or clustersof individuals may lead to moreefficient, streamlined subtype-specificintervention and prevention services(Achenbach, 1995; Bergman & Magnusson,1997). The overlap and tensionsbetween categorical and dimensionalclassification methods will be elucidatedthroughout this text. It is important todo so because a merger of these methodologiesis likely in the DSM-V and otherfuture diagnostic systems (Rounsavilleet al., 2002).Future Trendsassessment. The developments outlinedabove encompass observable behavior,structured and unstructured use of testsand interviews, and assessment of broadandnarrow-band aspects of personality, allwithin the context of a person’s situation/environment. Thus, in spite of variouscriticisms and some apparent decrease inthe use of personality assessment instruments(they were referring to projective devicesprimarily), all indications point to vigorousactivity in the area that promises tocontinue.” (p. 23; italics added).In the past, the technology of personalityassessment has been viewed as laggingbehind other areas of assessment, suchas intelligence and achievement testing(Martin, 1998). This conclusion is nolonger true. New measurement sciencerigor is being applied to the developmentof behavioral rating scales, interviewmethods, and diagnostic systems.Two trends of the past few decades arecontinuing; relatively less emphasis ontraining in projective methods (Belter &Piotrowski, 1999) and increased use ofrating scales (Archer & Newsom, 2000).In fact, a veritable explosion in the creationand publication of behavior ratingscales alone necessitated creation of this,our third edition of this text.The pace of change in personality assessmentis ever hastening. There is increasinginterest in the development of newchild assessment methods, providingclinicians with a wide array of assessmentoptions. In 1990, Tuma and Elbert(1990) identified test development andresearch trends that remain true to thepresent day.It is apparent that personality assessmentis undergoing rapid development in allareas: projective, objective, and behavioralassessment; clinical interviewing andinformal assessment; and environmentalChapter Summary1. Personality is typically considered to becomposed of traits, a more enduring setof characteristics of the individual.2. Formal personality measures emerged as alogical outgrowth of other efforts to measureindividual differences, most notablythe experimental methods of Wundt,Galton, and others.3. The Woodworth Personal Data Sheetwas published in 1918 as a result of the

20CHAPTER 1 HISTORICAL TRENDSsurge of interest in testing potentialsoldiers.4. The needs for diagnosis created by WorldWar I and World War II provided consider-ableimpetus for the development ofpersonality tests.5. A major assumption underlying projectivetesting is that the use of stimuli thatare prone to a variety of interpretationswill encourage clients to reveal informationthat they otherwise would notshare in response to direct questioning.6. The Rorschach test stimuli were originallyinkblots placed on paper that wasthen folded in half.7. The use of projective techniques withchildren dates back to the early part ofthis century, when Florence Goodenoughbegan to study children’s humanfigure drawings.8. The MMPI was one of the first tests touse an empirical approach for personalitytest development and used an itemselection method called empirical criterionkeying.9. The use of informant rating scales forthe assessment of child psychopathologytraces its roots to the assessment of adultpsychopathology in hospital settings.10. In a 1978 article in Psychological Bulletin,Achenbach and Edelbrock introducedthe terms internalizing and externalizingwhen referring to psychological disordersof childhood.11. The first edition of the DSM (Diagnosticand Statistical Manual) appeared in1952.12. The most recent edition of the manual,the DSM-IV-TR, is based on a morecomprehensive research base than anyof its predecessors.13. The 1974 Education of all HandicappedChildren’s Act, better knownas Public Law 94–142, and its reauthorization,the Individuals withDisabilities Education ImprovementAct (IDEIA), have mandated specialeducation and related services forchildren classified as emotionally disturbed.

C h a p t e r 2Measurement IssuesChapter Questionsl What type of information is yieldedfrom a T-score?l How does skewness affect scaling decisions?l How has factor analysis been used todevelop personality tests and diagnosticschedules?Users of instruments assessing personalityand other aspects of behavioral, emotional,and social functioning should havea thorough understanding of measurementprinciples. The discussion that follows,however, hardly qualifies as thoroughbecause measurement instruction is notthe purpose of this book. This chaptermerely points out some of the most importantmeasurement concepts for conductingassessments of youth.We assume that the user of this texthas had, at a minimum, undergraduatecourses in statistics, tests and measurements,as well as at least one graduatelevelmeasurement course. If a user ofthis text is not acquainted with someof the principles discussed here, then astatistics and/or measurement textbookshould be consulted. There are a numberof excellent measurement textbooksavailable, including Anastasi and Urbina(1998) as well as Allen and Yen (1979).The reader is also referred to the Codeof Fair Testing Practices in Education(Joint Committee on Testing Practices,2004) for a discussion of the appropriateprocedures for test development, testselection, scoring, interpretation, andcommunication of results.This chapter begins by defining thenature of the tests that assess psychologicalconstructs. Then, a review of basic principles21P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_2, © Springer Science+Business Media, LLC 2010

22 CHAPTER 2 MEASUREMENT ISSUESof statistics and measurement is presented,including topics ranging from measures ofcentral tendency to factor analysis. The lastpart of the chapter introduces measurementissues that are specific to the use and interpretationof personality tests and similarinstruments.Defining Personality TestsThere is a plethora of methods, includingtests, designed to assess similar-soundingpsychological constructs, including personalityscales, behavior rating scales, anddiagnostic schedules. The available personalitymeasures differ to such an extentthat they can be subtyped in order to clarifytheir psychometric properties. A definitionfor a psychological test, taken from an early,well-known personality assessment text,may be a good starting point. Kleinmuntz(1967) defines a psychological (includingpersonality) test by observing, “A psychologicaltest is a standardized instrument orsystematic procedure designed to obtain anobjective measure of a sample of behavior”(pp. 27–28). This rather broad definitionprovides a useful starting point for conceptualizingthe great variety of measuresavailable.The central characteristic of this definitionis the notion of standardization ofbehavioral sampling. Standardization has atleast two meanings: standardization in thesense of collecting a sample for the purposeof norm referencing and standardization asadministration of the measure accordingto a consistent set of rules. Most of themeasures discussed in this volume fit thefirst notion of standardization in that theyare norm-referenced. That is, these measuresuse norm groups for gauging a child’sperformance in comparison to some referencegroup. Furthermore, the principleof administration structure or consistencyapplies to all of the measures in this text.For example, respondents should completethe measure in an environment free of distractionsand should clearly comprehendthe response-format (e.g., true/false, frequencyratings, etc.) and time frame (e.g.,the last 6 months) referenced by the test.Standardized procedure emanates fromexperimental psychology, where laboratorycontrol is central to obtaining reliableand valid results (Kamphaus, 2001). Similarly,in the case of personality assessment,standardized administration procedureis necessary to produce reliable and validmeasurements of behavior.All psychological tests take a sample ofbehavior from which the findings are subsequentlygeneralized (Anastasi & Urbina,1998). This ability to generalize findings isthe central strength of psychological testsand is probably the reason for their widespreaduse. Without these tests, psychologicalmeasurement would be impracticalbecause of the time and expense required.Of course, a sample can always be in error, afact that should always be considered wheninterpreting results (Dahlstrom, 1993).Types of TestsHow does one identify an instrument thatassesses personality or behavioral, social, andemotional functioning? Personality tests havetraditionally attempted to assess personalitytraits such as introversion, agreeableness, andanxiety. As noted in Chap. 1, traits are usuallyconsidered to be relatively stable characteristicsof the individual (Martin, 1988). Forchildren and adolescents, such characteristicsmay be similarly conceptualized underthe term personality traits or, typically foryounger children, temperament. Research hasclearly indicated that individual differencesin a variety of personality domains in youthare measurable (see Shiner & Caspi, 2003),and relatively stable (e.g., Durbin, Hayden,Klein, & Olino, 2007; Hampson, Andrews,Barckley, & Peterson, 2007).

CHAPTER 2 MEASUREMENT ISSUES23Rating scales, one of the most popularchild assessment methods, may fall into yetanother category of test called diagnosticschedules (Kamphaus et al., 1995). Kamphauset al. define a diagnostic schedule as aspecialized psychometric method that providesa structured procedure for collectingand categorizing behavioral data that correspondto diagnostic categories or systems.A diagnostic schedule, then, is not designedto assess a trait, but rather to diagnose a syndrome.How does one identify a diagnosticschedule? One clue is the source of the itempool. The Children’s Depression Inventory(CDI; Kovacs, 1992) is a popular measureof childhood depression that used the DSMas its item source. It was designed to assessthe symptoms of depression in order toassist with making the diagnosis of depression.It was not designed to assess a stablepersonality trait or temperament but ratherto allow the examiner to make the diagnosisof depression with confidence. In fact, a cutscore that indicates the possible presence ofclinical depression is often used for interpretingscores (Kovacs).However, adding further complexityto understanding how rating scalesfit within the array of tools availablefor clinical assessments is the fact thatmany widely used rating scales cannotbe considered diagnostic schedules. Forexample, although the Behavior AssessmentSystem for Children-2 (BASC-2;Reynolds & Kamphaus, 2004) assessesclinically relevant domains for youth(e.g., hyperactivity, aggression, depression,anxiety), elevated scores on thosedomains do not necessarily mean that theindividual being assessed meets the criteriafor a corresponding diagnosis. Suchrating systems routinely have items thatdo not directly map onto the diagnosticcriteria. Rather, the content of thesescales may be indicative of aspects ofthe young person’s functioning that maylend themselves to recommendationsfor intervention, as well as help signala diagnosis or impairment in a particulardomain. Furthermore, some rating scalesmay blend the elements of personalitytests and diagnostic schedules, makingsound clinical judgment essentialin drawing the most sound conclusionsfrom the data collected. The primarypurpose of the assessment (e.g., diagnosticclarification vs. identifying areas ofbehavioral, social, or emotional concern)should guide the selection of diagnosticschedules and/or behavior rating scales.Further, we would argue that if given achoice, clinicians should initially seektools that provide a broad screening of avariety of possible problems rather thannarrowing in too quickly on a specificdiagnosis.Mash and Hunsley (2005) have articulatedthe problems with considering afocus on specific diagnoses as synonymouswith psychological assessment:“Although formal diagnostic systems…provideone alternative for framing the rangeof disorders and problems to be considered,there is no need to limit the range ofproblems to those detailed in a diagnosticsystem. Refraining from excessive relianceon formal diagnostic systems is warrantedgiven the well-documented shortcomings inthe nature and development of such systems(e.g., Beutler & Malik, 2002; Mash & Dozois,2003; Scotti, Morris, McNeil, & Hawkins,1996) and the lack of evidence that suchdiagnostic systems provide the best wayto match a treatment to a child (Bickman,2002)” (p. 368).Despite these concerns, diagnostic schedulesor checklists may still play a criticalrole in helping to address a referral questionand make treatment recommendationsthat are diagnostically-relevant (e.g.,classroom accommodations for a childwho meets criteria for ADHD). Diagnosticschedules have evolved from behavioralassessment methods, as has the DSM,which now emphasizes the tally of behaviors

24 CHAPTER 2 MEASUREMENT ISSUES(symptoms) in order to make a diagnosticdecision. Personality tests and many ratingscales, on the other hand, are rooted in thepsychometric tradition in which such testsare designed to assess traits across a continuum.While such instruments may not leaddirectly to a diagnostic decision, as notedabove, they can play other important rolesby identifying traits that have implicationsfor the course or prognosis of a disorder, oreven for treatment.While diagnostic schedules are practicalfor making diagnostic decisions, suchmeasures have limitations for studyingthe nature of individual differences or forcontributing to other important aspectsof the assessment process. These limitationsare inherent in diagnostic schedulesbecause they often lack a clear theoreticalbasis or evidence of a priori defined trait(s)that can be supported with construct validityevidence. Therefore, the emergence ofdiagnostic schedules as the instruments ofchoice for much of assessment practice isevidence of the profound impact of behavioral-baseddiagnostic systems on psychometrictest development, particularly overthe last decade or two, as well as the (realor perceived) need to provide diagnoses asa result of all assessments due to managedhealth care.Appropriate conclusions that couldbe drawn based on diagnostic schedulesinclude statements like the following:l Tonya suffers from major depression,single episode, severe.l Tony exhibits nearly enough symptomsto be diagnosed as having conduct disorder.l Traci has attention problems that areworse than those of 99% of the childrenher age.Alternatively, conclusions that could beoffered based on psychometric tests of personalityor behavioral, emotional, or socialfunctioning could include:l Allison shows evidence of poor adaptabilityto new situations and changes inroutine, which puts her at risk for schooladjustment problems.l Patrick’s high score on the sensationseeking scale warrants consideration aspart of his vocational counseling andeducational planning.l Maria’s somatization tendencies revealthe need for counseling in order toreduce her frequency of emergencyclinic visits.l Andersen’s apparent signs of depressionindicate a need for further evaluation andintervention.A central difference between these interpretivestatements is that those made basedon diagnostic schedules are dependent ondiagnostic nosologies. A variation of thispremise is the third statement exemplifyingdiagnostic-based conclusions, whichmay result from a norm-referenced behaviorrating scale that has a scale devoted toinattention. Such norm-based informationcan typically be gleaned from personalitytests or other rating scales as well.The interpretive statements made basedon psychometric tests, however, can beoffered independently of diagnosis. Theseconclusions are based on the measurementof traits or tendencies that may or may notrepresent diagnostic symptoms or signs,and yet, these conclusions contribute substantiallyto the assessment process.Widely used rating scales such as thoseto be discussed later in this volume haveseveral scales with the same name as adiagnostic category such as depression oranxiety. At the same time, such measuresare scaled similarly to traditional personalitytests with standard scores based onnorms.Although research is emerging onthis issue (e.g., Ferdinand, 2008; Kerr,Lunkenheimer, & Olson, 2007), generallyspeaking, we do not know the extentto which these scales demonstrate the

CHAPTER 2 MEASUREMENT ISSUES25stability associated with traits or the diagnosticaccuracy of the DSM system. Theirpopularity for clinical practice, however,continues to increase due to their costeffectiveness and time efficiency. Furthermore,rating scales allow for the rapid andaccurate identification of domains of deviantbehavior that may require diagnosis ortreatment (Hart & Lahey, 1999).In this volume, the term personality testwill occasionally be used generically toapply to personality trait measures, diagnosticschedules, syndrome scales, andrelated measures, always assuming that thereader is aware of the distinctions betweensubtypes of measures.Scores, Norms,and DistributionsTypes of ScoresIn this section, some of the basic propertiesof score types are reviewed, with particularemphasis on the T-score standard scoremetric and its variants. The properties ofthese scores will be highlighted in order toencourage psychometrically appropriatescore interpretation.Raw ScoresThe first score that the clinician encountersafter summing item scores is usuallycalled a raw score. Raw scores, on mosttests, are simply the sum of the item scores.The term raw is probably fitting for thesescores in that they give little informationabout a child’s performance as compared tohis or her peers. Raw scores are not particularlyhelpful for norm-referenced interpretation.Raw scores merely identify thenumber of behaviors or symptoms present,not how deviant this amount of symptomatologyis from the norm nor how impairingit is for the individual.Norm-Referenced ScoresPersonality test interpretation often focuseson norm-referenced interpretation, the comparisonof children’s scores to some standardor norm. For the purposes of assessingpsychological constructs, scores are usuallycompared to those of children the sameage. Norm-referenced achievement tests,by contrast, may compare children’s scoresto those of others in the same grade, andcollege admission counselors may comparean incoming student’s GPA to that of freshmenwho entered the year before.Norm referencing is of importancein personality and behavioral assessmentbecause it allows the clinician to gauge deviance,which is often central to the referralquestion. Parents who refer a child for apsychological evaluation often have normreferencingin mind. They ask questionssuch as “Is her activity level normal for herage?” or “Everyone says he is just a boy,but fire setting isn’t normal, is it?” Normreferencingallows the clinician to answersuch questions objectively. The remainingscores discussed in this section are normreferencedscores that allow the clinician tomake these important comparisons.Standard ScoresThe standard score is a type of derived scorethat has traditionally been the most popularfor psychometric test interpretation. Standardscores convert raw scores to a distributionwith a set mean and standard deviationand with equal units along the scale (Anastasi& Urbina, 1998). The typical standardscore scale used for personality tests andbehavior rating scales is the T-score, whichhas a mean of 50 and standard deviation of10. Another popular standard score that iscoming into more frequent use for personalitytest interpretation has the mean setat 100 and the standard deviation at 15,similar to the IQ metric (see Table 2.1).Because they have equal units along the

26 CHAPTER 2 MEASUREMENT ISSUESTable 2.1 Standard Score, T-Score, Scaled Score, and Percentile Rank Conversion TableStandardScoreM = 100SD = 15T-ScoreM = 50SD = 10ScaledScoreM = 10SD = 3PercentileRankStandardScoreM = 100SD = 15T-ScoreM = 50SD = 10ScaledScoreM = 10SD = 3PercentileRank160 90 99.99 128 69 97159 89 99.99 127 68 97158 89 99.99 126 67 96157 88 99.99 125 67 15 95156 87 99.99 124 66 95155 87 99.99 123 65 94154 86 99.99 122 65 92153 85 99.98 121 64 92152 85 99.97 120 63 14 91151 84 99.96 119 63 89150 83 99.95 118 62 88149 83 99.94 117 61 87148 82 99.93 116 61 86147 81 99.91 115 60 13 84146 81 19 99.89 114 59 83145 80 99.87 113 59 81144 79 99.84 112 58 79143 79 99.80 111 57 77142 78 99.75 110 57 12 75141 77 99.70 109 56 73140 77 18 99.64 108 55 71139 76 99.57 108 55 69138 75 99 107 55 67137 75 99 106 54 65136 74 99 105 53 11 65135 73 17 99 104 53 62134 73 99 103 52 57133 72 99 102 51 55132 71 98 101 51 52131 71 98 100 50 10 50130 70 16 98 99 49 48129 69 97 98 49 45(Continues)

CHAPTER 2 MEASUREMENT ISSUES27Table 2.1 (Continued)StandardScoreM = 100SD = 15T-ScoreM = 50SD = 10ScaledScoreM = 10SD = 3PercentileRankStandardScoreM = 100SD = 15T-ScoreM = 50SD = 10ScaledScoreM = 10SD = 3PercentileRank97 48 43 68 29 296 47 40 67 28 195 47 9 38 66 27 194 46 35 65 27 3 193 45 33 64 26 193 45 31 63 25 192 45 29 63 25 191 44 27 62 25 190 43 8 25 61 24 .4989 43 23 60 23 2 .3688 42 21 59 23 .3087 41 19 58 22 .2586 41 17 57 21 .2085 40 7 16 56 21 .1684 39 14 55 20 1 .1383 39 13 54 19 .1182 38 12 53 19 .0981 37 11 52 18 .0780 37 6 9 51 17 .0679 36 8 50 17 .0578 35 8 49 16 .0478 35 7 48 15 .0377 35 6 48 15 .0276 34 5 47 15 .0275 33 5 5 46 14 .0174 33 4 45 13 .0173 32 3 44 13 .0172 31 3 43 12 .0171 31 3 42 11 .0170 30 4 2 41 11 .0169 29 2 40 10 .01

28 CHAPTER 2 MEASUREMENT ISSUES0.13 % 2.14 % 13.59% 34.13% 34.13% 13.59% 2.14 % 0.13 %-4SD -3SD -2SD -1SD Mean +1SD +2SD +3SD +4SDTest Scorez scoreI-4I I I I I I I I-3 -2 -1 0 +1 +2 +3 +4T scoreI I I I I I I I I10 20 30 40 50 60 70 80 90PercentileRankI I I I I I I I I I I I I1 5 10 20 30 40 50 60 70 80 90 95 99Figure 2.1A normal distribution of scoresscale, standard scores are useful for statisticalanalyses and for making comparisonsacross tests. The equal units (or intervals)that are characteristic of standard scoresare shown for various standard scores andpercentile ranks in Table 2.1. This tablemay also be useful for converting a scorefrom one scale to another. In the T-scoremetric, the distance between 20 and 30 isthe same as that between 45 and 55.Standard scores are particularly usefulfor test interpretation because they allowfor comparisons among various subscales,scales, and composites yielded by the sametest, allowing the clinician to comparetraits. In other words, standard scores allowthe clinician to answer questions such as“Is she more anxious than depressed?” thusfacilitating profile analysis. Most modernpersonality tests use T-scores.In a normal distribution (a frequentlyuntenable assumption in personality andbehavior assessment, as is shown in a latersection), a normalized standard score dividesup the same proportions of the normal curve.However, because many scales on syndromemeasures in particular are heavily skewed(the most frequent scenario is that mostindividuals are not experiencing psychopathologyand a few are, resulting in positiveskewness), some test developers opt for theuse of linear T-scores. Linear T scores maintainthe skewed shape of the raw score distribution,which means that the same T-scoreon different scales may divide up differentportions of the norming sample. Specifically,50% of the norming sample may scorebelow a linear T-score of 50 on the Anxietyscale, whereas 55% of the norming samplemay score below a linear T-score of 50 on the

CHAPTER 2 MEASUREMENT ISSUES29Aggression scale. Essentially, then, the useof linear T-scores makes the relationship ofpercentile ranks to T-scores unique for eachscale.Percentile RanksA percentile rank gives an individual’srelative position within the norm group.Percentile ranks are very useful for communicatingwith parents, administrators,educators, and others who do not have anextensive background in scaling methods(Kamphaus, 2001). It is relatively easy forparents to understand that a child’s percentilerank of 50 is higher than approximately50% of the norm group and lower thanapproximately 50% of the norm group.This type of interpretation works well solong as the parent understands the differencebetween the percentile rank and thepercent of items passed.Figure 2.1 shows that percentileranks have one major disadvantage incomparison to standard scores. Percentileranks have unequal units alongtheir scale. The distribution in Fig. 2.1(and in Table 2.1) shows that the differencebetween the 1st and 5th percentile ranksis larger than the difference between the40th and 50th percentile ranks. In otherwords, percentile ranks in the middle ofthe distribution tend to overemphasize differencesbetween standard scores, whereaspercentile ranks at the tails of the distributiontend to underemphasize differences inperformance (Kamphaus, 2001).Here is an example of how confusingthis property of having unequal units canbe. A clinician would typically describe aT-score of 55 as average. When placed onthe percentile rank distribution, however,a T-score of 55 corresponds to a percentilerank of 69 in a normal distribution ofscores (see Table 2.1). The percentile rankof 69 sounds as though it is higher thanaverage. Examples such as this clearly showthe caveats needed when dealing with anordinal (unequal scale units) scale of measurementsuch as the percentile rank scale.It is important to remember that the ordinalproperties of the scale are due to thefact that the percentile rank merely placesa score in the distribution. In most distributions,the majority of the scores arein the middle of the distribution, causingsmall differences between standard scoresin the middle to produce large differencesin percentile ranks.Uniform T-scoresA uniform T-score (UT) is a special typeof T score that was used for developmentof the MMPI-2 norms (Tellegen & Ben-Porath, 1992). This derived score is aT-score like all other normalized standardscores with the exception that it maintainssome (but not all) of the skewness of theoriginal raw score distributions. The UTis like a normalized T-score in that therelationship between percentile ranks andT-scores is constant across scales, and itresembles a linear T-score metric in thatsome of the skewness in the raw score distributionis retained. The problem of a lack ofpercentile rank comparability across scalesis described by Tellegen and Ben-Porath(1992) in reference to the MMPI-2:“For example, the raw score distribution ofScale 8, Schizophrenia (Sc), is more positively(i.e., right-) skewed than that of Scale9, Hypomania (Ma). This means that a linearT-score of, say, 80 represents different relativestandings on these two scales in the normativesample. For women in the MMPI-2 normativesample, the percentile values of a linear T-scoreof 80 are 98.6 for Scale 8 and 99.8 for Scale 9;for men, the corresponding values are similar,98.6 and 99.7” (p. 145).In order for the UT scale score to have theproperties of percentile rank comparabilityacross scales and reflection of raw score distributionskewness, the UT-score is based

30 CHAPTER 2 MEASUREMENT ISSUESon the average skewness value across all ofthe clinical scales (Tellegen & Ben-Porath,1992). This approach meets the objectivesoutlined by the developers, but it relies onthe assumption that the skewness of theMMPI-2 clinical scales is similar. Thereis, however, evidence that some MMPI-2clinical scales (e.g., Hypochondriasis andSchizophrenia) are far more skewed thanothers (see Tellegen & Ben-Porath). Itappears that the UT is a compromise metricthat meets test development objectiveswhile, at the same time, not addressingcompletely the issue of different skewnessacross scales. More research and clinicalexperience with the UT metric is necessaryto determine whether or not this methodshould be adopted by other test developers.Norm DevelopmentSamplingNorm development is one area in which thetechnology of personality and behavioralassessment has generally lagged behindthat of intelligence or achievement testing(Martin, 1988). Intelligence tests, forexample, have routinely collected stratifiednational samples of children to use asa normative base. Stratification is used tocollect these samples in order to match, tothe extent possible, the characteristics of thepopulation at large. Common stratificationvariables include age, gender, race, geographicregion, community size, and parentalsocio-economic status (SES; Kamphaus,2001). These variables are used presumablybecause they are related to score differences.Of these widely used stratification variables,SES is known to produce the most substantialscore differences on intelligence measures(Kamphaus). The precedent, then,is set for the norming of personality andbehavioral assessment tools.This precedent, however, has not beenfollowed in several important respects. Untilrecently, many relatively popular personalityscales have not done a good job of stratifyingtheir samples. Some norming samples do notcontrol for geographic region, and others failto control for SES. The result is a normativestandard of unknown utility. Whilepoor norming is less likely to be toleratedin intelligence and academic achievementassessment, it is less frequently criticizedor even noted in discussions of personalityassessment. We will, however, note thecharacteristics of norming samples in subsequentsections of this text. This is importantbecause users of personality and behaviortests should know the characteristics of atest’s norming sample in order to make thebest decisions and gauge the amount of confidenceto place in the obtained scores.Intelligence, achievement, and adaptivebehavior tests typically feature interpretationbased on a national norm sample. Incontrast, a national normative standardhas often not been offered for personalitytests. A substantial number of personalitytests offer only local norms, a subset of thenational normative sample. Local normsanswer different questions than do nationalnorms. Hence, their potential utility hasto be evaluated prior to test selection andinterpretation.Local NormsLocal norms, or norms based on a specificpopulation in a specific setting or location,may sometimes be more useful thannational norms, particularly in terms oftheir relevance for the clinician’s work, andin some cases, recency relative to nationalnorms (Elliott & Bretzing, 1980; Petersen,Kolen, & Hoover, 1989). In order for localnorms to be meaningful, however, therange of their usefulness must be definedclearly.Regardless of the use of local or nationalnorms, typical norm-referenced questionsof interest to psychologists are diagnosticones. Common questions might include:

CHAPTER 2 MEASUREMENT ISSUES31l Does Lindsey have attention-deficit/hyperactivity disorder?l Is Jose clinically depressed?l Is Stephanie more anxious than otherchildren her age?One of the goals of diagnostic practiceis consistency, which is fostered by thepublication of diagnostic criteria. Consistentmethods of diagnosis allow cliniciansto communicate clearly with one another.If, for example, Dr. Ob Session in Seattlesays that a patient is suffering from conductdisorder, then Dr. Sid Ego in Atlantawill know what to expect from this adolescentwhen he enters his office for followuptreatment.National norms similarly promote consistency.If a clinician concludes that a childhas clinically significant attention problemsbased on a deviant score on an inattentionscale, then others may reasonably concludethat this child has attention difficultiesthat are unusual for her age. Popular tests,however, may offer different local normsthat can hamper consistent communication.Similarly, local norms also will be lessgeneralizable to the general population ofchildren who do or do not meet criteria fora particular diagnosis. The clinician mustthen balance these disadvantages of localnorms with the potential for local normsto be more relevant to the population withwhich he/she works.Gender-Based NormsPersonality and behavior measures areunusual in that gender-referenced (local)norms are sometimes offered by test developers.This practice is unusual in comparisonto other domains of assessment where,although significant gender differences exist,national combined gender norms are typicallythe only ones provided. Specifically,intelligence, academic achievement, andadaptive behavior scales produce mean scoredifferences between gender groups, but localnorms by gender are rarely offered. Whythen are gender local norms commonlyoffered for personality tests? Tradition couldbe the most parsimonious explanation.When comparing a child to his orher gender group, the effects of genderdifferences in behavior are removed.Another way of expressing this is to say that,when gender norms are utilized, roughlythe same proportion of boys as girls isidentified as having problems. Because, forexample, boys tend to have more symptomsof hyperactivity than girls (DSM-IV, APA,1994), the use of gender local norms woulderase this difference in epidemiology.Gender norm-referencing would identifyapproximately the same percentage of girlsand boys as hyperactive, such that a boywould require more severe symptomatologyto be identified as elevated on hyperactivityrelative to other boys. Depression isanother example of how gender normsmay affect diagnostic rates. Much evidencesuggests that girls express more depressivesymptomatology than boys in adolescence(Weiss & Weisz, 1988). The use of gendernorms for a depression scale would resultin the same number of adolescent boysas girls exceeding a particular cut score,whereas general national norms wouldretain the known greater prevalence amongadolescent girls.Are gender local norms a problem? Notso long as clinicians are clear about thequestions they are asking. A gender normquestion would be “Is Traci hyperactivewhen compared to other girls her age?”whereas a national norm question wouldbe “Is Traci hyperactive in comparison toother children her age?” General nationalnorms are preferred when a diagnosticquestion is asked. An example of a diagnosticquestion is, “Does Frank have enoughsymptoms of depression to warrant a diagnosis?”The DSM-IV diagnostic criteria donot have differing thresholds for boys, so agender norm would be inappropriate.

32 CHAPTER 2 MEASUREMENT ISSUESAge-Based NormsBecause there are substantial differencesacross age groups, intelligence and academicachievement tests routinely offernorms separately by age groups, typicallyusing age ranges of 1 year or less. By contrast,age ranges as large as 5–7 years arefrequently used for personality tests. Thistradition of articulating norms for largerage groups may be attributable to personalitytraits often having smaller normativesamples than intelligence and achievementtests and a lack of age group differencesin personality and behavior characteristics(Martin, 1988). Some data suggestthat the latter explanation may be moreappropriate. That is, differences betweenadjacent age groups are often insignificantfor behavior rating scales, whereas moremeaningful differences only occur overlonger developmental periods (e.g., Reynolds& Kamphaus, 2004).Clinical NormsA more unique norm group is a sample ofchildren who have been previously diagnosedas having a mental health problem.This clinical norm-referenced comparisoncan answer questions such as:l How aggressive is Sheila in comparisonto other children who are receivingpsychological services?l Are Tonya’s psychotic symptoms unusualin comparison to other children who arereferred for psychological evaluation?There is not a clear precedent for the developmentof clinical norms. The relevantdemographic stratification variables havenot been identified, making it difficult tojudge the quality of clinical norms. Shouldclinical norms, for example, attempt tomimic the epidemiology of childhood disordersincluding 10% depression cases, 5%ADHD cases, and so on? Should normsattempt to match the epidemiology of specificdisorders within child clinic-referredpopulations (i.e., include mostly externalizingdisorders)? Should norms be offeredseparately by diagnostic category to offera more exact comparison? Or shouldattempts be made to address each of theseissues?Until such standards emerge, cliniciansshould seek clinical norms that are at leastwell-described. A clear description of thesample will allow the clinician to determineif the clinical norm group has the potentialto answer questions of interest. For example,the clinician who works in an inpatientsetting may have more interest in aclinical sample of inpatients, whereas othersmay prefer that clinical norms be based ona referral population. If the clinical normgroup for a test is not well-described, theclinician cannot meaningfully interpretthe norm-referenced comparisons.The normal curve refers to the graphicdepiction of a distribution of test scoresthat is symmetrical (normal), resemblinga bell. In a normal distribution, there area few people with very low scores (thesepeople are represented by the tail of thecurve on the left in Fig. 2.1), a few withvery high scores (the tail on the right), andmany individuals with scores near the average(the highest point in the curve).When a distribution is normal or bellshaped,as is the case in Fig. 2.1, the standarddeviation always divides up the sameproportion. Specifically, ±1 standard deviationalways includes approximately 68% ofthe cases in a normal distribution, and ±2standard deviations always include approximately95% of the cases. The normal curveis also sometimes referred to as the normalprobability, or Gaussian curve.Normal distributions, however, cannotbe assumed for personality tests orbehavior ratings. While intelligence andacademic tests often produce near-normaldistributions, personality tests oftenproduce skewed distributions. Examples

CHAPTER 2 MEASUREMENT ISSUES33of skewed distributions are shown in Figs.2.2 and 2.3. The distribution depicted inFig. 2.2 is negatively skewed; a positiveskew is shown in Fig. 2.3. A mnemonicfor remembering the distinction betweenpositive and negative skewness is to notethat the valence of the skewness appliesto the tail, when positive is on the rightand negative is on the left.It is understandable that diagnosticschedules and syndrome scales such asbehavior rating scales produce skewed distributions.After all, only a small proportionof the population is experiencing aparticular disorder at some point in time,and the majority of individuals are free ofsuch symptomatology (positive skew). Onthe other hand, it is quite likely that thedistributions for many adaptive skills orFigure 2.2A hypothetical example of a negativelyskewed distribution of scoresFigure 2.3A hypothetical example of a positivelyskewed distribution of scoresbehaviors would be negatively skewed, inthat the majority of the population wouldpossess high levels of such skills, particularlywith age (c.f., Sparrow, Cichetti, &Balla, 2005).The often skewed distributions obtainedfor personality measures, particularly diagnosticschedules, produce more controversyregarding scaling methods. If, forexample, a distribution is heavily skewed,should normalized standard scores (whichforce normality on the shape of the standardscore distribution regardless of theshape of the raw score distribution) orlinear transformations (which maintainthe shape of the raw score distribution) beused? Petersen et al. (1989) maintain that“usually there is no good theoretical reasonfor normalizing scores” (p. 226), andwe concur with this opinion.What differences does the scaling methodmake (i.e., normalized versus lineartransformations)? The primary differenceis in the relationship between the standardscores (T-scores) and percentile ranksyielded by a test. The positively skeweddistribution shown in Fig. 2.3 is a goodexample of how this relationship can beaffected. If this distribution was normalized(i.e., forced normal by converting rawscores to normal deviates and then the normaldeviates to T-scores), then a T-scoreof 70 will always be at the 98th percentile.If linear transformations were used for thescale distribution shown in Fig. 2.3, thenthe corresponding percentile rank wouldmost certainly be something other than 98.Clearly the type of standard score used forscaling a test affects diagnostic and, perhaps,treatment decisions. If normalizedstandard scores were used for a positivelyskewed scale (e.g., one measuring conductproblems), then potentially more childrenwould be identified as having significantproblems. On the other hand, normalizedstandard scores make the clinician’s jobeasier by fostering interpretation acrossscales. Herein lies the debate: Is the inter-

34 CHAPTER 2 MEASUREMENT ISSUESpretive convenience of normalized standardscores worth the trade-off in lack ofprecision?Clinicians will find that many testsuse normalized standard scores (usuallyexpressed in a T-score metric) even whenclear evidence of significant skewnessexists. We suggest that readers note thescaling method used by tests discussed inthis volume as they consider the strengthsand weaknesses of each measure.ReliabilityThe reliability of a test refers to the degreeto which test scores are free from measurementerror and includes the presumedstability, consistency, and repeatability ofscores for a given individual (AERA, APA,NCME, 1999).The reliability of a personality test isexpressed by the computation of a reliabilitycoefficient, which is a special typeof correlation coefficient. One essentialdifference between a reliability coefficientand a correlation coefficient is that reliabilitycoefficients are typically not negative,while negative correlation coefficientsare eminently possible. Reliability coefficientsrange, then, from 0 to +1. Reliabilitycoefficients represent the amountof reliable variance associated with a test.In other words, a reliability coefficient isnot squared, as is the case with correlationcoefficients, to calculate the amount of reliablevariance (Anastasi & Urbina, 1998).For example, the reliable variance of a testwith a reliability coefficient of .90 is 90%,an unusually easy computation!The error variance associated with atest is also easy to calculate. It is doneby subtracting the reliability coefficientfrom 1 (perfect reliability). Taking theprevious example, the error variance fora test with a reliability coefficient of .90is 10% (1 - .90).The sources of measurement error,while potentially crucial for interpretation,are often not specified, leaving the psychologistto engage in speculation. Errormay result from changes in the patient’sattitude toward assessment or cooperation,malingering, rater biases, patients’ healthstatus, subjective scoring algorithms, oritem content that is incomprehensible tothe examinee, among other factors.For this reason, it is important to considerstatistics that document both the reliableand error variance of a scale or test.In addition, multiple reliability coefficientsand error estimates based on classical andmodern test theory methods are necessaryto guide clinical and research practice.Logically, then, it follows that no singleestimate of reliability or error discussed inthis section is adequate to support routineuse of a test of assessment procedure.Test–Retest MethodA popular method for computing the stabilityof personality test scores is the testretestmethod. In this method the sametest, for example the MMPI-A, is administeredto the same group of individualsunder the same or similar conditionsover a brief period of time (typically 2–4weeks). The correlation between the firstand second administrations of the test isthen computed, yielding a test-retest reliabilitycoefficient that is optimally veryclose to 1.0. Of course, the importance ofsuch reliability depends on the constructbeing assessed. If clinicians seek to assesschanges in specific, discrete behaviors asa result of an intervention, for example,then test-retest reliability becomes lessof a concern. On the other hand, if aclinician seeks to evaluate what are presumablyrelatively stable indicators ofbehavioral functioning or personality,the test-retest reliability of the measurebecomes paramount.

CHAPTER 2 MEASUREMENT ISSUES35Internal ConsistencyCoefficientsAnother type of reliability coefficient typicallyreported in test manuals is an internal consistencycoefficient. This estimate differsfrom test-retest or stability coefficients inthat it does not directly assess the stabilityof the measure of personality over time.Internal consistency coefficients assesswhat the name implies-the average correlationamong the items in a test or scale.In other words, this index of reliabilityassesses the homogeneity of the test itempool. Internal consistency coefficients areinexpensively produced, since they onlyrequire one administration of the test.Typical formula used for the computationof internal consistency coefficients includesplit-half coefficients, Kuder Richardson20, and Coefficient (or Cronbach’s) Alpha.On occasion, there are differencesbetween internal consistency and test-retestcoefficients that can affect test interpretation.A test may, for example, have a relativelypoor internal consistency coefficientand yet a strong test-retest coefficient (Kamphaus,2001). Because internal consistencycoefficients are imperfect estimates of stabilitycoefficients, both types of coefficientsshould be recorded in the manual for a test(AERA, APA, NCME, 1999). It is then upto the professional making use of the test todetermine if the reliability is suitable for thepurpose for which the tool is to be used.Variables that Affect ReliabilityClinicians who use personality or behaviortests should recognize factors that canaffect reliability. Some factors that the clinicianshould keep in mind when estimatingthe reliability of a test for a particularchild include the following:1. Reliability can differ for different scorelevels. A test that is very reliable foremotionally disturbed students is notnecessarily as reliable for nondisabledstudents without research evidence tosupport its use (AERA, APA, NCME,1999).2. Reliability can suffer when there isa long interval between assessments(Nitko, 1983).3. Reliability can be affected by rater orchild characteristics such as age, readinglevel, and fatigue. Reliability of personalitymeasurement, for example, maydrop if the child does not understandthe test items.4. Analogously, error may be introduced ifa poor translation of a test is used.Reliable Specific VarianceSubtest specificity is the amount of reliablespecific variance that can be attributed toa single subtest or scale. Kaufman (1979)popularized the use of subtest specificityin clinical assessment as a way of gaugingthe amount of confidence a clinicianshould have in conclusions that are basedon a single subtest. In effect, knowledge ofsubtest specificity makes clinicians morecautious about drawing conclusions basedon a single scale.A reliability coefficient represents theamount of reliable variance associated with ascale. An example would be an anxiety scaletaken from a larger battery of 13 tests, all ofwhich are part of a major personality test battery.The anxiety scale has a test-retest reliabilitycoefficient of .82. On the surface, thistest appears reliable. If this scale producesthe child’s highest score, the examiner maywish to say that the child has a problem withanxiety. The examiner can then make thisstatement with confidence because the test isrelatively reliable, right? Not necessarily. AsKaufman (1979) points out, the conclusionbeing drawn by the clinician is about someskill, trait, or ability (in this case, anxiety) that

36 CHAPTER 2 MEASUREMENT ISSUESis specific or measured only by this one scale.The reliability coefficient, on the other hand,reflects not just reliable specific variance butalso reliable shared variances. Subtest specificityis typically computed in the followingway (Kamphaus, 2001):1. Compute the multiple correlation (R)between the scale in question and allother scales in the battery, and square it(R 2 ). This computation yields the amountof reliable shared variance between thescale in question, in this case anxiety, andthe other scales in the battery.2. Subtract the squared multiple correlationcoefficient from the reliabilitycoefficient, or r tt. If R 2 = .30, .82 - .30 =.52. This formula yields the reliable specificvariance.3. Compare the amount of reliable specificvariance (.52) to the amount of errorvariance (1 – .82 = .18). If the reliablespecific variance exceeds the error varianceby .20 or more, then the scale isconsidered to have adequate specificityfor interpretive purposes. By convention,if the reliable specific varianceexceeds the error variance by .19 orless, then the test lacks specificity, andit should be cautiously interpreted. Ifthe reliable specific variance does notexceed the error variance, then interpretationof the scale is ill-advised.Standard Error of MeasurementThe standard error of measurement (SEM)gives an indication of the amount of errorassociated with test scores. In more technicalterms, the SEM is the standard deviationof the error distribution of scores. Thereliability coefficient of a test is one way ofexpressing the amount of error associatedwith a test score in order to allow the userto gauge the level of confidence that shouldbe placed in the obtained scores. An examinermay report a personality test score for achild as being 63 with a test-retest reliabilitycoefficient of .95. This practice, however,is unorthodox and clumsy. The typicalpractice is to report a test score along withthe test’s standard error of measurement,as is frequently done for opinion polls conductedby the popular media (e.g., the errorrate or margin of error of this poll is…).The standard error of measurement is simplyanother way of reflecting the amount oferror associated with a test score.In classical test theory, if a child wereadministered a personality test 100 timesunder identical conditions, he or shewould not obtain the same score on all 100administrations. Rather, the child wouldobtain a distribution of scores that approximatesa normal curve. This error distributionwould have a mean. The mean ofthis theoretical distribution of scores is thechild’s true score. A true score is a theoreticalconstruct that can only be estimated. This errordistribution, like other distributions, notonly has a mean, but it can also be dividedinto standard deviations. In an error distribution,however, instead of being calleda standard deviation, it is called the SEM.As one would predict, then, in this errordistribution of scores ±1 SEM dividesup the same portion of the normal curve(68%) as does a standard deviation, and ±2SEMs divide up the same proportion ofthe error distribution (95%) as ±2 standarddeviations do for a normal distribution ofobtained scores.Confidence BandsA confidence band is a probability statementabout the likelihood that a particularrange of scores includes a child’s truescore. As is done with opinion polls, cliniciansuse the SEM to show the amountof error, or unreliability, associated withobtained scores. Obtained scores arethen banded with error. “Banding” isfrequently accomplished by subtracting1 SEM from, and adding 1 SEM to, the

CHAPTER 2 MEASUREMENT ISSUES37obtained score. If, for example, the childobtained a T-score of 73 on the ReynoldsChild Depression Scale (RCDS; Reynolds,1989), one could apply the theory ofstandard error of measurement to bandthis score with error. For the total RCDSsample, the standard error of measurementrounds to 4 T-score points. Giventhat ±1 SEM includes approximately68% of the error distribution of scores,the clinician could then say that there is a68% likelihood that the child’s true scorelies somewhere in the range of 69–77. Anexaminer who wanted to use a more conservative±2 SEMs could say that thereis a 95% probability that the child’s truescore lies somewhere between 65 and 81.Confidence bands can be obtained for avariety of levels if one knows the SEMof the scale. Some manuals include confidencebands at the 68%, 85%, 90%,95%, and 99% levels.Construct ValidityValidity is defined as “the degree to whichaccumulated evidence and theory supportspecific interpretations of test scoresentailed by proposed uses of a test” (AERA,APA, NCME, 1999, p. 184). There are anumber of different ways of evaluating thevalidity of a test. Some of the more commontypes of validity evidence will be discussedin this section. Validity is the most importantpsychometric characteristic of a test.A test can be extremely well normed andextremely reliable and yet have no validityfor the assessment of personality. Onecould, for example, develop a very goodtest of fine motor skill, but if one triedto make interpretations about someone’spersonality from this test, such interpretationswould not be valid. That is, validity isessentially an issue pertaining to the uses ofa test and the interpretations that one seeksto make from test results.Virtually every aspect of a test eithercontributes to or detracts from its abilityto measure the construct of personality orbehavior, or, in other words, its constructvalidity. Construct validity is the degree towhich a test measures some hypotheticalconstruct. As such, the construct validityof a personality test cannot be establishedbased on a single research investigation orthe study of only one type of validity (e.g.,factor analysis). Construct validity is basedon the long-term accumulation of researchevidence about a particular instrument,using a variety of procedures for the assessmentof validity.Based on the information provided inthe previous paragraphs, it is clear that astatement that a test is valid or invalid isinappropriate. Instead, certain interpretationscan have more or less evidence tosupport their validity and the accumulationof evidence in support of these interpretationsis always ongoing (AERA, APA,NCME, 1999).Content ValidityOne of the reasons that many people woulddisagree with using a test of vocabularyknowledge as a measure of personality isthat it does not appear to possess valid content.Content validity refers to the appropriatesampling of a particular contentdomain. Content validity has been mostclosely associated with the development oftests of academic achievement (Anastasi &Urbina, 1998). Typically, procedures forthe establishment of content validity arejudgmental (Petersen, Kolen, & Hoover,1989).Personality test developers have oftenrelied on empirical test developmentmethods, in which items are assigned toscales based on statistical properties only(such as factor loadings, to be discussedlater), and many manuals do not providea clear indication of the source of items. In

38 CHAPTER 2 MEASUREMENT ISSUESsome cases, the item source is clear, such aswith the Children’s Depression Inventory(CDI; Kovacs, 1991), where items werebased on accepted diagnostic nosologiessuch as the DSM. Even in such cases, however,personality test developers usually donot go to the lengths of other test developersto document adequate sampling ofthe content (or psychopathology) domain.Few personality tests or behavioral ratingscales, for example, use panels of experts todevelop item content.Problems with regard to content validitymay be identified as cases of constructunderrepresentation or construct irrelevance.A depression scale may suffer constructunderrepresentation, for example,if it lacks both cognitive (e.g., excessiveself-deprecation) and vegetative symptomsof depression (e.g., problems sleeping).In this scenario, it may be said that thereare not enough items on the scale that areknown to be “indicators,” or symptoms ofdepression, resulting in questionable contentvalidity.The reader will note in later chaptersthat construct irrelevant items are a moreserious problem in behavior assessment.This problem is likely to occur when onlyempirical methods are used to constructscales and select items for scales (i.e., factoranalysis). Examples of construct irrelevanceare listed in Chap. 17 as they relateto the assessment of ADHD. In comparisonto some others, ADHD is a well-studiedcondition with a widely agreed-uponset of symptoms (e.g., motor hyperactivityand inattention). What if, however, anitem such as “My child is adopted” wasplaced on an inattention scale of a parentrating scale? As is noted in Chap. 17,such an item would likely be identified as asource of construct irrelevant variance forthis scale, a source that would lead to a lessvalid assessment of attention problems forthe child undergoing evaluation.In our view, construct irrelevance andconstruct underrepresentation are likelyto become problems at the item selectionstage of test development. We, therefore,caution test users to carefully review theprocess of item selection and scale constructionfor each test that they utilize. Bydoing so, we think that clinicians will bebetter able to judge the implications of testcontent for interpretation.Criterion-Related ValidityCriterion-related validity assesses thedegree to which tests relate to other tests ina theoretically appropriate manner. Thereare two kinds of criterion-related validity:concurrent and predictive.Concurrent ValidityThis type of validity stipulates that a testshould show substantial correlations withother measures to which it is theoreticallyrelated. One of the important criteria forthe evaluation of personality or behaviormeasures since their inception has beenthat they show a substantial correlationwith other indicators of psychopathology,such as well-validated tests or clinicians’ratings or diagnoses. The typical concurrentvalidity investigation involves administeringa new behavior rating scale and an existingwell-validated measure of psychopathologyto a group of children. If a correlation of .20is obtained, then the concurrent validity ofthe new test would be in question. A .75 correlation,on the other hand, would be supportiveof the validity of the new test.Predictive ValidityPredictive validity refers to the ability of a testto predict (as shown by its correlation) somelater criterion. This type of research investigationis conducted very similarly to a concurrentvalidity study, with one important exception.The critical difference is that in a predictivevalidity study the new personality test is first

CHAPTER 2 MEASUREMENT ISSUES39administered to a group of children, and thensometime in the future- perhaps two months,three months, or even six years-a criterionmeasure (such as clinicians’ ratings of adjustment)is administered to the same group ofchildren (see Verhulst et al., 1994).Correlations with Other TestsOne can use correlations with other teststo evaluate the validity of a behavior orpersonality test. In a sense, this method isa special type of concurrent validity study.The difference is that the correlation is notbetween a personality measure and somecriterion variable, such as clinicians’ ratingsof adjustment, but between a personalitytest and a measure of the same construct,another personality measure. For example,if a new test of anxiety is published, itshould show a substantial relationship withprevious measures, but not an extremelyhigh relationship (Anastasi & Urbina,1998). If a new personality test correlates.99 with a previous personality test, then itis not needed, as it is simply another formof an existing test and does not contributeto increasing our understanding of the constructof personality. If a new anxiety scalecorrelates only .15 with existing well-validatedanxiety scales, it is also likely not tobe a good measure of personality. New personalitytests should show a moderate tostrong relationship with existing tests, yetcontribute something new to our understandingof the construct of interest.Convergent/Discriminant ValidityConvergent validity is established when ascale correlates with constructs with which itis hypothesized to have a strong relationship.Discriminant validity is supported when apersonality measure has a poor correlationwith a construct with which it is hypothesizedto be unrelated. These types of validitymay be important to consider if there are noexisting, well-normed, or relatively recentmeasures of a construct. That is, one may notbe able to judge the criterion-related validityof a measure because no other suitable measuresof that particular construct exist. Ofcourse, convergent and discriminant validityare important indicators of validity for existing/establishedmeasures as well.If one were assessing the convergentand discriminant validity of a measure ofanxiety, one would expect high correlationswith other measures of anxiety andmoderate correlations with other measuresof depression, given the well documentedassociation between anxiety and depression(Klein et al., 2005). However, one wouldexpect only minimal correlations betweenanxiety and measures of learning problems,thus providing support for its divergentvalidity.Factor AnalysisFactor analysis is a popular technique forvalidating modern tests of personality thattraces its roots to the work of the eminentstatistician Karl Pearson (1901). Factoranalysis has become increasingly popularas a technique for test validation. A wealthof factor-analytic studies dates to the 1960swhen computers became available. Factoranalysis is difficult to explain in only a fewparagraphs. Those readers who are interestedin learning factor analysis need a separatecourse on this technique and a greatdeal of independent reading and experience.A thorough discussion of factor-analytictechniques can be found in Gorsuch(1988). An introductory-level discussioncan be found in Anastasi & Urbina (1998)and Kamphaus (2001).Factor analysis is a data reduction techniquethat attempts to explain variancein the most efficient way. Most scales oritems included in a test correlate with oneanother. It is theorized that this correlationis the result of one or more common factors.The purpose of factor analysis is toreduce the correlations between all scales

40 CHAPTER 2 MEASUREMENT ISSUES(or items) in a test to a smaller set of commonfactors. This smaller set of commonfactors will presumably be more interpretablethan all of the scales in a personality testbattery considered as individual entities.Factor analysis begins with the computationof an intercorrelation matrix showingthe correlations among all of the itemsor scales in a test battery. Most studiesof behavior or personality tests use itemintercorrelations as input. These intercorrelationsthen serve as the input to a factoranalyticprogram that is part of a popularstatistical analysis package.The output from a factor analysis thatis frequently reported first in test validationresearch is a factor matrix showingthe factor loading of each subtest on eachfactor. A factor loading is, in most cases ofexploratory factor analysis, the correlationbetween a scale and a larger factor. 1Factor loadings range from -1 to +1 just ascorrelation coefficients do. Selected factorloadings for the MMPI-A factor analysisof the standardization sample (Butcheret al., 1992) are shown in Table 2.2. A highpositive correlation between a scale and afactor means the same thing as a high positivecorrelation between two scales in thatthey tend to covary to a great extent. Onecan see from Table 2.2 that the Hysteriascale is highly correlated with Factor 1, forexample, and that Mania is not highly correlatedwith Factor 1, but it is highly correlatedwith Factor 2.Once the factor matrix, as shown inTable 2.2, is obtained, the researcher mustlabel the obtained factors. This labelingis not based on statistical procedures, buton the theoretical knowledge and perspectiveof the individual researcher. For the1When orthogonal (independent or uncorrelated)rotation techniques are used (and these techniquesare very frequently used in test validation research),the factor loading represents the correlation betweenthe subtest and a factor. This is not the case whenoblique or correlated methods of factor analysis areused (Anastasi & Urbina, 1998).MMPI-A, there is general agreement as tothe names of the factors. The first factor istypically referred to as general maladjustmentand the second as overcontrol. Thethird and fourth factors are named afterthe scales with the highest loadings oneach: social introversion and masculinityfemininity(Butcher et al., 1992).Test developers often eliminate scales oritems based on factor analyses. They alsocommonly design their composite scoresbased on factor-analytic results. This processwas not followed in the developmentof the MMPI-A, as this test was developedlong before the ready availability offactor-analytic procedures. Although theMMPI-A appears to be a four-factor test, itproduces 10 clinical scale T-scores, and nocomposite scores corresponding to the fourobtained factors are offered. More recentlydeveloped tests, such as the CBCL, madeheavy use of factor analytic methods in thedevelopment of scale and composite scores(see Chap. 7).Generally, consumers of factor-analyticresearch seek comparability between thefactors and composite scores offered forinterpretation. If there is, for example,a one-to-one relationship between thenumber of factors found and the numberof composite scores produced, then thevalidity of the composite scores is likelyenhanced.Confirmatory Factor AnalysisThe procedures discussed thus far aregenerally referred to as exploratory factoranalyticprocedures. A newer factor-analytictechnique is called confirmatory factoranalysis (Kamphaus, 2001). These twofactor-analytic procedures differ in somevery important ways. In exploratory factoranalysis, the number of factors to be yieldedis typically dictated by the characteristicsof the intercorrelation matrix. That is, thenumber of factors selected is based on theamount of variance that each factor explains

CHAPTER 2 MEASUREMENT ISSUES41Table 2.2 Selected MMPI-A Factor Loadings1General MaladjustmentFactors2Social Overcontrol3Introversion4Masculinity FemininityHs .77 .09 .31 .05D .69 −.23 .51 −.08Hy .88 −.15 −.22 −.15Pd .71 .28 .21 .19Mf .08 .01 .07 −.84Pa .70 .19 .23 .25Pt .52 .39 .67 .06Sc .61 .38 .53 .36Ma .31 .78 −.04 .33Si .19 .10 .91 .02Note: These are varimax rotated factor loadings.Source: Adapted from Butcher et al., 1992.in the correlation matrix. If a factor, forexample, explains 70% of the variance inthe correlation matrix, then it is typicallyincluded as a viable factor in further aspectsof the factor analysis. If, on the other hand,the factor only accounts for 2% of thevariance in a factor matrix, then it may notbe included as a viable factor.In confirmatory factor analysis, the numberof factors is not dictated by data, but ratherby the theory underlying the test under investigation.In confirmatory factor analysis, thenumber of factors is selected a priori, as wellas the scales that load on each factor (Keith,1990). The primary test in confirmatory factoranalyses is the correspondence (i.e., fit)between the factor structure dictated a prioriand the obtained data. If there is a great dealof correspondence between the hypothesizedstructure and the obtained factor structure,then the validity of the personality testis supported (hence the term confirmatory)and the theory is confirmed. If, for example,a researcher hypothesized the existence offour factors in a particular personality test,the confirmatory factor analysis will test howwell the data from a specific sample conformto this hyopothesized test structure.Thorough confirmatory factor-analyticstudies use a variety of statistics to assessthe fit of the hypothesized factor structureto the data. These statistics may include achi-square statistic, goodness-of-fit index,adjusted goodness-of-fit index, or rootmean square residual (RMR). Several statisticsare desirable for checking the fitof a confirmatory factor analysis becauseall of these statistics have strengths andweaknesses. The chi-square statistic, forexample, is highly influenced by samplesize (Glutting & Kaplan, 1990).Cluster AnalysisSimilarly to factor analysis, cluster analysisattempts to reduce the complexity of a dataset. In factor analysis, it is typical to try toreduce a large number of variables (e.g.,items) to a smaller set. In cluster analysis,researchers are most often interested ingrouping individuals (as opposed to variables)

42 CHAPTER 2 MEASUREMENT ISSUESinto groups of people who share commoncharacteristics. Ward’s (1963) hierarchicalagglomerative method is one example of apopular cluster-analytic technique.Several steps are common to cluster-analytictechniques, including the following:1. Collect a sample of individuals whohave been administered one test yieldingmultiple scores or a battery of tests.2. For each variable (e.g., depressionscores), compute the distance betweeneach pair of children.3. These distances between each individualon each variable are then used toproduce a proximity matrix. This matrixserves as the input for the cluster analysisin the same way that correlation orcovariance matrices are used as input infactor analysis.4. Apply a cluster-analytic method that sortsindividuals based on the distances betweenindividuals that were plotted in the proximitymatrix. In simple terms, clusteringmethods in this step match individualswith the smallest distance between individualson a particular variable.5. This sorting process continues untilgroups of individuals are formed thatare homogeneous (i.e., have profiles ofscores of similar level and shape).6. Just as in factor analysis, the researcherhas to decide next on the numberof clusters that is the most clinicallymeaningful. Statistical indexes areprovided as an aid to the researcher inthis step.Cluster-analytic techniques are useful inpsychopathology research for identifyingsubtypes of disorders or for designingdiagnostic systems (Borgen & Barnett,1987). Cluster-analytic techniques havefrequently been applied to identify subgroupsbased on their performance ona particular personality measure (e.g.,LaCombe et al., 1991).In a series of investigations, Kamphausand colleagues have used cluster analysisof large data sets to identify children withsubsyndromal behavior problems (Huberty,Kamphaus, & DiStefano, 1997; Kamphaus,Huberty, DiStefano, & Petoskey,1997; Kamphaus et al., 1999). These studiesof elementary school children suggestthat there are numerous children with profilessuggestive of functional impairmentin school or at home who, nevertheless,are either not diagnosed or do not meetaccepted diagnostic criteria. Thus, thesecluster analyses helped to classify childrenwithout mental health diagnoses but whomay require prevention or treatment.Sensitivity and SpecificityIdentification of a diagnosis is one of theprimary reasons for conducting an evaluation.A test that is to be used for such apurpose should possess evidence of sensitivity,or the ability to identify true positives(i.e., the percentage of children whoactually have the disorder). A prototypicalstudy might involve administering an electronicmeasure of inattention to a group ofchildren with ADHD and a group withoutany psychiatric diagnoses (“normals”). Inthis type of investigation, electronic measuresof inattention often demonstrategood sensitivity by correctly identifyingthe vast majority of cases of ADHD, afinding that then triggers investigation ofspecificity. Specificity refers to the relativepercentage of true negatives, or the correctidentification of individuals who donot have the disorder as not having thedisorder. This same measure of inattentionmay also identify only 50% of the nondiagnosedsample as “normal.” Therefore, itmay have demonstrated good sensitivitybut inadequate specificity (i.e., a high rateof false positives). Electronic measures ofinattention often produce results of thisnature. In an exhaustive review of the

CHAPTER 2 MEASUREMENT ISSUES43literature on such measures, Riccio andReynolds (2003) have found that, whilesensitivity is typically good, evidence ofspecificity is often poor.In a later chapter, we will observe thatthe standards for this type of sensitivityand specificity have been raised considerablyby the most recent Standardsfor Educational and Psychological Testing(AERA, APA, NCME, 1999). Now testsmust demonstrate the ability to differentiateamong diagnostic categories-not justbetween a diagnostic group and normality.Unfortunately, few test manuals provideevidence of this nature and many journalarticles test only the relatively easy distinctionbetween some condition and normality.The clinician, however, routinely hasthe more difficult task of differentiatingamong diagnostic categories. Again, electronicmeasures of inattention have notshown good evidence of diagnostic groupdifferentiation. In fact, Riccio and Reynolds(2003) concluded that when childrenwith a number of problems are included,the proportion of children correctly classifieddrops significantly.More recent work has focused on thedevelopment of an evidence-base that willimprove problem specificity, or the abilityto distinguish particular problems fromeach other (Mash & Hunsley, 2005). Cliniciansroutinely are faced with this taskwhich is also often considered “differentialdiagnosis.” The call for a larger moresound evidence base also raises awarenessof positive predictive power (i.e., the abilityof an item to correctly identify a child witha particular problem) and negative predictivepower (i.e., the ability of an item tocorrectly identify a child without a problem;Pelham, Fabiano, & Massetti, 2005).More research on these issues can onlyserve to assist in clinical decision making,but increased evidence will still not replaceclinical judgment in integrating informationfrom a variety of sources that havevarying degrees of validity.Threats to ValidityReadabilityAn obvious, but easily overlooked, threatto validity is the lack of ability of theparent, teacher, or child to understandthe personality test items. While concernis often expressed about the ability ofchildren to read test items, parents mayalso have difficulty due to limited educationalattainment or cultural or linguisticdifferences. Harrington and Follett (1984)found that most tests available at thetime they conducted their study failed toaddress the issue in their test manuals.They provide several suggestions to thepractitioner for screening informants inorder to guard against readability servingas a threat to validity.For parents, Harrington and Follettrecommend having examiners read the testinstructions for the informant and paraphrase.Children can be asked to read someitems from the beginning, middle, and endof the instrument aloud so the examinercan gauge the child’s reading skill.Related to this point is the problem oftranslational equivalence or the degree to whicha translation of a test is equivalent to its originallanguage form (AERA, APA, NCME,1999). Evidence of translational equivalenceshould be offered to reassure the test user thata threat to validity is not present.Response SetsA response set is a tendency to answer questionsin a biased fashion, thus maskingthe true feelings of the informant. Theseresponse sets are often mentioned, andaddressed in construction and interpretation,in some personality tests .The social desirability response set is thetendency of the informant to respond toitems in a socially acceptable way (Anastasi& Urbina, 1998). Some personalitytests include items and scales to assess thepotential effects of such a response set.

44 CHAPTER 2 MEASUREMENT ISSUES“I like everyone that I meet” might bean item on such scales. The acquiescenceresponse set is the tendency to answer “true”or “yes” to a majority of the items (Kaplan& Saccuzzo, 1993). A third response setis called deviation, and it comes into playwhen an informant tends to give unusualor uncommon responses to items (Anastasi& Urbina, 1998).Guarding Against Validity ThreatsPersonality and behavior tests ofteninclude other validity scales or indexesin order to allow the examiner to detectvalidity threats. Some tests include fakebad scales, which assess the tendency toexaggerate problems. Computer scoringof personality tests has allowed for theinclusion of consistency indexes. One suchindex allows the examiner to determine ifthe informant is answering questions in apredictable pattern. A consistency indexmight be formed by identifying pairsof test items that correlate highly. If aninformant responds inconsistently to suchhighly correlated items, then his or herveracity may be suspect.Examiners often also conduct informalvalidity checks. One quick check is todetermine whether or not the informantresponded to enough items to make the testresult valid. Another elementary validitycheck involves scanning the form for patternedresponding. A form that routinelyalternates between true and false responsesmay reflect a patterning of responses.One way to limit the influence ofresponse sets is to ensure that informantsare clear about the clinician’s expectations.Some clients may also need to take thepersonality test under more controlled circumstances.If an examiner has reason tobelieve, for example, that a child is oppositional,then the self-report personalitymeasure may best be completed in thepresence of the examiner.UtilityAs described earlier, clinical utility is the“next frontier” in evidence-based assessment.By the time an assessment instrumentis well-known and widely used inclinical settings, it usually has demonstratedadequate reliability and construct validity.However, as Mash and Hunsley (2005)describe, the question of utility or whetherthe instrument provides “psychologistswith the kinds of information that can beused in ways that will make a meaningfuldifference in relation to diagnostic accuracy,case formulation considerations, andtreatment outcomes” (p. 365) remains. Thisconcept can also be applied to the inclusionof a particular informant in the assessmentprocess.In short, a rating scale, for example,may be a valid indicator of depression,but its utility indicates how valuable thatparticular rating scale is for an assessmentof depression relative to other measuresand relative to the cost (monetary andtime) involved in administering it. Validity,including incremental validity (i.e., theimproved assessment decision as a resultof adding a measure), is a necessary conditionfor utility, and establishing such validityevidence for an assessment tool, andespecially an entire assessment battery, isarduous. Nevertheless, various forms ofvalidity evidence are likewise not sufficientfor demonstrating utility. Ultimately, inaddition to cost effectiveness, the clinicianmust take into account the assessment’srole in translating to effective interventionand subsequent positive change for a child(Mash & Hunsley, 2005).Calls to examine the clinical utility ofassessment are not entirely recent (e.g.,Hayes, Nelson, & Jarrett, 1987). However,give the current state of affairs, our discussionof utility is necessarily brief. As themove toward evidence-based assessmentbecomes strengthened by a larger collection

CHAPTER 2 MEASUREMENT ISSUES45of empirical research and improved communicationabout assessment strategies,future volumes will hopefully be well poisedto take on a detailed review of evidence onclinical utility.ConclusionsKnowledge of psychometric principlesis crucial for the proper interpretationof personality tests. As psychometricsbecome more complex, clinicians have tobecome increasingly sophisticated regardingpsychometric theory. Because personalityassessment technology has generallylagged behind other forms of child assessment,knowledge of psychometric theorymust be considered more often by the clinicianwhen interpreting scores.Some personality tests, for example, donot include basic psychometric propertiessuch as standard errors of measurement inthe manual. Such oversights discourage theuser from considering the error associatedwith scores, which is a basic considerationfor scale interpretation. Omissions like thisone are rare in academic and intelligenceassessment. The application of the SEMis merely one example of the psychometricpitfalls to be overcome by the user ofpersonality tests. This chapter ends, however,on an optimistic note. Newer testsand recent revisions are providing moreevidence of validity and test limitations intheir manuals.Chapter Summary1. A T-score is a standard score that has amean of 50 and standard deviation of 10.2. A percentile rank gives an individual’srelative position within the norm group.3. In order to select a representativesample of the national populationfor any country, test developers typicallyuse what are called stratificationvariables.4. Local norms are those based on somemore circumscribed subset of a largerpopulation.5. The reliability of a test refers to thedegree to which its scores are repeatedover several measurements.6. The standard error of measurement(SEM) is the standard deviation of theerror distribution of scores.7. A confidence band is a probability statementabout the likelihood that a particularrange of scores includes a child’strue score.8. The reliability of a test may differ forvarious score levels.9. Construct validity is the degree to whichtests measure what they purport tomeasure.10. Factor analysis is a data reduction techniquethat attempts to explain the variancein a personality or behavior testparsimoniously.11. In cluster analysis researchers are mostoften interested in grouping individuals(as opposed to variables) into clustersthat share common behavior ortraits.12. Personality tests and behavioral ratingscales often include other validityscales or indexes in order to allow theexaminer to detect validity threats.13. Sensitivity refers to the ability of a testto identify true positives and specificityto the ability of a test to identify truenegatives.14. Clinical utility concerns how well aparticular tool provides necessaryinformation and does so in a uniqueand cost-effective manner relativeto other tools (or informants).

C h a p t e r 3Classification and DevelopmentalPsychopathologyChapter QuestionsScience and Assessmentl Why is understanding the basic researchon children’s and adolescents’ emotionaland behavioral functioning important toclinical assessment?l How are classification and assessmentrelated?l What are some of the models used forthe classification of the emotional andbehavioral functioning of children andadolescents?l What are some of the advantages anddangers of classification?l What are some of the most importantimplications of the basic research in thefield of developmental psychopathologyfor the clinical assessment of childrenand adolescents?A basic assumption underlying the writingof this text is that, to be competent in theclinical assessment of children and adolescents,much more knowledge is requiredthan being able to simply administer tests,this being the easier part. Many other crucialareas of expertise are necessary for appropriatelyselecting the tests to be administeredand for interpreting them after administration.One such area of expertise was the focusof the previous chapter: an understandingof the science of measuring psychologicalconstructs. However, a more basic level ofknowledge is needed for using measurementtheory appropriately. That is, one musthave a thorough understanding of the natureof the phenomenon being measured beforedetermining the best method for measuring47P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_3, © Springer Science+Business Media, LLC 2010

48 CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGYit. In particular, the constructs of interest discussedin this chapter are the emotional andbehavioral functioning of youth. The first hintat the importance of this basic understandingof psychological constructs was made in theprevious chapter on psychometrics. It shouldhave become clear that good psychometricproperties are not absolutes. They dependon the nature and characteristics of the specificpsychological construct being assessed.For example, childhood depression is frequentlycharacterized by multiple episodes ofdepression interspersed with periods of normalmood (Kovacs, 2001). Therefore, highstability estimates over lengthy time periodsshould not be expected. In fact, if such stabilityoccurs, then one is measuring somethingthat is not an episodic depression.In addition to appropriately utilizing psychometrictheory, understanding the natureof the phenomenon to be assessed is crucialto almost every aspect of a clinical assessment,from designing the assessment batteryand selecting the tests, to interpreting theinformation, and communicating it to thechild and parent. For these reasons, scienceand clinical practice are inextricably linked.There are many areas of basic research thatenhance an assessor’s ability to conduct psychologicalevaluations, but we have selectedtwo that we feel are the most critical to theclinical assessment of children and adolescents.First, a thorough understanding ofthe theories that guide the different modelsof classification are necessary because theframework used to define and classify psychologicalfunctioning determines how onedesigns and interprets an assessment battery.Secondly, the clinical assessment of childrenmust be conducted in the broad context ofdevelopmental psychopathology.Developmental psychopathology refersto an integration of two scientific disciplines:child development and childpsychopathology. The integration restson the basic assumption that the mostappropriate way to view the emotionaland behavioral functioning of children,both normal and problematic, is within acomprehensive framework that includesthe influence of developmental processes(Rutter & Garmezy, 1983). A noteddevelopmental psychopathologist, JudyGarber, summarized her views on thisfield as being “concerned with both thenormal processes of change and adaptation,and the abnormal reactions to stressor adversity, as well as the relationshipbetween the two” (Garber, 1984, p. 30).Thus, developmental psychopathology isa framework for understanding children’semotions and behaviors that has manyimplications for the assessment process.It is beyond the scope of this book toprovide an intensive and exhaustive discussionof classification theories, or the manyimportant findings in the field of developmentalpsychopathology. Instead, thischapter illustrates the criticality of thesetwo knowledge areas to the assessment process,and also provides a basic frameworkfor applying this knowledge to the assessmentof children and adolescents. The discussionthat follows highlights some of theissues in both areas that we feel have themost relevance to the assessment process.ClassificationClassification refers to the process of placingpsychological phenomena into distinctcategories according to some specifiedset of rules. There are two levels of classification.One level of classification isthe method of determining when a psychologicalfunctioning is abnormal, deviant,and/or in need of treatment, whilethe second level of classification is themethod of distinguishing among the differentdimensions or types of psychologicalfunctioning. Thus, clinical assessmentis considered partly, as a process of classification.It involves (1) determining whethersome areas of psychological functioning

CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGY49in a child and adolescent are pathologicaland need treatment and (2) determiningthe types of pathology that may be present.Alternatively, according to Achenbach(1982) “assessment and classificationare two facets of what should be a singleprocess: assessment aims to identify thedistinguishing features of individual cases;taxonomy (classification) is the grouping ofcases according to their distinguishing features”(p. 1). Therefore, understanding theissues involved in classification is essentialto clinical assessment.The first issue involves acknowledgingthat any classification system of psychologicalfunctioning will be imperfect. Psychologicalphenomena do not fall into specificcategories of normal and abnormal, or intoclear, non-overlapping dysfunction types.This seems to be especially true with children;there is often no clear demarcationof when a dimension of behavior should beconsidered normal and when it should beconsidered pathological. Further, there isoften a high degree of overlap among thevarious forms of psychopathology in children.Finally, any classification system isonly as good as the research used to createit. As the research advances, so should theclassification system.Therefore, any system of classification isbound to be imperfect. Due to this imperfection,many experts have argued againstthe need for any formal classification system.Instead, they argue that psychological functioningshould be assessed and describedidiosyncratically for each individual person.That is, each person is a unique individualwhose psychological functioning shouldsimply be described in ways that maintainthis uniqueness without comparing it withthat of other individuals or fitting it intoartificial categories. This argument has anintuitive appeal given the complexity ofhuman nature. However, there are severalcompelling arguments for the need for goodclassification systems, in spite of the factthat even the best system will be imperfect.The Need for ClassificationSystemsCommunicationThe main purpose of classification systemsis to enhance communication among professionals(Blashfield, 1984; Quay, 1986).A classification system defines the rules bywhich psychological constructs are defined.In the absence of such a system, psychologicalconstructs are defined by idiosyncraticrules developed by each professional, andone cannot understand the terminology usedby a professional unless the rules employedin defining the terms is understood. Forexample, the term depression is a psychologicalconstruct that has several meaningsin the psychological literature on children.It can refer to Major Depressive Episodes,as defined by the Diagnostic and StatisticalManual of Mental Disorders, FourthEdition-Text Revision (DSM-IV-TR; APA,2000). In contrast, it can also refer to elevationson a rating scale of depression (Curry& Craighead, 1993), or to a set of responseson a projective technique (Exner & Weiner,1982). The concept of masked depression isused to describe the belief that many childhoodproblems (e.g., hyperactivity, enuresis,learning disabilities) are the result ofan underlying depressive state (Cytryn &McKnew, 1974). Not surprisingly, each ofthese definitions identifies a different groupof children.Thus, simply saying that a child exhibitsdepression does not communicate much toanother professional unless there is furtherexplanation on how this classification wasmade. On the contrary, if one states that thechild meets the criteria for Major Depressionaccording to the DSM-IV-TR system,then a classification is said to be madeusing a system with clearly defined rules.Also, another professional will then havea clear idea of how depression is defined,even if he or she does not agree with theDSM-IV-TR system. However, this communicationrequires precision in the use

50 CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGYof terms. In this example, using the termMajor Depression would be misleading, andactually impair the communication, if theterm was used without ensuring that theDSM-IV-TR criteria were met (e.g., basedon responses from a projective technique).Documentation of Need for ServicesClassification systems allow for the documentationof the need for services. Thisencompasses documenting the need forspecial educational services for a child,determining the need for mental healthservices within a given catchment area,attempting to determine appropriate staffingpatterns within an institution, or documentingthe need for services to third-partypayers (e.g., insurance companies). Theseuses of the classification systems have beenthe most controversial because the imperfectionsinherent in the existing systemscan lead to very deleterious outcomes formany people. For example, a child can bedenied services by a school or payment forthe services by an insurance company canbe denied if the child’s problems do not fitinto the classification system being used.Unfortunately, the task of documentingneed is inextricably linked to classificationbecause it requires some method of differentiatingbetween those in need of servicesand those that are not in need of services.The solution is not to eliminate the classificationsystems, but instead (1) to developbetter systems of classification that moredirectly predict the need for services and(2) to educate other professionals on thelimitations of the classification systems, sothat they can be used more appropriatelyfor documenting the need for services.Dangers of ClassificationDue to the reasons already stated, it is generallyaccepted that explicit classificationsystems are needed. However, users of theclassification systems must be aware of thedangers and limitations of such systems.Because clinical assessment is a process ofclassification, the clinical assessor must beespecially cognizant of these issues. Manyof these dangers can be limited if classificationsystems are used appropriately. Therefore,in the discussion that follows, we havetried to not only outline the dangers ofclassification but also to present practicesthat minimize or eliminate potentiallyharmful effects.Because psychological phenomena, andthe persons they represent, do not fallneatly into categories, one loses informationby attempting to fit people into arbitrarycategories. People within the samecategory (e.g., Major Depression) share certaincharacteristics (e.g., depressed mood,loss of interest in activities, disturbances insleep, impaired concentration), but therealso exist many differences among personswithin a category (e.g., the number ofdepressive episodes, whether the depressionstarted after the death of a relative). Theshared characteristics should provide someimportant information about the personsin the category (e.g., prognosis, response totreatment), else the classification becomesuseless. However, given the loss of informationinherent in any classification grouping,classification should not be considered theonly information necessary for an adequatecase conceptualization. Instead, any classification,whether it is a diagnosis or an elevationon a behavior rating scale, should beone part of a larger description of the case.This approach allows one to take advantageof the positive aspects of formal classification(e.g., enhanced communication); yet,it acknowledges the limits of such systemsand integrates classification into a broaderunderstanding of the case. In this book,case studies that illustrate this approach areprovided. In each case, diagnoses or othermethods of classification are integrated intoa more complete clinical description of thechild being assessed.

CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGY51A second danger of classification systemsis that they foster the illusion of aclear break between normal and psychopathologicalfunctioning. For example, ifthe classification is based on an elevationon a rating scale (over a T-score of 70), itgives the illusion of a dramatic differencebetween children with T-scores of 69 (notclassified) and children with T-scores of70 (classified). Stating that this is an illusiondoes not imply that all psychologicaltraits are on a continuum with normality,because some are clearly not. For example,in Jerome Kagan’s work with behaviorallyinhibited children, there seem to be anumber of qualitative differences betweenchildren with behaviorally inhibited temperamentand those without this temperament(Kagan & Snidman, 1991). However,if one was using a measure of behavioralinhibition with some cut-off for classifyinginhibition (e.g., a T-score of 65), theremay be some children close to this threshold(e.g., T-scores of 60–64) who were notclassified due to imperfections in the measurementtechnique (Ghiselli, Campbell,& Zedeck, 1981). Therefore, whether theillusion of a clear break is due to a normallydistributed trait or due to measurementerror, it is still an illusion.A third danger of classification is thedanger of stigmatization associated withthe psychological labels, often the endresult of classification. How strong theeffect of labeling is on psychological functioningis not clear from research. There isno evidence to suggest that the act of labelingcreates significant pathology througha self-fulfilling prophesy. However, it isalso clear that labels can affect how others,either lay persons (e.g., Snyder, Tanke,& Berscheid, 1977) or clinicians (Rockett,Murrie, & Boccaccini, 2007), interact withchildren who have been diagnosed withcertain mental health problems. Giventhis potential danger, classificatory terms(labels) should be used cautiously and onlywhen there is a clear purpose for doing so(e.g., when it influences treatment considerations).Also, when such terms are used,great efforts should be made to clearlydefine the meaning of the term to avoidmisinterpretations. And, finally, termsshould be worded to emphasize the classificationof a psychological construct andnot classification of the person. For example,it is better to use the phrase “a childwith conduct disorder” rather than stating“a conduct disordered child”.Evaluating ClassificationSystemsThus far, we have argued that classificationsystems are necessary despite the potentialdangers and misuses. However, this is onlythe case for good classification systems. Ifa classification system tells little about aperson, then nothing is gained in terms ofcommunication, and all the dangers (e.g.,loss of information, stigmatization) aremaintained. Therefore, it is essential tocritically evaluate any system of classificationand, even within a system, to evaluatethe individual categories.As discussed in the previous chapter, whileillustrating the association between classificationand assessment, one finds that evaluationof classification systems is similar to theevaluation of assessment and procedures, ingeneral. Specifically, the primary considerationsfor evaluating a system are its reliabilityand the validity of interpretations derivedfrom it (Quay, 1986). In terms of reliability, auser of a system must be able to make classificationsconsistently, such as over short timeperiods (test-retest reliability) or betweentwo independent users who make the classification(interrater reliability). In order forclassification systems to be reliable, the rulesof classification must be simple and explicit.However, reliability is important primarilybecause it limits the validity of a classificationsystem. Therefore, the validity of a systemis of paramount importance. Classification

52 CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGYmust allow for some valid interpretations tobe made. That is, classification must meansomething. It should tell something aboutthe causes of the child’s emotional or behavioralproblems or the likely course of theproblems. Most importantly, it should tellwhether the child needs treatment, and if so,what type of treatment.Models of ClassificationSo far, our discussion of classification systemshas been on issues that transcend anysingle type of classification system. Theuses of a system, the dangers inherent inclassification, and the methods of evaluatingsystems are all pertinent, irrespectiveof the model on which a system is based.However, there are several different theoreticalmodels on which a classificationsystem can be based. A model is a specificframework for viewing classification, suchas whether abnormal behavior is viewed asa statistical deviance (“Is this level of functioningrare in the general population?”) orin terms of its functional impairment (“Doesit affect a person’s adaptive functioning?”).The theoretical model of a system willdetermine the rules of classification. Aswill become evident, the different typesof assessment techniques discussed in thisbook were designed to provide informationabout the different models of classification.The following sections will reviewtwo general models of classification thathave strongly influenced the classificationof children and adolescents, and have had amajor influence on the types of assessmentprocedures that have been developed.Medical ModelsThe first major model of classification, themedical model, was largely derived fromclinical experience with disturbed childrenand adolescents (Achenbach, 1982; Quay,1986). In this type of classification, a diagnosticentity is assumed to exist, and thesystem defines the characteristics that areindicative of this diagnosis. The approachis called a medical model approach because itassumes there is a disease entity, or a coredeficit, which is the disorder. It then definesthe symptoms that are indicative of the presenceof the disorder.There are two primary characteristics ofthe medical model approach to classification.First, because of the emphasis on a core deficit,medical model systems differ dramaticallydepending on the theory or theories usedto define the deficits considered to underliethe psychological disorders. That is, medicalmodel systems are strongly influenced bythe theory of abnormal behavior espoused bythe system, such as psychodynamic theoriesor biological theories. Second, because of theemphasis on a pathological core (e.g., the diseaseentity), medical model systems typicallymake sharp distinctions between disorderedand non-disordered individuals. There istypically an underlying assumption that thereare qualitative differences between individualswith and without a disorder.Multivariate ApproachesThe second major approach to classificationthat has been extremely influential inthe clinical assessment of children has beenlabeled the multivariate statistical (Quay,1986) or the psychometric approach(Achenbach, 1982). In this approach, multivariatestatistical techniques are used toisolate interrelated patterns of behavior.Therefore, unlike the clinically derivedsyndromes that are defined by theory andclinical observations, behavioral syndromesare defined by the statistical relationshipbetween behaviors or their patterns ofcovariation. In this approach, behaviorsform a syndrome if they are highly correlatedwith each other, and there is no necessaryassumption of a pathological coreto underlie the symptoms, as is the case inmedical models of classification.

CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGY53Table 3.1 An Example of a Multivariate Classification SystemExternalizingInternalizingInattention HI/ODD CD Social Anxiety Depression SAD/FearsDisorganizedInterruptsothersFights Timid/shy Sad Upset overseparationDistractible Stubborn Spreads rumors Not self-confident Low energy Worriedabout parentForgetful Loud Bullies Nervous in groups Anhedonia Afraid to leavehouseSloppy/messy Noisy StealsDaydreams Talks a lot LiesNote: Dimensions are based on a series of exploratory and confirmatory factor analyses of caretaker ratings of1,358 children and adolescents of ages 4–17 (Lahey et al., 2004). Behaviors listed are just examples and not thecomplete list of items used in the factor analyses. HI hyperactivity-impulsivity, ODD oppositional defiant disorder,CD conduct disorder, SAD separation anxiety disorder.In addition to being based on statisticalcovariation, the psychometric approachis also different from medical models ofclassification because it emphasizes quantitativedistinctions rather than qualitativedistinctions. Once behavioral syndromesare isolated through statistical analyses, achild’s level of functioning along the variousdimensions of behavior is determined.Behavioral syndromes are conceptualizedalong a continuum, from normal to deviant.Interpretations are typically made bycomparing an individual case to a representativenormative sample, and choosingsome level of functioning as being so rarein the average population that it should beconsidered deviant. Classifications are thusbased on how a child falls into a certaindimension of functioning (e.g., anxiety/withdrawal) relative to some comparisongroup (e.g., compared to other children ofthe same age group).In Table 3.1, we provide an example ofa multivariate approach to the classificationof childhood emotional and behavioralfunctioning reported by Lahey et al.(2004). This system was based on a series ofexploratory and confirmatory factor analysesof caretaker ratings of 1,358 children andadolescents of ages 4–17. It shows sixdimensions of behavior that can be subsumedunder two broad dimensions ofExternalizing and Internalizing problems.Classification in the Future:An Integration of Medical andMultivariate ApproachesThese two basic approaches to classificationare important for clinical assessmentbecause the design of assessmentinstruments is often consistent withone of these basic approaches. Moreimportantly, the interpretation ofassessment instruments is basically aprocess of classification. Therefore,it is often guided by these models orsome variation of them. The clinicalassessor should be aware of the issuesinvolved in the classification generally,and the advantages and disadvantagesof these two models of classificationspecifically, to aid in the interpretationof assessment measures.Research has indicated that both themedical and the multivariate models have

54 CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGYflaws that make their exclusive use problematic(see Quay, 1986). For example, thedependence of medical model approacheson theory has led to many “disorders” beingcreated with little support from research.Also, the medical model approach, with itsemphasis on qualitative distinction, masksa continuum with normality that seemsmost appropriate for understanding manydimensions of functioning. In contrast,the multivariate approach with its dependenceon statistical analyses in the absenceof clear theory has resulted in syndromesthat are hard to generalize across samplesand with different sets of symptoms. Also,while some psychological phenomena inchildren and adolescents are best conceptualizedon a continuum with normality,there are others that may fit with morequalitative distinctions (e.g., Kagan &Snidman, 1991; Lahey et al., 1990) andare not captured well by the multivariateapproach.As a result, future classificationsshould look towards an integration of theapproaches. For example, clinical diagnosescan be improved by conducting multivariateanalyses to see if the covariationof symptoms for the diagnosis is supported(e.g., Frick, et al., 1993; Lahey et al.,2004). However, there are other ways inwhich the correspondence between statisticallyderived syndromes and the clinicallyderived diagnoses can be explicitly tested(e.g., Eiraldi, Power, Karustis, & Goldstein,2000), thereby improving the validityof both the approaches and leading to classificationsystems that accommodate thediverse nature of psychological constructs.In the following section, we provide anoverview of one of the most commonlyused classification systems, the Diagnosticand Statistical Manual of Mental Disorders,which is published by the American PsychiatricAssociation. Although initially thissystem of classification was based largelyon a medical model system of classification,more recent revisions have attemptedto capture the best characteristics of thetwo major classification approaches.Diagnostic and StatisticalManual of Mental Disorders,Fourth Edition (DSM-IV; APA,1994 & DSM-IVTR; APA, 2000)The DSM approach to defining psychiatricdisorders has undergone dramaticchanges in its many revisions since its firstpublication in 1952. The biggest changecame with the publication of its thirdrevision in 1980. In the first two editions,the definition of disorder was clearlybased on a medical model approach toclassification. The definition assumedan underlying pathological core, and theconceptualization of the core was largelybased on psychodynamic theory. In thethird edition, there was an explicit switchfrom a medical model view of disordersand a dependence on the psychodynamictheory. In the third and subsequent editions,a functional approach of viewingdisorders was used in which mentaldisorders were defined as “a clinicallysignificant behavioral or psychologicalsyndrome or pattern that occurs in anindividual and is typically associated withpresent distress (e.g., a painful symptom)or disability (i.e., impairment in one ormore important areas of functioning), orwith a significantly increased risk of suffering,death, pain, disability, or importantloss of freedom.” (p. xxxi, AmericanPsychiatric Association, 2000).Another major change in the thirdedition, also maintained in subsequentrevisions, is an increase in the level ofspecificity with which disorders aredefined. In the first two editions of themanual, disorders were often poorlydefined, leading to problems of obtaininghigh levels of reliability in the diagnosticclassifications (Spitzer & Cantwell,1980). In contrast, later revisions include

CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGY55more detailed diagnostic definitions withexplicit symptom lists, which have led toan increase in the reliability of the system(e.g., Spitzer, Davies, & Barkley, 1990).To illustrate this change, the DSM-II(APA, 1968) definition of HyperkineticReaction of Childhood is contrasted withthe analogous DSM-III-R definition ofAttention-Deficit Hyperactivity Disorderin Box 3.1.As a result of these changes, the mostrecent revision of the manual (DSM-IV;APA, 1994) has characteristics of boththe medical and multivariate models ofclassification. For example, its functionalapproach to defining disorders, which doesBox 3.1A Comparison of DSM-II and DSM-III-R Diagnostic CriteriaDSM-II: Hyperkinetic Reactionof Childhood (Adolescence)This disorder is characterized by overactivity,restlessness, distractibility,and short attention span, especially inyoung children; the behavior usuallydiminishes in adolescence. If this behavior iscaused by brain damage, it should be diagnosedunder the appropriate non-psychoticorganic brain syndrome.DSM-III-R: Attention-DeficitHyperactivity DisorderNote: Consider a criterion met only if thebehavior is considerably more frequent thanthat of most people of the same mental age.A. A disturbance of at least 6 months duringwhich at least eight of the following arepresent:1. Often fidgets with hands or feet orsquirms in seat2. Has difficulty remaining seated whenrequired to do so3. Is easily distracted by extraneous stimuli4. Has difficulty awaiting turn in games orgroup situations5. Often blurts out answers to questionsbefore they have been completed6. Has difficulty following throughinstructions from others (not due tooppositional behavior or failure of comprehension),e.g., fails to finish chores7. Has difficulty sustaining attention intasks or play activities8. Often shifts from one uncompletedactivity to another9. Has difficulty playing quietly10. Often talks excessively11. Often interrupts or intrudes on others,e.g., butts into other children’s games12. Often does not seem to listen to whatis being said to him or her.13. Often loses things necessary for tasksor activities at school or at home (e.g.,toys, pencils, books, assignments)14. Often engages in physically dangerousactivities without considering possibleconsequences (not for the purpose ofthrill seeking), e.g., runs into streetwithout lookingB. Onset before the age of 7.C. Does not meet criteria for a PervasiveDevelopmental Disorder.Sources: Diagnosis and Statistical Manual of Mental Disorders, Second Edition, American PsychiatricAssociation, 1967 and Diagnosis and Statistical Manual of Mental Disorders, Third Edition, Revised,American Psychiatric Association, 1987. Reproduced with permission of the publisher.

56 CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGYnot assume a pathological core, is consistentwith the multivariate approach toclassification. Also, many of the disordersare based, at least in part, on the patternsof symptom covariation, which is the hallmarkof multivariate models (e.g., Frick,et al., 1994). In contrast to the multivariateapproach, DSM-IV definitions classifydisorders into discrete categories, whichis more consistent with the medical modelapproach, although many of the cut-offswere empirically determined rather thanbeing based purely on theoretical considerations(Lahey, et al., 1994).One of the major changes in DSM-IVfrom its predecessors is the emphasis onusers having access to the basic researchunderlying the various diagnosticcategories. For example, in the manual,each disorder is initially introduced bysummarizing the current research on itsbasic characteristics, associated features likeage, gender, and cultural trends, prevalence,course, and familial pattern (APA, 1994).These introductory descriptions wereenhanced in a later version of the manual:the Diagnostic and Statistical Manualof Mental Disorders, Fourth Edition,Text Revision (DSMIV-TR; AmericanPsychiatric Association, 2000).In this revision, no changes were madeto the basic structure and diagnostic criteriafor the individual disorders, nor were anydisorders added or deleted. Instead, workgroups were formed to identify any errorsor omissions in the introductory materialfor each disorder, and to provide additionalmaterial to enhance the description of thedisorders and their basic characteristics,if these could be justified by a review ofthe relevant research. In addition, thepublication of the DSM-IV and DSM-IV-TRwere accompanied by the publication of anumber of edited volumes called “DSM-IVSource Books” (e.g., Widiger et al., 1994)that provide more extended reviews of theresearch that led to the diagnostic criteriaincluded in these manuals. As such, thesemost recent revisions of the DSM havetried to enhance the ability of users ofthis system to gain access to the researchfindings related to the disorders includedin the manual.DSM-IV maintained the multiaxial systemof classification that was initiated inDSM-III. As discussed previously in thischapter, a major disadvantage of any classificationsystem is its inability to captureall relevant dimensions of a person’s functioningwithin a single given category ordiagnosis. For a broader view of classification,DSM-IV specifies several dimensionsof functioning (axes) that are relevantto understanding a person’s functioning.Specifically, Axis I (Clinical Disorders)and II (Personality Disorders/MentalRetardation) are fairly typical of otherclassification systems and comprise themajor categories of mental disorders. Box3.2 provides a summary of the Axis I andAxis II diagnoses that are most relevantfor children and adolescents. However,DSM-IV includes three other dimensionson which a child can be classified. AxisIII allows the system user to indicate anyphysical disorder that is potentially relevantfor the understanding or managingof a case. Axis IV allows for a reporting ofthe psychosocial and environmental stressorsthat may affect the diagnosis, treatment,and prognosis of mental disordersin Axes I and II. Axis V provides a scale toindicate the highest level of adaptive functioning(psychological, social, and occupational/educational)that is currently beingexhibited or the highest level of adaptivefunctioning that has been exhibited withinthe past year. Clearly, this multiaxialapproach of DSM-IV is not sufficient totake the place of an adequate case formulation.However, it highlights the need toplace diagnoses in the context of manyother important aspects of a person’s psychologicalfunctioning.

CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGY57Box 3.2A Summary of DSM-IV Axes I and II Diagnoses Relevant to Children and AdolescentsIntellectualLearningMotor SkillsPervasiveDevelopmentalBehavioralMental RetardationMathematics DisorderDisorder of WrittenExpressionReading DisorderLanguage and SpeechExpressive Language DisorderMixed Receptive-ExpressiveLanguage DisorderPhonological DisorderStutteringSelective MutismDevelopmental CoordinationDisorderAutistic DisorderRhett’s DisorderChildhood DisintegrativeDisorderAsperger’s DisorderAttention-DeficitHyperactivity DisorderOppositional Defiant DisorderConduct DisorderEmotionalIdentityPhysical(Eating)Physical(Motor)Physical(Elimination)Bipolar Disorders (I &II)*Cyclothymia*Adjustment Disorder withDepressed Mood*Gender Identity DisorderofChildhoodReactive AttachmentDisorder of Infancy orEarly ChildhoodAnorexia Nervosa*Bulimia Nervosa*PicaRumination DisorderTourette’s DisorderChronic Motor or VocalTic DisorderTransient Tic DisorderStereotypic MovementDisorderEncopresisEnuresisEmotional(Anxiety)Emotional(Mood)Separation Anxiety DisorderGeneralized Anxiety Disorder *Panic Disorder *Agoraphobia *Social Phobia*Obsessive CompulsiveDisorder *Post-Traumatic StressDisorder *Adjustment Disorder withAnxious Mood*Major Depression *Dysthymia*Physical(Somatic)PsychosisSubstance–RelatedDisordersSomatization Disorder*Conversion Disorder*Pain Disorder *Hypochrondriasis*Body DysmorphicDisorder *Adjustment Disorder withPhysical Complaints*Schizophrenia*Alcohol (Amphetamine,Cannabis, etc.)Dependence**Denotes disorders that have the same criteria for children and adults.Note: The selection of disorders most relevant to children and adolescents and the grouping of disorders weremade by the authors and not by DSM-1V.

58 CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGYDevelopmentalPsychopathologyAs mentioned earlier, the overridingprinciple of developmental psychopathologyis that children’s emotional andbehavioral functioning must be understoodwithin a developmental context(Rutter & Garmezy, 1983). Therefore,it follows that the assessment of children’semotional and behavioral functioning mustalso be conducted within a developmentalframework. Such themes as understandingbehavior in a developmental contextand conducting assessment within a developmentalframework are broad principlesthat have several important specific implicationsfor the assessment process.Developmental NormsFirst, a developmental approach recognizesthat a child’s emotional and behavioralfunctioning must be understood withinthe context of developmental norms. Tobe specific, there are numerous behaviorsof children that are common at one age,but relatively uncommon at others. Forexample, bedwetting is quite common priorto age 5, and even at age 5, it is present in 15 -20% of children (Doleys, 1977; Walker,Milling, & Bonner, 1988). Similarly, childhoodfears tend to be quite common, andthe types of fears that are most commonshow a regular progression with childdevelopment (Campbell, 1986). For example,separation anxiety is not uncommonin infants toward the end of the first yearof life (Bowlby, 1969), whereas fears ofthe dark and imaginary creatures are quitecommon in preschool and school-age childrenbut decrease in prevalence with age(Bauer, 1976).These are just a few of the many development-relatedchanges in the prevalenceof specific child behaviors. Knowledge ofthese developmental changes in behavioris crucial to clinical assessment becausethe same behavior may be developmentallyappropriate at one age but indicativeof pathology at another. Therefore, assessmentof children and adolescents mustallow development-based interpretations.The critical nature of these interpretationsimplies that selection of assessmenttechniques must be based, at least in part,on the availability of age-specific norms.Further, given the rapid developmentalchanges experienced by children and adolescents,comparisons must be made withinfairly limited age groups. Whereas foradults using a comparison group that spansthe ages from 25 to 35 may be justifiable,a comparison group for children that spansthe ages 5–15 would be meaningless, giventhe many changes in development thatare subsumed within this period. Becausethe normative information provided by anassessment instrument and the appropriateuse of norm-referenced information bythe assessor are critical components to theclinical assessment of children, these issuesare discussed in great detail throughoutthis book.Developmental ProcessesUnfortunately, many assessors believe thatsimply comparing the assessment informationwith age norms is all that is neededto take a developmental approach to childand adolescent assessment. This is a muchtoo limited view of development, and howcan it be applied to understanding bothnormal and pathological outcomes in children.A developmental approach is a “process-oriented”approach. Put simply, thismeans that any developmental outcome,be it a normal personality dimension or aproblematic behavioral pattern, is the endresult of an interaction of numerous interrelatedmaturation processes (e.g., socioemotional,cognitive, linguistic, biological).

CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGY59This process-oriented approach has severalimportant implications for how assessorsconceptualize what they are trying to assessand how they go about doing it.First, given the interrelated nature ofthe maturation processes, this approachrecognizes that a focus on any single type ofdevelopmental process will provide only alimited, and sometimes misleading, understandingof a child’s psychosocial functioning.For example, understanding the familyenvironment of a child with behavioralproblems will only provide a limited perspectiveon how these behavioral problemsdeveloped without considering the child’stemperament, which may make the childmore difficult to raise. Such transactionalrelations among developmental processesnecessitate that an assessor design batteriesthat assess the many different types ofprocesses and take into consideration thepotential transactional relations amongthese processes when interpreting the testingresults.Second, the process-oriented developmentalapproach recognizes that the samedevelopmental process (e.g., a permissiverearing environment) may result in differentoutcomes (e.g., some children whoare creative, others who are dependent,and others who are antisocial), leadingto a concept called “multifinality”. Thecomplementary concept is “equifinality”and refers to the possibility that thesame outcome (e.g., antisocial behavior)can result from very different developmentalprocesses among individuals (e.g.,very strong emotional reactivity leadingto strong, angry and aggressive reactionsor weak emotional reactivity leading toproblems in the experience of empathyand guilt). The implications for the assessmentprocess are that the assessors need toexpect that the same personality pattern orpsychopathological condition may resultfrom very different processes across individuals,and assessment batteries need to bedesigned to assess these “causal pathways.”This, in fact, may be one of the most criticalconcepts in a developmental psychopathologicalapproach to assessment becauseit places the assessment process in theimportant role of uncovering the uniquecausal pathway that leads to a child’s currentfunctioning so that interventions canbe better tailored to the unique needs ofthe child (see, e.g., Frick, 2006).Third, how the various processes unfoldover development leads to specific “tasks”that may make certain behaviors more likelyto occur at certain points in development.These unique demands, or “developmentaltasks,” lead to many of the age-relatedchanges in children’s emotions and behaviorsdiscussed previously. Comparing achild’s behavioral or emotional functioningto the developmental norms can helpdetermine whether the child’s functioningis deviant compared to other children whoare experiencing similar developmentaldemands. However, simply comparing achild’s behavior to developmental normsand determining whether or not the child’sfunctioning is deviant compared to otherchildren of the same age does not allowthe assessor to determine whether (1) thechild’s problems should be considered anexaggeration of the normal maturationalprocesses operating at that developmentstage or (2) the child’s problems should beconsidered as a qualitative deviation fromnormal development (i.e., not consistentwith the specific demands of that developmentalstage), with the latter often beingindicative of a more severe pathologicalprocess.These two different interpretationsof deviations from developmental normscan be illustrated in the assessment ofconduct problems in children and adolescents.Research has documented anincrease in the acting-out behavior forboth boys and girls that coincides withthe onset of adolescence (e.g., Offord,et al., 1989). The first implication of thisfinding is that the assessment of adolescent

60 CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGYconduct problems should be based on acomparison with the adolescent norms,so that age-specific deviations can bedetermined. However, to interpret theseage-specific deviations, it is also helpfulto realize that an increase in acting-outbehavior in adolescence is consistentwith the identity formation processoutlined in Erik Erikson’s psychosocialtheory of personality development(Erikson, 1968). Specifically, Eriksoncharacterizes adolescence as a time whenyouths are struggling with the developmentof an individual identity, one thatis separate from their parents. Rebellionand questioning of authority aremanifestations of adolescents’ rejectionof parental and societal values, as theystruggle to develop their unique identity.Understanding this process allows for anadditional interpretation when one documentsdeviations from age norms. If ayouth exhibits developmentally deviantlevels of conduct problems for the firsttime in adolescence, the conduct problemsmay be best conceptualized as anexaggeration of a normal developmentalprocess (e.g., identity development).In contrast, a preadolescent child whoexhibits deviant levels of conduct problemsis showing a behavior that seemsmore qualitatively different from whatis expected from normal developmentalprocesses. It may, therefore, be an indicationof more severe pathology. This isconsistent with research, indicating thatconduct problems that have onset in adolescenceare more likely to be transient,whereas conduct problems with a prepubertalonset tend to be more severe andchronic (Moffitt, 2003).Stability and ContinuityIssues regarding the stability of childhoodbehavioral and emotional functioning areimportant from a developmental perspectiveto psychopathology. These complex issueshave important implications for the assessmentprocess. The basic issue of stability isnot unique to the assessment of children andadolescents, but has been a long-standingcontroversy in psychological assessmentthroughout the life span. For example, manyhave questioned the concept of personality,because it implies a consistency of behaviorover place and time that is often not apparentin human behavior (Mischel, 1968). Thisissue is more relevant to children than adultsbecause childhood behavior seems to beless stable over time and situation, makingthe concept of personality in children evenmore controversial. Our view of the debate,which is similar to the view of many othertheorists (e.g., Buss, 1995; Martin, 1998), isthat the concept of personality can be usefulif conceptualized appropriately, but dangerousif viewed wrongly.For example, many measures of children’sbehavior or other aspects of personalityshow much less stability in childrenthan do analogous constructs in adults.Specifically, Roberts and DelVeccio (2000)reported a meta-analysis of 152 longitudinalstudies assessing the average stabilityof personality traits in different age groupsover a 6–7-year period. They reported thatthe average stability coefficient for childrenand adolescents was 0.31, comparedto 0.54 for young adults, 0.64 for middleagedadults, and 0.75 for adults over theage of 50. These findings are not surprisinggiven the rapid developmental changesthat occur in childhood. However, thefindings have important implications forthe interpretation of personality measuresin children. Specifically, interpretations ofdispositional characteristics must be madecautiously in children so that there is noimplication of strong stability over time,unless data are available to support suchan interpretation. Given that the data arelacking in most cases, the term personalitymay be misleading for many domains ofchild behavior.

CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGY61Although we feel that this caution iswarranted, there are also several ways inwhich this general statement must be qualified.First, there is clearly some continuityin children’s behavior, and the degree ofstability (or instability) seems to be dependenton the domain of behavior beingassessed. For example, research generallyindicates that externalizing behaviors (e.g.,hyperactivity, aggression, antisocial behavior)tend to be more stable over time thaninternalizing behaviors (e.g., fears, depression)(e.g., Frick & Loney, 1999; Ollendick& King, 1994). Therefore, interpretationsof the stability of behavior must be dependenton the dimension of behavior that isof interest.In addition, aggregates of behaviors(behavioral domains) tend to be more stablethan discrete behaviors. For example,Silverman and Nelles (1989) reported onthe 1-year stability of mothers’ reports offears in their children between the ages of 8and 11. Over the 1-year study period, therewas only a 10% overlap between Time 1and Time 2 in the ten specific objects orsituations that mothers reported as elicitingthe most fear in their children. However,the correlation between the absolutenumber of fears was quite high. Althoughthe specific types of fears were not stableover the study period, the level of fears wasstable. Some have argued that aggregationallows one to pick up generalized responsetendencies that are not captured by discretebehaviors (Martin, 1988). However,this increase in stability through the aggregationof behaviors can also be conceptualizedfrom basic measurement theory. It hasconsistently been shown that increasingthe number of items on a measure of a traitalso increases its reliability (Ghiselli et al.,1981). Hence, the increased stability maybe a function of a more reliable method ofmeasurement.Finally, stability can be affected bywhether one is viewing the developmentaloutcome or the processes that may haveled to this outcome. For example, the typeof adjustment problem may be episodic, asin the case of depression, but the factorsthat led to this problem, such as problemsin emotional regulation, may continue evenafter the depression has remitted and couldplace the child at risk for other adjustmentproblems in the future. As such, depressionmay not appear stable, but the problems inemotional regulation are stable (Keenan,2000). Again, this illustrates the relevance ofthe process-oriented approach to assessmentthat is consistent with the developmentalpsychopathology framework. This frameworkillustrates the importance of not simplyassessing the developmental outcomes (e.g.,psychopathological conditions), but alsoassessing the various interacting processesthat lead to these outcomes.Situational StabilityExplicit in the developmental psychopathologicalperspective is a transactional view ofbehavior. That is, a child’s behavior influenceshis or her context and is also shapedby the context. As a result, one expects ahigh degree of situational variability inchildren’s behavior based on the differingdemands present across situations.Providing some of the best data onthis issue, Achenbach, McConaughy, andHowell (1987) conducted a meta-analysisof 119 studies that reported correlationsbetween the reports of different informantson children’s and adolescents’ (ages1 1/2 to 19 years) emotional and behavioralfunctioning. The correlations between differenttypes of informants (e.g., parentteacher)were fairly low, averaging about0.28. This low correlation is not a goodindicator of cross-situational specificity byitself, because reduced correlations couldalso be due to the individual bias of differentinformants rather than to actual differencesin a child’s behavior across settings.However, the mean correlations between

62 CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGYinformants who typically observe the childin similar situations (between two parentsor two teachers) were generally muchhigher, averaging about 0.60. The relativelylow correlations (0.28 vs. 0.60) acrossthe informant types compared to thosewithin the informant types serve as a morerelevant indicator of the high variability inchildren’s behavior across settings.These findings by Achenbach et al.(1987) were replicated in a later meta-analysisby Renk and Phares (2004). Also, themodest correlation among ratings fromdifferent informants may not be limited toonly children or adolescents. For example,the average correlation between the adult’sratings of their own personality and theratings that their spouse make of themonly show an average correlation of 0.39(McCrae, Stone, Fagan, & Costa, 1998).Thus, in general, the correlation betweendifferent raters of a person’s personalityand behavior across different informantsappears to be quite modest across the agerange. Achenbach et al. (1987) highlightseveral important implications of thesefindings to the assessment process. Theseare summarized in Box 3.3.In addition, there are several issues oncross-situational specificity that are analogousto those discussed on the stabilityof childhood behavior. First, like stability,the low correlations across situationsmay depend on the type of behavior beingassessed. Specifically, externalizing problemstend to show higher correlations acrossinformants than internalizing problems(Achenbach et al., 1987; Renk & Phares,2004). Second, the situational specificityof behavior may be a function of whetheror not aggregated domains of behaviorare studied, or whether discrete behaviorsare studied. For example, Biederman,Keenan, and Faroane (1990) compared thereports of both the parent and teacher onthe symptoms of attention deficit disorder(ADD). Individual symptoms showed anaverage correlation across home and schoolsettings of about 0.20. In contrast, on adiagnostic level there was a much higheragreement. There was a 90% probabilityof a teacher reporting enough symptoms toreach a diagnosis of ADD if the child wasdiagnosed by parents’ report. Similar tothe findings on stability, this suggests that,although individual behaviors (symptoms)may show a high level of specificity acrosssituations, the broader construct (diagnosis)of aggregated behaviors seems to showgreater consistency across situations.ComorbiditiesComorbidity is a medical term that refers tothe presence of two or more diseases thatoccur simultaneously in an individual. Thisterm has also been applied in the psychologicalliterature to denote the presence of twoor more disorders, or two or more problematicareas of adjustment co-occurring withinthe same individual. There can be severalreasons for comorbidity. For example,comorbidity can involve the co-occurrenceof two independent disorders, two disordershaving a common underlying etiology,or two disorders having a causal relationbetween themselves (Kendall & Clarkin,1992). Unfortunately, research in most areasof psychology has not allowed for a cleardelineation of the various causes of comorbidityamong psychological problems.Despite this inadequate understandingof the causes of comorbidity, this conceptis important for the clinical assessment ofchildren, for it is clear that comorbidityis the rule, rather than the exception, inchildren with psychological difficulties(Bird, Gould, & Staghezza, 1993). Specifically,children’s problems are rarelycircumscribed to a single problem area;instead, children tend to have problemsin multiple areas of adjustment. Forexample, in children with severe conductproblems, 50%–90% have a co-occurringADD, 62% have a co-occurring anxietydisorder, 25% have a learning disability,

CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGY63Box 3.3Research Note: Meta-Analysis of Cross-Informant Correlations for Child/AdolescentBehavioral and Emotional ProblemsAs noted in the text, Achenbach et al. (1987)conducted a meta-analysis of 119 studiesthat reported correlations between differentinformants on children’s and adolescents’emotional and behavioral functioning. Themeta-analysis included studies that correlatedratings of parents, teachers, mental healthworkers, observers, peers, and self-reportratings. The overall findings suggested thatcorrelations between different types of informants(e.g., parents and teachers) are fairlylow, averaging about 0.28, indicating a highdegree of variability in the report of a child’semotional and behavioral functioning acrossdifferent types of informants. The correlationswere higher when calculated betweensimilar informants (e.g., between two parentsor between two teachers), averaging about0.60. This suggests that the low correlationsbetween different informants may be atleast partially due to differences in children’sbehavior in different settings, rather thanto idiosyncratic methods of rating behavioracross informants.The authors of this meta-analysis discussseveral important implications of their findingsto the clinical assessment of children.1. “The high correlations between pairs ofinformants who see children in similar settingssuggest that data from a single parent,teacher, observer, etc. would providea reasonable sample of what would be providedby other informants of the same typewho see the child under generally similarconditions” (p. 227).2. “In contrast, the low correlations betweendifferent types of informants suggest thateach type of informant provides substantiallyunique information that is notprovided by other informants. The highdegree of situational specificity poses aspecific challenge to clinical assessmentsintended to categorize disorders accordingto fixed rules” (p. 227), such as theDiagnostic and Statistical Manual of MentalDisorders-Fourth Edition, Revised (APA,1994). Such systems specify symptomsthat must be judged present or absent,and the low correlations across settingssuggest that in most cases the presence orabsence will depend on the setting.3. As a result of the high degree of specificity,clinical assessments of childrenshould obtain information from differentinformants who see the child in differentsettings. “In such assessments, disagreementbetween informants’ reports areas instructive as agreement because theycan highlight variations in judgments ofa child’s functioning across situations” (p.228). In Chap. 15, we discuss the issuesinvolved in deciding how to interpretthese variations in reports of a child’s oradolescent’s emotional and behavioralfunctioning.Sources: Achenbach, T. M., McConaughy, S. H., & Howell, C. T. (1987). Child/adolescent behavioral andemotional problems: Implications of cross-informant correlations for situational specificity. Psychological Bulletin,101, 213–232.and 50% have substantial problems inpeer relationships (see Frick & O’Brien,1994; Strauss et al., 1988). Similar ratesof comorbidity are found in many othertypes of child psychopathology.Research attempting to understandcomorbidity in childhood psychopathologyhas had a major impact on our understandingof the causes of several childhooddisorders (e.g., Hinshaw, 1987). However,comorbidity also has a more immediateimpact on the clinical assessment of childrenand adolescents. In a special issueof the Journal of Consulting and Clinical

64 CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGYPsychology (Kendall & Clarkin, 1992), severalarticles highlighted the unique treatmentneeds of children with various types ofcomorbid psychopathological conditions.The high degree of comorbidity and itsimportance to the design of effectivetreatment programs makes the assessmentof comorbid conditions crucial inthe clinical assessment of children andadolescents.Because of the importance of comorbidity,most clinical assessments of childrenand adolescents should be comprehensive.Specifically, assessments must cover multipleareas of functioning so that, not onlyare the primary referral problems assessedadequately, but potential comorbid problemsin adjustment are also assessed. Ingeneral, the clinical assessment must bedesigned with a thorough understandingof the high degree of comorbidity presentin child psychopathology and the mostcommon patterns of comorbidity that arespecific to the referral question.Practical Implications for AssessmentAlthough we have tried to summarizesome of the major findings in the field ofdevelopmental psychopathology that haveparticular relevance to clinical assessment,sometimes it is difficult to translate researchinto guidelines for practice. The followingis a summary of some of the major implicationsof the findings discussed in this sectionapplied to the clinical assessment ofchildren and adolescents. These implicationsare expanded and applied to specificsituations throughout this book.1. A competent assessor needs to beknowledgeable in several areas ofbasic psychological research to competentlyassess children and adolescents.In addition to competence inmeasurement theory, knowledge ofdevelopmental processes and basiccharacteristics of childhood psychopathologyis also essential.2. Children’s behaviors and emotions mustbe understood within a developmentalcontext. Therefore, an important characteristicof assessment instrumentsfor children is their ability to providedevelopmentally sensitive normativecomparisons. On a more general level,appropriate interpretations of testscores, even if they are based on agespecificnorms, should be guided by aknowledge of developmental processes,and their effect on a child’s behavioraland emotional functioning.3. Children’s behavior is heavily dependenton the contexts in which the childis participating. Therefore, assessmentsof children must be based onmultiple sources of information thatassess a child’s functioning in multiplecontexts. In addition, an assessment ofthe relevant aspects of the many importantcontexts in which a child functions(e.g., at school, at home, with peers) iscrucial in understanding the variationsin a child’s behavior across settings.4. Most assessments of children mustbe comprehensive. This is necessarybecause of the need to adequately assessthe many important situational contextsthat influence a child’s adjustment.Children often exhibit problems inmultiple areas of functioning that spanemotional, behavioral, learning, andsocial domains. Effective treatmentsmust be based on the unique strengthsand weaknesses of the child across thesemultiple psychological arenas.ConclusionsThe main theme of this chapter and, in fact,of this entire text, is that appropriate assessmentpractices are based on the knowledgeof the basic characteristics of the phenomenabeing assessed. As a result, the competentassessor is knowledgeable not only intest administration but is also well versed

CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGY65in psychometric theory, child development,and childhood psychopathology.Clinical assessment can be conceptualizedas a process of classification. Therefore,understanding the issues involved inclassifying psychological functioning isimportant. Formal classification systemsare needed to promote communicationbetween professionals, to utilize researchfor understanding individual cases, and todocument the need for services. However,classification systems can also be quitedangerous if they are poor systems or ifthey are not used appropriately.Two models of classification have had agreat influence on our understanding andassessment of children’s emotional andbehavioral functioning: clinically-derivedmedical models and statistically-derivedmultivariate models. Understanding thebasic assumptions of these approachesto classification, and understanding theadvantages and disadvantages of each areimportant for interpreting the assessmentinformation.Being knowledgeable about basicresearch within the field of developmentalpsychology is also crucial in conductingand interpreting psychological assessmentsof children and adolescents. This researchillustrates the importance of conductingand interpreting assessments within adevelopmental context, the importance ofunderstanding the stability and situationalspecificity of children’s psychological functioning,and the importance of comorbidityin childhood psychopathology. Theseresearch findings have many practicalapplications to the assessment process;these applications are discussed throughoutthis text.Chapter Summary1. To be competent in the clinicalassessment of children and adolescents,one must be knowledgeable about thecurrent research base on emotionaland behavioral disorders.2. Classification refers to a set of rulesthat delineates some levels or types ofpsychological functioning as pathologicaland places these significant areas ofpathology into distinct categories oralong certain dimensions.3. Appropriately developed and competentlyused classification systems can aid incommunication among professionals, inapplying research to clinical practice, andin documenting the need for services.4. Poor classification systems or inappropriatelyused classification systemscan foster an illusion of few differencesamong individuals within a givencategory, can foster an illusion of a clearbreak between normality and pathology,and can lead to stigmatization.5. Medical model approaches toclassification assume an underlyingdisease entity, and tend to classifypeople into distinct categories.6. Multivariate approaches baseclassification on patterns of behavioralcovariation and tend to classifybehavior along continuous dimensions,from normality to pathology.7. The Diagnostic and Statistical Manualof Mental Disorders, Fourth Edition TextRevision (APA, 2000) is one of the mostcommonly used classification systemswith characteristics of both medicalmodel and multivariate approaches.8. Developmental psychopathology providesa framework for understanding theadjustment of children and adolescents.9. Based on this framework, assessmentsmust be conducted with a knowledge ofage- specific patterns of behavior, witha knowledge of normal developmentalprocesses, and with consideration ofissues regarding the stability of behaviorover time and across situations.10. Because research has shown that childrenwith problems in one area of

66 CHAPTER 3 CLASSIFICATION AND DEVELOPMENTAL PSYCHOPATHOLOGYadjustment typically have problemsin multiple areas, most assessmentsof children need to cover multipledomains of functioning.11. Because research has shown that children’sbehaviors are heavily dependenton the contexts in which they occur,clinical assessments must assess achild across multiple contexts andassess the characteristics of the mostimportant contexts in which a childfunctions.

C h a p t e r 4Standards and FairnessChapter Questionsl Are professional guidelines availablefor the assessment of individuals fromdiverse cultural, linguistic, ethnic, economicand other backgrounds?l Why is it problematic to use the phrase,“valid test?”The majority of problems that occurin applied clinical assessment are duenot to inherent flaws in the tests, butto the inappropriate use of tests, andmisinterpretation of their results, byclinicians (Anastasi, 1992). Test misuse isprimarily due to substandard practice byclinicians, just as most auto accidents arecaused by driver error and not by the carper se. Even a widely used and accepted testcan become a tool for disserving a client.And these cases of misuse are common,and include misuses ranging from incorrectscoring to interpretations of scores thathave not been shown to be valid by severalindependent research studies (Eydeet al., 1993). Consequently, psychologicalassessment practice has long been governedby peer-developed guidelines and standardsthat have proliferated and become moreexplicit and sophisticated as the fieldmatures (AERA, APA, NCME, 1999). Thischapter is devoted to providing an executivesummary of some of the major publicationsin this area, especially those developed byrelevant learned societies. It also providesguidance for practice based on some of themost widely cited ethical principles, teststandards, regulations, and recent treatisesthat give suggestions for assessing diverseclientele. (A self examination for enhancingretention of these issues is given in Box 4.1.)67P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_4, © Springer Science+Business Media, LLC 2010

68 CHAPTER 4 STANDARDS AND FAIRNESSBox 4.1Ethics and Standards Self-ExaminationChecklistPeriodic completion of this checklist mayserve as a quick reference for the clinician tocue adherence to optimal practice methods.Principle/Guideline Questions1. Do I have adequate training to use thetests/methods that I plan to use?2. How might the individual’s backgroundcultural,linguistic, social, economic, orotherwise-affect the planning of my evaluationor the interpretation of my results?3. Are the tests that I am using validated forthe specific purposes that I have in mind?4. Are there particularly unreliable scalesthat I should refrain from interpreting?5. Have I received informed consent andassent prior to initiating the evaluation?6. Will I provide feedback to the client or tothe others concerned, such as the child’sparents, teachers, or pediatrician?7. Did I adequately protect patient privacy?8. Do I have written permission to shareconfidential information with concernedparties?9. Whom do I need to assist with in thisexamination-a translator, patient, socialworker, community member, etc.?10. Have I consulted a professional colleagueregarding questionable issues asneeded?Use and MisuseThere are many elements of competent testusage. According to one empirical studyof test usage, there are exactly 86 of them(Eyde et al., 1993). These competenciesrange from “accepting responsibility forcompetent use of the test to “not makingphotocopies of copyrighted materials” to“restricting test administration to qualifiedpersonnel.” In their unique casebook, Eydeand her colleagues assembled 78 case studiesof test “misuse,” accompanied by instructionalquestions and documentation of eachviolation of the 86 elements. This workrepresents a unique effort to document therelationship between clinician behavior andassessment practices that makes it a recommendedreading for students of assessment.Furthermore, the vignettes provide ampleevidence that guidelines and standards arenecessary for promoting optimal assessmentpractices.APA Ethical Principlesof PsychologistsVirtually every professional organizationadopts some ethical responsibility for itsmembers. An initial step is the developmentand dissemination of ethical principlesfor the members of the organization.Many organizations also adjudicate ethicalcomplaints against members made by thepublic or others.Psychology has a long history of involvementin test development and assessmentpractice, resulting in the frequent use ofthe term psychometrics. Hence, the ethicalstandards published by the American PsychologicalAssociation (APA) are amongthe most well known sets of ethical principlespromulgating standards for assessmentpractice. This section presents some of therelevant APA principles and provides sampleapplications of their use. These standardsprovide helpful guidance regardingthe restriction of test use from a professionalcontext only, requirements for evidenceor scientifically-based test interpretation,restriction of test use to qualified persons,adherence to relevant testing guidelinesand standards of practice, proper and fullexplanation of test results, and maintainingthe security of test items and other contentthat may make the test useless if released tothe general public.

CHAPTER 4 STANDARDS AND FAIRNESS69It can be challenging to maintain thesecurity of test content given the access torecords by many sources, and numerousrequests from parents, lawyers, or patientsthemselves to see the actual record formsused for the evaluation. Some internet sitesprovide sample items that are analogous toitems found in popular tests, such as theMMPI-2. However, for the most part, testitem content can be shielded from wouldbetest takers. When faced with questionsabout test security, it has commonly beenconsidered a good practice to a) explainthe problems associated with release ifitems have the ability to practice psychologicalassessment with others, or b) agreeto release test record forms only to otherqualified professionals, to interpret themappropriately for the person making therequest. These two responses are an oversimplificationof the various request typesand potential responses, suggesting thatconsultation with colleagues will be consideredwise under these circumstances.Test Standardseducational testing, employment testing,and program evaluation and public policy.These standards are well-articulatedand thorough, which portends that theyare influential in court proceedingsand forensic work, formation of publicpolicy, and, hopefully, in the assessmenttraining of psychologists and other usersof tests (Kamphaus, 1998).One of the vital points presented by thetest standards is that it is incorrect to usethe phrase “the validity of the test,” becauseit cannot be concluded that a particular testis valid for all children under all assessmentsituations (AERA, APA, NCME, 1999). Thevalidity of a test should be gauged properlyin relation to every assessment situation inwhich it may be put to use. Clinicians whoassess children’s personalities must, therefore,learn how to use more than one personalitytest well in order to validly assesschildren, families, institutions, or systems.A small sampling of some of the importantaspects of psychological test use and validationare given in the following excerptsfrom the test standards.The Standards for Educational and PsychologicalTesting (AERA, APA, NCME,1999) discuss in great detail the manyissues introduced by the ethical principles.In fact, this volume should bepart of every psychologist ’ s library, and itwould serve as a useful adjunct to this orany other textbook dealing with clinicalassessment.This latest version of the test standardsis ambitious, and includes chapterson validity, reliability, test development,scales and norms, test administration,scoring and reporting, supporting documentationfor tests, fairness, rightsof test takers, testing individuals fromdiverse linguistic backgrounds, testingindividuals with disabilities, responsibilitiesof test users, psychological testing,Evidence-Based InterpretationGiven the scope and complexity of thestandards, only a few of them can be summarizedhere. Of course, there is alwayssome loss of content when one summarizesan original source. Therefore, again, thereader is advised to read the original standards.Also, the Test Standards should bea required study for all doctoral programsin psychology, as is currently the case forrelevant ethical standards.While it is understood that it is simplisticto say that a test is valid or not valid, itis, therefore, also incorrect to consider aparticular test interpretation to be validfor all children being evaluated. First, testdevelopers must present a rationale foreach intended interpretation for their test

70 CHAPTER 4 STANDARDS AND FAIRNESS(AERA, APA, NCME, 1999). This rationalemust include documentation of validityevidence and theory relevant to the interpretationin a “comprehensive summary”(Standard 1.1; AERA, APA, NCME, 1999).In addition, although test developers arerequired to produce such evidence, individualpsychologists have to evaluate the qualityof the evidence, as it relates to specificcircumstances.With regard to validity, the followingstandards are also offered (paraphrasedfrom AERA, APA, NCME, 1999).1. Uses and interpretations of test scores,intended populations of application,and the construct(s) assessed must beclearly stated by the test developer.(Standard 1.2)2. Users should be cautioned not to interprettests in a manner inconsistent withthe available evidence. (Standard 1.3)3. Psychologists must justify and collectevidence for an interpretation of a testthat is not justified. (Standard 1.4)4. Validity of study samples should bedescribed in detail. (Standard 1.5)5. Procedures for selecting the test contentshould be specified. (Standard1.6)6. If experts are used in the test design,their roles and credentials should bedescribed in detail. (Standard 1.7)7. A rationale should be offered for interpretationof items or item subsets whensuch interpretations are advised by thedeveloper. (Standard 1.10)8. The relationships among scores withina test should be supported with evidence.(Standard 1.11)9. A rationale and scientific evidenceshould be provided for all interpretationsof score differences and profiles.(Standard 1.12)10. The conditions under which validityevidence was gathered should bedescribed. (Standard 1.13)11. The relationship between a test andscores on other measures should be theoreticallyconsistent. (Standard 1.14)12. When statistical adjustments, such asthose for restriction of range, are made,both adjusted and unadjusted valuesshould be given. (Standard 1.18)13. If a test is used to recommend alternativetreatments, evidence of differentialtreatment outcomes should be providedwhen feasible. (Standard 1.19)14. When unintended consequencesof test use occur, testinvalidity should be ruled out as acause for such consequences. (Standard1.24)With regard to psychologists, the followingstandards are given.1. Psychologists should limit their assessmentpractice to the use of tests that arequalified by training. (Standard 12.1)2. Psychologists should refrain from makingbiased interpretations that serve asspecial interests. (Standard 12.2)3. Tests should be selected only if theyare suitable for the characteristics ofthe patient. (Standard 12.3)4. Evidence must be provided if it is suggestedthat an interpretation can bemade based on combinations of testscores. (Standard 12.4)5. Tests used in combination to make adiagnosis must show adequate validity(sensitivity and specificity), and thepsychologist must meet the user qualificationsrequired to interpret the testsinvolved. (Standard 12.5)6. Psychologists should choose tests fordifferential diagnosis only if evidenceshows that the test can differentiatebetween clinical samples of interest,not just between a clinical sampleand the general population. (Standard12.6)

CHAPTER 4 STANDARDS AND FAIRNESS717. When a test is used to help make adiagnosis, the diagnostic category mustbe carefully defined. (Standard 12.7)8. Psychologists must ensure that psychometristswho work under theirsupervision are adequately trained.(Standard 12.8)9. Psychologists should describe the testingprocedures in a language understandableby the patient. (Standard 12.10)10. Confidentiality of the results should bemaintained consistent with legal andethical requirements. (Standard 12.11)11. Psychologists should use a setting andequipment necessary to obtain validresults. If this setting is not available, thetest should also be administered underoptimal conditions when possible forcomparison purposes. (Standard 12.12)12. Psychologists should be familiar withthe reliability and validity of evidenceof each test they use, and they shouldprovide a logical and coherent analysisof the results that support their inferences.(Standard 12.13)13. Qualitative information, such as backgroundinformation and observations,should be considered when making testinterpretations. (Standard 12.14)14. The quality of actuarial or computerbasedinterpretations and the norms onwhich they are based should be evaluatedfor their quality. (Standard 12.15)15. Psychologists should not imply that arelationship exists between test resultsand prognoses or treatment outcomesunless evidence is available for patientswho are similar to the patient beingevaluated. (Standard 12.16)16. Interpretations that suggest how apatient will perform on other measuresor outcomes should be supported byevidence of criterion-related validity.(Standard 12.17)17. Psychologists should base their interpretationson several sources of data,test results, or other evidence, andthey should be cognizant of the theory,empirical evidence, and limitations ofeach test used. (Standard 12.18)18. Construct irrelevant factors (e.g., motivation,response sets, health factors,suboptimal testing conditions) shouldbe considered as an alternative explanationfor a set of test results. (Standard12.19)19. Normally, psychologists should discussthe results with the patient in a languagethat he or she can understand.(Standard 12.20)One reason for delineating these twosets of standards is to give the reader anappreciation of their scope, which is trulyimpressive. Awareness of the scope of thestandards should alert psychologists that acompliance with these standards requiresdedicated effort and self monitoring.A few themes in these standards deserveelaboration. First, interpretation must beevidence-based and, when it is not, the factshould be made known to all consumers ofthe results. We suggest that psychologistsask themselves a few questions about a testinterpretation in order to encourage compliancewith the standards. If we conclude,for example, that a child has depression, ourself-monitoring questions could include:1. On what assessment results do we basethis interpretation?2. Do the structured interviews andself-report measures have adequateevidence of reliability, sensitivity, andspecificity?3. Is this scientific evidence based onadolescent samples with similar demographics(e.g., sex, age, ethnicity, geographicregion, language, etc.)?4. Is this evidence based on studies thatuse clinical control samples in additionto comparisons with the generalpopulation?

72 CHAPTER 4 STANDARDS AND FAIRNESS5. Are the scales and interviews free ofconstruct-irrelevant variance (e.g.,“threatens others,” which is not a coresymptom of depression but is an indicatorof aggression)?6. Is there an evidence of constructunder-representation (e.g., scale doesnot include any “vegetative” symptomsof depression)?7. Are the procedures for developingdiagnostic interview and scale contentdescribed and are reasonable?8. Were the interview results, test findings,patient background/history, andother evidence integrated into a coherentrationale for the diagnosis?9. Could there be an alternate explanationfor the conclusions drawn by theconstruct-irrelevant variance (e.g.,child was coached on the content ofsome test items)?10. What is the definition of depression(e.g., DSM-IV diagnostic criteria)?11. Is a classification or diagnostic decisionmade only to receive insurance reimbursementor other remuneration,financial or otherwise?12. Have these results been shared with theclient in a manner that he or she can fullyunderstand?13. Has confidentiality of the clients’results been maintained?Of course, a different set of questions maybe posed for non-diagnostic interpretationssuch as prognosis, treatment, or programevaluation.The issue of confidentiality is worthadditional comment. Rights to privacycontinue to be threatened due to changesin health insurance practices and increasingaccess to electronically stored information,among other factors (Alderman &Kennedy, 1995).Standards for Privacy of IndividuallyIdentifiable Health Information have beenissued by the U.S. Department of Healthand Human Services (Federal Register:December 28, 2000, volume 65, number250). These standards are remarkable inat least three ways: (1) They specificallyinclude records of psychologists who provide“qualified psychologist services”; (2)Psychotherapy notes that are “separatedfrom the rest of the individual’s medicalrecord” are generally excluded fromrelease, including release to the patient; (3)A patient may be denied access to his orher medical record if“a licensed health care professional has determined,in the exercise of professionaljudgment, that the access requested is reasonablylikely to endanger the life or physicalsafety of the individual or another person.”(p. 82823)These medical record release standardsare, therefore, reflective of some of thevicissitudes of mental health care. Theexclusion of psychotherapy notes from themedical record shows a high regard for thesensitivity of such information. For thepurposes of this text, however, it shouldbe noted that psychotherapy notes are notde-fined to include the “results of clinicaltests,” “diagnosis,” and so on.With regard to the other aspects ofthe medical record that may include psychologicaltest and assessment results andinterpretations, a psychologist may alsodeny access if the “protected health information”makes reference to another individualwho may suffer harm if informationis released. Finally, release may be deniedto the patient ’ s personal representative, ifthe provider thinks that harm may occur.Of course, a patient may ask for a reviewof any denial request and certain exclusions(e.g., criminal activities) from thisdenial provision are stipulated. However,these regulations appear to give psychologistssome discretion regarding the releaseof patient records.

CHAPTER 4 STANDARDS AND FAIRNESS73BIAS and Cultural CompetenceTest BiasThe perception that bias is inherent inpsychological tests has spurred many challengesand accusations (Kamphaus, 2001).A review of some of the major technicalissues follows.Mean Score DifferencesMost definitions of test bias do not usuallyconsider the issue of mean score differencesas a meaningful test of bias (Reynolds& Kaiser, 1990). Instead, importantquestions related to the validity of a personalitytest’s inferences across groupsforms a test of bias. In this approach tobias, an evidence of the construct validity fora personality test score inference differingacross groups exist. Numerous studieshave addressed these technical issues. Forthe purpose of this chapter, the definitionof test bias offered by Reynolds and Kaiseris most appropriate.“Test bias refers in a global sense to systematicerror in the estimation of some true value fora group of individuals. The key word hereis systematic; all measures contain error, butthis error is assumed to be random unlessshown to be otherwise.” (p. 624)Given this definition, one would expectscore differences to indicate genuine differencesin behavior or personality, if thetest works the same way (i.e., measures thesame constructs in a valid way) for variousgroups. An interesting finding with regardto ratings of child behavior is that fewmean differences exist between culturaland linguistic groups even for tests developedprimarily for use in the United States.Crijnen, Achenbach, and Verhulst (1997),for example, found remarkably small differencesbetween groups for parent ratingsusing the Child Behavior Checklist. Theystudied the results for 13,697 children andadolescents from 12 cultures includingChina, Israel, Sweden, German, Jamaica,and the United States. Their results have astriking similarity across culture, includingsimilarities in cross-sectional changes associatedwith age. They also noted that sexdifferences were invariant across cultures:With no significant exceptions, boys obtainedhigher externalizing scores but lowerinternalizing scores than girls. This genderdifference in the kinds of problems that parentsreport might thus be a “cultural universal”…(p. 1276).Several aspects of this study were corroboratedfor four cultural groups, for bothparent and teacher ratings in a study byKamphaus et al. (2000). This investigationevaluated the differences between U.S.Anglo, U.S. African American, U.S. Hispanic,and Colombian (Medellin) samplesfor both parent and teacher ratings fromthe BASC. Kamphaus et al. also found thatdifferences were small between groups forboth parent and teacher ratings, and sexdifferences were consistent regardless ofcultural/linguistic groups.Taken together, these studies, and manyothers, revealed that differences amongracial, ethnic, cultural, and linguisticgroups for behavior problem measures arerelatively small, a trend that stands in contrastto other realms of testing (i.e., intelligenceand academic achievement) withlarge differences (Kamphaus, 2001). And,while sex differences are relatively smallfor cognitive measures (i.e., intelligence),they are greater for child behavior problemmeasures. While interesting, however,these studies of mean score differences donot serve as clear indices of bias.Content Validity BiasContent validity was one of the first areasof investigation of test bias. This searchfor bias is understandable given that a

74 CHAPTER 4 STANDARDS AND FAIRNESSfrequent bias concern is usually directedat item content that seems inappropriateor, perhaps, even offensive to a group ofindividuals. Again, a very helpful definitionof content validity bias can be taken fromReynolds and Kaiser (1990):“An item or subscale of a test is considered tobe biased in content when it is demonstratedto be relatively more difficult for membersof one group than for members of another ina situation where the general ability level ofthe groups being compared is held constantand no reasonable theoretical rationale existsto explain group differences on the item (orsubscale) in question.” (p. 625)Numerous procedures have been proposedfor assessing bias in individual items, butthe logic behind item bias detection techniquesis fairly simple (Kamphaus, 2001).The fundamental aspect of most statisticalmethods that assess bias across cultural orgender groups is to match the groups onan overall score level, which is the first stepin the procedure. If, for example, one waslooking for gender bias in a pool of personalitytest items, one would first match boysand girls on their overall test score, be itstandard or raw score. So, if one wanted toevaluate biased items in the MMPI-A, forexample, one would first statistically groupthe cases, with perhaps all the boys andgirls with f-scores above 90 on a particularscale as one group, those between 80 and89 as another group, those between 70 and79 as a third group, and so on (it should benoted, however, that this is not the exactprocedure used by most item bias techniquesbut an oversimplification of suchprocedures). Subsequently, some statisticaltest of significance is applied to checkif, within these various score groups, thereare still significant differences in responseto the items of one group or another.This discussion relates to another itembias detection technique: judgmental biasreviews. The procedure used by somepublishers is to have groups of individualsreview the items carefully. This procedureensures that members of a number of culturalgroups review the items to determinenot only the potential bias, but also theitems that may be inappropriate for variouscultural groups. There is, however, muchdisagreement between judgmental reviewsof items and statistical analyses of bias. Itappears that statistical analyses of bias aremore reliable (Reynolds and Kaiser, 1990).In an investigation of judgmental biasreviews for intelligence test items, Sandovaland Mille (1979) compared the ratings of45 WISC-R items by 38 African American,22 Mexican American, and 40 undergraduatestudents. This study found that minorityand non-minority judges did not differin their ability to identify the culturallybiased items. The conclusions of Sandovaland Mille were that: (1) Judges are not ableto detect items that are more difficult for aminority child than for a Caucasian child,and (2) the item selection for minority childrenby judges of ethnic background did notshow any difference.Item bias, however, may be subtle anddifficult to detect. Canino and Bravo (1999)cite an example of a problem with contentequivalence on the Diagnostic InterviewSchedule for Children (DISC; Shafferet al., 2000). They found that, initially, itwas difficult to translate items regardingseasonal depression into Spanish. Later itwas found that even successful translationof the symptoms was of no value becauseseasonal depression never occurred for thechildren in sunny Puerto Rico.Construct Validity BiasA workable definition of construct validitybias by Reynolds and Kaiser (1990) reads as:“Bias exists in regard to construct validitywhen a test is shown to measure differenthypothetical traits (psychological constructs)for one group or another, or to measure thesame trait but with differing degrees of accuracy”(p. 632).

CHAPTER 4 STANDARDS AND FAIRNESS75The most popular method used for thestudy of construct validity bias is factoranalysis. Numerous researchers have usedsimilar procedures. The central characteristicof these procedures is to conductfactor analyses separately for various culturaland gender groups, and determineif a similar factor structure is yielded foreach group. The most popular procedurefor assessing agreement between factorstructures across groups is a coefficient ofcongruence, which is interpreted similar toa correlation coefficient.Lachar and Gruber (1994) provide anexample of this method for the PersonalityInventory for Youth (see Chap. 6). Theyconducted factor analyses separately by genderand ethnicity, and then compared the factorsyielded separately for the groups. Theirfindings were similar to those for abilitytests (Kamphaus, 2001) in that correlationsbetween the obtained factors were uniformlyhigh, in the low 0.90 at their worst.Predictive Validity BiasThe final type of bias that has received agreat deal of attention is predictive validitybias. A working definition of predictivevalidity bias is“A test is considered biased with respect topredictive validity if the inference drawnfrom the test score is not made with thesmallest feasible random error or if there isconstant error in an inference for predictionas a function of membership in a particulargroup”. (p. 638) (Reynolds & Kaiser, 1990)The issue of predictive or criterion-relatedvalidity is that these coefficients should notdiffer significantly across cultural or gendergroups. One of the typical procedures in thisresearch literature is to compare the predictivevalidity coefficients across groups. Astudy might compare the ability of a depressionmeasure to predict future adjustmentfor various groups, for example.If different predictive validity coefficientswere obtained for two or more groups, theresults would be called slope bias. In order tounderstand the concept of slope bias, it ishelpful to recall how correlation coefficientsare learned in introductory statistics courses.Such procedures are typically taught by havingthe students collect data on two variablesand plot the scores of a group of individualson these two variables. This plot results in ascatter plot. Then students compute a correlationcoefficient and draw a line of bestfit through the scatter plot. This line of bestfit is a visual representation of the slope. Acorrelation coefficient (predictive validitycoefficient) of 0.90 would produce a slopethat is very different from that obtainedwith a correlation coefficient of 0.30. Consequently,this form of bias in prediction isoften referred to as slope bias.Summary Comments on BiasWhile psychometric evidence of test biascan be found, little compelling evidenceof bias is found for various groups residingin the United States (Figueroa, 1990). As aresult, the focus has now changed to implicatetest misuse as the major contributingfactor to improper assessment of individualand groups of children. This misuse, however,includes more than individuals. Government,school district, or other entitiesmay, for example, create unwise policiesthat inadvertently produce biased and untowardoutcomes for children, such as imposingstrict cut-off scores that affect assessmentand conceptualization of the case.FairnessThe term fairness refers to “… the principlethat every test taker should be assessed inan equitable way” (AERA, APA, NCME,1999, p. 175). Some issues related to achievingthis objective are discussed in this section.The renewed focus on test use comes

76 CHAPTER 4 STANDARDS AND FAIRNESSat a time when psychologists are seekingto improve test use for various linguisticand cultural groups. Cultural plurality hasposed a challenge to assessment and diagnosticpractice since the early days of themental tests. The testing movement wasforced early on to change tests and testingpractice in the United States because ofthe tremendous influx of new immigrants.Between 1901 and 1910, over nine millionimmigrants entered the United States-moreimmigrants than the combined populationsof New York, Maryland, and New Hampshirein 1900 (French & Hale, 1990). Onecomponent of the initial appearance of theWechsler scales as an alternative to the Stanford-Binetmonopoly of the time was the factthat Wechsler included a performance scale thatcould be used with some success with non-English speakers (Kamphaus, 2001).Little has changed since the days of massmigration to the United States. In manyways, psychologists have used the same strategiesfor dealing with clients from diversecultures. A popular approach involves adaptingexisting assessment instruments. TheThompson adaptation of the TAT for adultsof African American heritage during the1930s is one of the early examples of suchattempts. Psychologists with multi-culturalexpertise should be able to adequately assessthe needs of a child from a culture that maydiffer from their own, even if test instrumentsthat are not specifically designed forthe child ’ s culture are the only ones available.A study by Malgady and Costantino(1998) highlighted the need for developingnew cultural competencies. They evaluatedthe effects of language and ethnicityof the clinician on diagnostic decision making,using Spanish-dominant adult patientsof Puerto Rican and Dominican descent.These patients were then seen by boardcertifiedpsychiatrists and clinical psychologistsfor diagnostic interviews. Patients werematched on DSM-IV diagnosis and dividedinto four interview language and clinicianassessment groups: English only/non-Hispanic clinician; Spanish only/Hispanicclinician; English only/Hispanic clinician;Bilingual/Hispanic clinician. Several of theMalgady and Costantino’s results are noteworthy.First, they found no differences inthe diagnoses or symptoms between psychiatristsand psychologists. Symptom severitywas highest among patients interviewed inSpanish by a bilingual interviewer. Symptomseverity was rated lowest when a patient wasinterviewed in English by an Anglo clinician.While concern has been expressed that a clinicianwho does not share ethnicity and languagewith a patient will pathologize (Cohen& Kasen, 1999), these results suggest that itis also possible that such a mismatch betweenpatient and clinician could lead to failure inidentifying psychopathology, which couldexacerbate the symptoms of those deniedaccess to treatment. Either way, researchdoes suggest a need for broader training ofclinicians in multi-cultural competencies.Emic Versus Etic PerspectivesAn emic perspective refers to behavior thatis considered specific to a culture, whereasan etic perspective presupposes that much ofthe behavior and laws of psychology areapplicable cross-culturally. Anastasi (1992)proposes that both perspectives are validby theorizing that learned behavior maybe culture-specific (emit) but that the “lawsof learning” apply cross-culturally (etic).She hypothesizes further that hierarchicalmodels of personality may be most usefulin studying their behavior, as is the casefor studies on intelligence when supportingevidence can be found for a “g” factorand for specific traits (e.g., spatial ability)at lower levels of the hierarchy. There is,for example, evidence that several temperamenttraits can be identified cross-culturally(Martin, 1988).Inappropriate, ill-informed, or insensitiveinterpretations may also be made of

CHAPTER 4 STANDARDS AND FAIRNESS77the “clinical” data. An examiner may concludethat a 13-year-old girl of Asian heritageis socially introvert, shy, and perhapsin need of assertive training because of herbehavior during an interview with a maleclinician. She may have been demure andmade no eye contact. The examiner maydraw such a conclusion despite the factthat she appeared friendly and outgoingwhen she was observed on the school playgroundand seemed to interact openly withher family members. This client may not,in fact, be pathologically shy; rather, shemay be adhering to a prohibition againstmaking eye contact with a male becauseof cultural values that suggest that this isa sexually seductive behavior (or an indicationof a lack of respect) that is deemedinappropriate for her (Hasegawa, 1989). Inthis case, the clinician was simply ignoringrelevant data, and the clinician ’ s lack offamiliarity with the child’s culture resultedin an erroneous interpretation.The clinician, however, must alsoremember the importance of individualizinginterpretation. Within a culturalgroup, variability can be substantial (Zuckerman,1990). It may be assumed by somethat Vietnamese and Chinese children havesimilar values due to early Chinese dominationand the inculcation of Vietnameseculture with Confucian ethics. There havealso been other influences on this culturethat may affect a child ’ s behavior, includingEuropean Roman Catholicism, brought bythe French conquest of 1958, the influenceof U.S. culture from the Vietnam War,and Buddhist influences from neighboringCambodia (Huang, 1989). Classifyingchildren by race, culture, or languagebackground is an appealing approach forresearchers and clinicians alike that isfraught with errors, primarily due to thetendency to overgeneralize a particulargroup of people (Zuckerman, 1990).Inclan and Herron (1985) cite the “cultureof poverty” as another subculture thatmay affect a variety of groups. This “culture”is formed by a clash between those who haveachieved material wealth and prosperity andthose who struggle to achieve economic paritywith little hope of doing so. Childrenreared in a culture of poverty possess identifiablecharacteristics: an orientation to presenttime, inability to delay gratification, impulsivity,sense of predetermined fate, resentmentof authority, alienation and distrust of others,and lack of emphasis on rigor, discipline,and perseverance (Inclan & Herron, 1989).They note that some impoverished parentsof adolescents may be assessed by a therapistas being too rigid and controlling theiryoungsters at a time when parents shouldbe giving their children more freedom. It ispossible, however, that poor parents may beall too familiar with the culture of povertyand may be seeking control, not for its ownsake, but rather to ensure that their child oradolescent does not fall prey to the negativeconsequences of the behavior associated withthat culture (Inclan & Herron, 1989).These examples demonstrate the needfor clinicians to develop an enlarged knowledgebase in order to deal effectively withtheir referral population. Just as cliniciansneed to have knowledge of behavioral principles,psychometrics, child development,child psychopathology, and physiologicalpsychology to conduct an evaluation competently,it is increasingly clear that theymust know the history, culture, and languageof their community extremely wellin order to not use assessment proceduresinappropriately, and to avoid making naïveand inappropriate interpretations.Guidelines for AssessingChildren from DiverseBackgroundsResources for assessing children fromdiverse groups are now more readily available(e.g., Dana, 1999; Geisinger, 1992).Two developments that can assist practitionersare (1) the availability of guidelines

78 CHAPTER 4 STANDARDS AND FAIRNESSfrom blue ribbon panels and committeesand (2) the increasing availability of formalmeasures of acculturation.Numerous sets of guidelines providespecific advice for the psychologist who isunsure of what procedures to use in questionablesituations. The Guidelines for Providers ofPsychological Services to Ethnic, Linguistic, andCulturally Diverse Populations give specificand helpful advice to the clinician seeking tocarry out a competent evaluation of a childfor whom cultural/social/linguistic issuesloom large (see www.apa.org).A good example of the guidance to begained from such publications deals withthe frequently occurring situation of a languagedifference between the examiner andthe child or other family members. Guideline6a suggests that a cascade of threeoptions applies to the examiner faced withsuch a case: (1) Refer the child to a clinicianwho can communicate in the client’spreferred language; (2) if this is not possible,use a translator who also possessesprofessional training; and, lastly, (3) oneis advised to use a paraprofessional fromthe community to translate. Moreover, thenext guideline, 6b, highlights the potentialthreat to validity of using a translator whohas a dual relationship with the client (e.g.,a grandparent).Assessing AcculturationThe previously discussed guidelines forconsidering cultural and linguistic issueshint at the need to more carefully assessan individual’s level of adoption of theso-called “dominant culture,” which, ofcourse, could change from one neighborhoodto the other. The guidelines indicatethe need to collect information, such as thenumber of generations of residence withinthe dominant culture, number of yearsof residence, dominant language fluency,community resources, and so on. This datacollection is an informal means of assessinglevel of acculturation. There are, however,more formal (some are quantifiable) methodsfor assessing acculturation. In fact, ithas been argued that the ready availabilityof such measures warrants their routine usein assessment practice (Geisinger, 1992).Marin (1992) defines the constructsrelevant to assessing ethnic identity andacculturation. He cites three componentsof ethnic identity: (1) “birth and gestationalhistory, (2) culture-specific behaviors andpractices (e.g., language), and (3) culturespecificattitudes that include adherenceto a culture’s values and norms as well asin-group and out-group attitudes” (p. 236).The process of acculturation is defined as“… changes in individuals that are producedby contact with one or more culturalgroups” (p. 237). Several instruments arenow available for assessing the ethnic identificationof individuals and the degree towhich acculturation has taken place.Dana (1993) provides a detailed compendiumof measures of acculturation andidentification with a particular culture.Some of these scales are listed below.African American MeasuresDevelopmental Inventory of Black Consciousness(DIB-C; Milliones, 1980)Racial Identity Attitude Scale (RIAS;Helms, 1986)African Self-Consciousness Scale (ASC;Baldwin & Bell, 1985)Asian American MeasuresEthnic Identity Questionnaire (EIQ;Masuda, Matsumoto, & Meredith, 1970)Hispanic American MeasuresAcculturation Rating Scale for MexicanAmericans (ARSMA; Cuellar, Harris, &Jasso, 1980)Children’s Acculturation Scale (Franco,1983)

CHAPTER 4 STANDARDS AND FAIRNESS79Cuban Behavioral Identity Questionnaire(CBIQ; Garcia & Lega, 1979)Hispanic Acculturation Scale (HAC;Marin et al., 1987)Children ’ s Hispanic Background Scale(CHBS; Martinez, Norman, & Delaney,1984)Cultural Life Style Inventory (Mendoza,1989)Multi-dimensional Scale of CulturalDifferences (MSCD; Olmedo, Martinez,& Martinez, 1978)Multi-cultural Experience Inventory(MEI; Ramirez, 1984)Behavioral Acculturation Scale (BAS;Szapocznik, Scopetta, & Aranalde, 1978)Bicultural Involvement Questionnaire(BIQ; Szapocznik & Kurtines, 1980)ConclusionsThe child psychologist of today has tobecome steeped in various ethical, legal,professional, language and cultural issues, andstandards of practice that face the profession.It is necessary for the practitioner to seekthis knowledge through experiences duringgraduate school and beyond. Continuousprofessional development is especiallyimportant in order to achieve fairness in theassessment process. Knowledge of cultural,linguistic, technology change, and othereffects on assessment remains in its infancy,thus portending considerable change in thefuture.and appropriate use of assessments andinterventions, limiting use of psychologicaltests to qualified professionals, andmaintenance of test security.2. The Standards for Educational and PsychologicalTesting (AERA, APA, NCME, 1999)provide a rich resource for maintainingstandards of practice.3. While psychometric evidence of content-,construct-, or criterion-relatedvalidity test bias can be found, littlecompelling evidence of bias is found forvarious groups residing in the UnitedStates. As a result, the focus has nowchanged to implicate test misuse as themajor contributing factor in improperassessment of children.4. The renewed focus on test use comes ata time when psychologists are seekingto improve test use for various culturaland linguistic groups.5. The emic perspective refers to behaviorthat is thought to be specific to a culture,whereas the etic perceptive presupposesthe behavior theory and lawsof psychology that are applicable crossculturally.6. The Guidelines for Providers of PsychologicalServices to Ethnic, Linguistic, andCulturally Diverse Populations give specificand helpful advice to the clinicianseeking to carry out a competentevaluation of a child for whom cultural/social linguistic issues provide threatsto test validity.7. There are now formal and quantifiablemethods for assessing acculturation.Chapter SummarySome of the APA ethics principles of relevanceto assessment include1. Evaluation, diagnosis, or interventionin a professional context, competence

C h a p t e r 5Planning the Evaluationand Rapport BuildingChapter Questionsl Why is it important to carefully plan anevaluation?l What information is necessary for planninga focused clinical assessment?l What are some of the important considerationsin determining whether or not achild should be tested and who should dothe testing?l What is a scientific approach to testing?l What is rapport and why is it more difficultto develop in the clinical assessmentof children and adolescents than inmany other clinical endeavors?l How can informed consent be considereda rapport-building strategy?l What are some of the important strategiesthat can aid in developing rapportwith children and adolescents?Non-specifics in ClinicalAssessmentA recurrent theme in this text is thatan assessor needs to have knowledge ofseveral areas of basic research to appropriatelyselect and interpret psychologicaltests for children and adolescents. Inthis chapter, we consider another areaof competence crucial to clinical assessmentthat goes beyond knowing howto administer specific tests. It is ratherdifficult to discuss this competence inobjective terms because it relates to difficulttopics for research and, as a result,there is only limited objective data toguide this practice. Instead, much inthis chapter is guided by clinical experience,not just our own experience,but the experience of other practicing81P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_5, © Springer Science+Business Media, LLC 2010

82 chapter 5 Planning the Evaluation and Rapport Buildingpsychologists who have written in thisarea.This chapter deals with setting anappropriate context in which testingtakes place. This is not simply the physicalcontext of testing, but the activitiesof the assessor that allow the clinicalassessment to achieve its goals. Manyof the issues discussed involve clinicalskills that are difficult to teach, but oftenrequire refinement based on practicalexperience in testing children and adolescents.However, an analogy can bemade with the literature on psychotherapy.Many useful guides for practicingclinicians have been published that dealwith the non-specifics of psychotherapy.The term non-specifics has been usedto refer to several contextual factors,within which the psychotherapy techniquestake place, such as the relationshipbetween therapist and client or theprocess by which a therapist engages aclient in a therapeutic setting (Karver,Handlesman, Fields, & Bickman, 2006).In this chapter, we attempt to deal withthe non-specifics in the clinical assessmentof children and adolescents.One critical component of setting anappropriate context for an evaluation iscareful planning. In the following section,we discuss a basic framework fordesigning clinical assessments for childrenand adolescents. Within this basicframework, however, evaluations mustbe tailored to the needs of the individualcase. The critical developmental issues,the most relevant areas of adjustment tobe assessed, and the most important elementsof a child’s or adolescent’s environmentwill all vary from case to case.As a result, it is inappropriate to developspecific guidelines for designing evaluations.Instead, in this section we attemptto provide a framework for designingassessments that can be tailored to mostassessment situations.Clarifying the ReferralQuestionA crucial part of planning any evaluationis having enough information,prior to beginning the testing, to makeat least some initial decisions on thestructure and content of the assessmentprocess. This is not to say that the assessmentprocess should be so structuredfrom the outset that changes after testingis underway are not possible. However,obtaining crucial information before thefirst testing session enhances the likelihoodthat one will provide a focused and appropriateassessment. Almost every testingagency, whether clinic, school, hospital, orprivate practitioner, has some establishedintake process that provides preliminaryinformation on the child or adolescentbeing tested. There is no single best way tostructure the intake procedure. However,there are some pieces of information thatshould be obtained routinely in any intakeprocess, in addition to any basic information(e.g., name, address, phone number,insurance coverage) that is required by theagency.Purpose of TestingThe most important piece of intake informationfor planning an evaluation is theintended purpose of the evaluation. Amajor flaw in many clinical assessments isa lack of focus. From the outset, an evaluationshould have clearly specified goalsand objectives. As discussed previously, itis erroneous to think in terms of an assessmenttechnique or battery being valid orinvalid. Results of the evaluation can bevalid for specific interpretations. Therefore,what interpretations one anticipates makingat the end of the evaluation shouldguide the selection of tests for the assess-

chapter 5 Planning the Evaluation and Rapport Building83ment battery. For example, if an assessmentis primarily intended to determine schoolplacement, then the focus of the evaluationwill not be to determine whether ornot a psychiatric diagnosis is warranted,but to determine whether or not the childmeets the eligibility requirements of theschool system. The assessor may feelthat more information is needed to makeappropriate recommendations to meet achild’s psychological and educational needsthan is required by these criteria. However,enough information to determine the eligibilityshould be part of the assessment,if this is the primary referral question.In our experience, it is not uncommonfor an otherwise sound and competentlyconducted evaluation to be useless forthe specific purposes for which a childwas referred.There are many other examples showinghow the intended use of the assessmentinformation determines the measures tobe used. This may be as broad as definingwhat areas need to be covered for acertain purpose (e.g., some residentialtreatment centers require a personalityassessment prior to acceptance) to as specificas requiring certain tests (e.g., someschool systems require specific tests to begiven for special education placement).The assessor should not give a test that,in his or her professional judgment, isinappropriate for a particular use or isinappropriate for a particular client.However, if, at the time of referral, theintended use of the assessment is clarifiedand there is some question as to howappropriate certain requirements are fora given case, the assessor can attempt toaddress these issues before beginning theevaluation.Often the person or agency referring achild or adolescent for testing is not surehow the test results will be used. Instead,the child is referred because the agency isunsure of the nature of a child ’ s problem (oreven whether there is a problem), and thereferrer is unsure of what can be done tohelp the child. There are many variationson this theme, but, in essence, the goal ofthe assessment is to diagnose the source ofa child ’ s difficulty and to make treatmentrecommendations based on this diagnosis.In Chap. 3, we discussed many importantissues in making diagnostic decisions.However, Martin (1988) provides a succinctand practical analysis of the specificgoals involved in diagnosis. These areto (1) predict future behavior, (2) differentiatebetween abnormal and normalbehavior, (3) make differential diagnoses,and (4) delineate individual differencesin competencies and disabilities. Martinalso provides some interesting recommendationsfor planning the evaluationto maximize the reliability of the diagnosticprocess. These are summarized inBox 5.1.Description of ReferralProblemsIn addition to understanding the purposeof the testing referral, it is also importantto obtain an initial description of the difficultiesthat a child is experiencing thatled to the referral. One of the reasons thatclinical assessments are so fascinating isthat, if done right, the assessment is a typeof scientific inquiry. Based on the intakeinformation, the assessor should havesome initial hypotheses for understandinga given case that will be tested during theevaluation. These hypotheses will guidethe initial planning of the evaluation andinitial test selection. As in any good scientificendeavor, we must be clear of thedata that would support and those thatwould not support the various hypotheses.In contrast to many other scientific enterprises,however, the hypotheses can, andshould, change during the investigation.

84 chapter 5 Planning the Evaluation and Rapport BuildingBox 5.1Planning the Evaluation to Enhance ReliabilityIn our chapter on psychometric theory, we discussedreliability as a key concept in understandingthe psychological measurement. Reliabilityis often considered as a property of individualtests. However, Martin (1988) discusses severalissues in planning an assessment battery thatcan maximize the reliability of the informationthat is obtained. Key to Martin’s approach is hisconceptualization of four primary sources oferror variance that can affect the reliability ofmeasurement of children’s social and emotionalfunctioning:(1) Temporal variance – changes in behaviorover time(2) Source or rater variance – differences ininformation due to characteristics of theinformant(3) Setting variance – differences due to differentdemand characteristics across settings(4) Instrument variance – unreliability inherentin individual instrumentsMartin uses the basic concept in measurementtheory to describe how these sources of errorvariance can be controlled in an assessment.Specifically, the primary method of controllingerror variance and increasing reliabilityis through aggregation. As the length ofa test increases, the reliability of the scoresincreases. Thus, to control the temporal variance,repeated measurements on several occasionsshould be obtained. Similarly, to reducesource and setting variance, informationshould be obtained from multiple sources andacross multiple settings. The implication ofthese psychometric considerations is the needfor a comprehensive evaluation.The final source of error variance in Martin’ s scheme is the instrument variance. Likethe other sources of variance, aggregatinginformation across instruments is a crucialmethod for increasing reliability. However,this is only the case if additional tests providereliable information. If one adds unreliabletests to a battery, then aggregationactually decreases the reliability of the battery.Clinicians, who have a favorite test that theyuse in the batteries, will often justify theiruse of the test, even if it has been provenunreliable, by the statement “I only use it asone part of a more comprehensive battery.”This is clearly better than using the test inisolation. However, adding a piece of unreliableinformation will only reduce the reliabilityof the aggregated information. In aseparate publication, Martin (1982) gives theexample of three umpires calling a baseballgame. If one of the umpires is blind, his callswill only serve to reduce the reliability ofthe calls made by the entire umpiring team.The moral of the story: Aggregation onlyincreases the reliability of the informationobtained if the individual tests are selectedto enhance reliability.Source: Martin (1988). Assessment of Personality and Behavior Problems Infancy through Adolescence. New York:Guilford Press.As data accumulate on a case and it becomesclear that initial impressions of a case werewrong, the assessor must revise the assessmentaccordingly. To employ this scientificapproach to clinical assessment, enoughpreliminary information on a child’s functioningmust be obtained prior to startingthe evaluation, so that initial hypothesescan be formed. A case example that utilizesthis approach is provided in Box 5.2.

chapter 5 Planning the Evaluation and Rapport Building85Box 5.2A Scientific Approach to Clinical Assessment: A Case ExampleJoshua is a 10-year-old boy who was referredto the outpatient psychiatry department of alarge inner-city pediatric hospital for testing.The intake worker determined that Joshuawas being referred by his parents becausehe was in danger of failing the fifth grade.According to the intake information, Joshuawas having great difficulty paying attentionin class and completing assignments.He was also described as being excessivelyfidgety and restless. The intake informationindicated that these school problems werenew this school year. He had been an A\Bstudent in the four previous school grades,which made his current poor performanceespecially puzzling.Based on this information, several initialhypotheses were formulated. It could be thatsimilar problems were experienced in the pastgrades but they had just increased in severityin the fifth grade; in which case, dispositionalcauses were possible such as an attention deficitdisorder and/or a learning disability. Alternatively,if this recent onset was supported inthe evaluation (through interviewing parentsabout past school performance, obtainingschool records, interviewing past teachers), itmay be that Joshua had experienced or was experiencingsome type of newly occurring stressor(e.g., parental divorce, sexual abuse) that wasresulting in the deterioration in behavior. Theevaluation was designed to test these initialhypotheses.Interestingly, during the assessment ofpotential stressors, Joshua’s mother reportedthat he had been involved in an automobileaccident during the summer prior to enteringthe fifth grade. He had sustained a closed headinjury and had lost consciousness for severalminutes. He was released from the hospitalwith no noticeable effects of the injury. Afterobtaining this information, another hypothesisbecame possible. Joshua might have sustainedneurological damage from the accident thatwas affecting his behavior. As a result, hewas referred for a neurological exam, whichuncovered neurological damage that seemedto be the most likely cause of his behavioraldifficulties. Although the initial hypotheseswere not correct, this illustrates how a scientificapproach to hypothesis testing can beuseful in structuring the assessment process.Designing the EvaluationWe concluded Chap. 3 by providing severalguidelines for clinical assessmentsof children that followed from researchin developmental psychopathology. Inthis section, we take these research-basedguidelines and use them to develop practicalconsiderations in designing clinicalassessments of children. Once again, theserecommendations are designed to providea generic framework that can be tailored tothe needs of the individual case.Developmental ConsiderationsFrom the discussion of developmental psychopathologyprovided in Chap. 3, it is clearthat assessments of children’s emotional andbehavioral adjustment need to be sensitiveto a number of important developmentalissues. First, a basic tenet of developmentalpsychopathology is the importanceof taking a “process-oriented” approachto conceptualizing children ’ s adjustment.As a result, it is important that clinicalassessments not only involve a standardizedand comprehensive assessment of a child’s

86 chapter 5 Planning the Evaluation and Rapport Buildingbehavioral and emotional adjustment, butit is also important that they include anassessment of the developmental processes(e.g., temperamental tendencies, familycontext) that may be related to the child’scurrent pattern of adjustment. This morecomprehensive assessment requires thatassessors maintain a current knowledge ofthe research related to the type of problemsthey encounter in their evaluations, sothat they have an adequate understandingof the processes that may be involved inthe development of these problems.This research provides the assessorwith some initial hypotheses regardingthe problems that led to a child’s referralfor testing. Measures are selected so thatthese hypotheses can then be tested in theevaluation. For example, there is researchto suggest that there are distinct subgroupsof children with conduct problems, someof whom react very strongly to peer provocationand emotional stimuli, and some ofwhom show a lack of reactivity to emotionalcues, leading them to ignore the potentialconsequences of their behavior on others(Frick, 2006). Solely assessing the child’slevel and severity of conduct problems andmaking a diagnosis of “Conduct Disorder,”without assessing the child’s affectiveand interpersonal style, would not allowone to distinguish between these differentsubgroups that may require differentapproaches to treatment (Frick & McMahon,2008).Secondly, it is important for the psychologistto consider the developmentalstage of the child to be assessed in designingan assessment battery. For example, itis important, when selecting tests for a battery,to determine whether the tests providegood norm-referenced scores for thedevelopmental stage of the child or adolescentbeing assessed. Because this is soimportant, a significant focus of the laterchapters (which provide reviews of specifictesting instruments) is on the descriptionof the instruments’ norm-referencedscores. It is evident from these reviewsthat the adequacy of these scores can varyacross developmental stages (e.g., havinga very limited normative sample for olderadolescents). In addition to specific tests,some testing modalities may be more orless appropriate depending on the developmentallevel of the child. For example,in the chapter on structured interviews, wediscuss research suggesting that the childself-report format on these interviews maybe unreliable before age 9.Determining the RelevantPsychological DomainsA fairly ubiquitous finding in researchon childhood psychopathology is thehigh degree of overlap or comorbidity inproblem behaviors (Jensen, 2003). Thatis, children with problems in one area ofemotional or behavioral functioning are athigh risk of having problems in other areasof emotional or behavioral functioning, aswell as problems in social and cognitivearenas. In addition, a key assumption to adevelopmental approach to understandingchildren’s adjustment is that all outcomesare influenced by multiple interactingprocesses. As a result, most evaluationsof children and adolescents must be fairlycomprehensive to ensure that all areas thatcould be relevant to treatment planningare assessed. In planning an evaluation, oneshould consider the most likely comorbiditiesassociated with the referral problemand the most likely factors that can leadto such problems, and design the evaluationto provide an adequate assessment ofthese areas. From the referral information,one may also gather some clues as to howintensive the assessment of these potentiallyimportant domains should be.For example, consider a referral of a7-year-old boy who is having significantproblems of being disorganized, being veryimpulsive, and having difficulty staying in

chapter 5 Planning the Evaluation and Rapport Building87his seat. An initial hypothesis may be thatthe child has attention-deficit hyperactivitydisorder (ADHD), and an evaluation isdesigned to test this hypothesis and to testthe many different processes (e.g., poorexecutive functioning) that could lead to thisdisorder (see Chap. 17). In addition, fromresearch on ADHD one knows that approximately30% of children with the disorderhave a co-occurring learning disability (Frick& Kimonis, 2008). Therefore, one needsto determine how this comorbidity can beassessed. However, in the initial intake, thechild’s mother states that her son has no realproblems academically, other than losing hisassignments frequently, and, in fact, he hasonly made two B’s on his report cards sinceentering school. Based on this piece of information,one may decide not to conduct anintensive evaluation of a potential learningdisability unless, during the course of theevaluation, some evidence of learning problemsis discovered.Screening of ImportantContextsResearch has indicated that children’sbehavior is strongly influenced by factorsin their psychosocial environment.Therefore, an important considerationin planning an evaluation is determiningthe aspects of a child’s environment beingassessed (e.g., teaching styles of specificteachers, affective tone of family interactions)and the assessment methodology(e.g., naturalistic observations, behaviorrating scales). However, the relevance ofcontext will vary from child to child. Theintake information should provide enoughinformation so that an evaluation can beplanned, in which (1) informants fromeach of a child ’ s relevant contexts provideinformation on the child ’ s functioningand (2) the contexts that seem to have themost impact on a child ’ s functioning can beassessed in greater detail.One of the most influential contexts forthe majority of children is the family. Achapter in this text (Chap. 12) is devoted tothe assessment of a child’s family environment.However, what constitutes a familyfor a child is becoming increasingly diverse,and the intake can yield some preliminaryinformation on the family structure (e.g.,marital status of parents, degree of contactwith non-resident parents, other adult caretakersin the home) that provides the assessorwith some clues as to the best methodof structuring an evaluation of the familycontext.Practical Considerations inDesigning an EvaluationIn an important clinical endeavor like psychologicaltesting that can have importantconsequences for a child, one does not liketo consider mundane factors such as timeand expense in designing the evaluation.Clearly, these factors should not outweighwhat is in the best interest of the childbeing tested. However, sometimes thesefactors are unavoidable and often expediencyis in the best interest of the child.For example, an adolescent who has animpending court date for a juvenile offensemay need to have an evaluation completedbefore this date to help in determining themost appropriate placement and the mostappropriate services. One should take carenot to be so influenced by expediency thattreatment decisions are misguided by poorassessment results. But one must considerwhat can be meaningfully obtained withinthe available time frame, and possibly makeas part of the outcome of the evaluation arecommendation for the additional testingthat might be beneficial as time allows.How much to weigh cost and timeconstraints will vary from case to case.However, we feel that one should ask thefollowing two questions in designing anyevaluation:

88 chapter 5 Planning the Evaluation and Rapport Building1. What is the essential information neededto answer the referral question(s)?2. What is the most economical meansof obtaining this essential informationwithout compromising the usefulness ofinformation?To Test or Not to TestWhen a child is referred for testing, an importantquestion that should be asked is whether ornot an evaluation is in the child’s best interest.We feel that simply because an evaluation isrequested by someone is not sufficient reasonto conduct the evaluation. A professionalmust make the decision as to whether or notan evaluation is likely to benefit the child oradolescent. Often this question is ignored forfinancial reasons. If you don’t do the evaluation,you don’t get paid. However, we feelthat a clinical assessor has the ethical obligationto estimate the potential benefit of theevaluation to the child and then convey thisdetermination to the referring agency.There can be several reasons why anevaluation would not be in a child ’ s bestinterest. For example, a child ’ s parentmay seek multiple evaluations becausethe parent does not agree with the findingsof previous evaluations. We feel thatsecond opinions are not inappropriate inmany cases. However, if this is not consideredcarefully, a child may be subjectedto numerous intrusive evaluations thatare not necessary and the evaluator mayinadvertently reinforce a parent’s denialof a child ’ s special needs. Alternatively, theperson referring a child or adolescent mayhave unrealistic expectations from whatan evaluation can accomplish, or the reasonfor the evaluation may be insufficientto justify performing the evaluation. Anexample that illustrates both of these issuesis a child who is referred by a parent todetermine his future sexual orientation.Even if one determines that a child oradolescent may benefit from an evaluation,one must also question whether or not theassessor is the appropriate person to conductthe evaluation. The appropriatenessof an assessor may simply be a matter ofone ’ s competence, either because of uniquecharacteristics of the child (e.g., age, culture)or because of the specific nature ofthe referral question. Assessors must holdclosely to the principle noted in the Standardsfor Educational and PsychologicalTesting published jointly by the AmericanEducational Research Association,American Psychological Association, andNational Council on Measurement Educationthat “test users should not attemptto interpret the scores of test takers whosespecial needs or characteristics are outsidethe range of the user’s qualifications”(Standard 11.3, p. 114).In addition to competence, a clinicalassessor must also question whether or notpersonal reasons might prevent him orher from conducting an objective evaluation.For example, an examiner may have apersonal relationship with a child or familythat might interfere with the ability toobjectively administer and interpret tests.Alternatively, the assessor may have personalissues related to the referral problemthat might prevent him or her from beingable to competently perform the evaluation.For example, a psychologist whohimself is dealing with memories of a pastsexual abuse may not be able to conductan evaluation of another sexual abuse victimbecause he is unable to transcend hisown issues related to the abuse. Thereare no specific guidelines for determiningwhen personal issues would interferewith an evaluation. Our point is to suggestthat assessors should routinely questionwhether or not they are appropriateto conduct an evaluation, and they shouldconsult with colleagues if there is any questionregarding their ability to competentlyconduct the evaluation.

chapter 5 Planning the Evaluation and Rapport Building89Rapport BuildingThere is no aspect of the assessment processthat is as difficult to define and toteach as the concept of rapport. However,rapport is a critical component of testingchildren and adolescents (Fuchs & Fuchs,1986), although it is rarely discussed inchild assessment texts (e.g., Achenbach &McConaughy, 1987; Ollendick & Hersen,1993) or in administration manuals for testsdesigned for children (see Fuchs, 1987).The Longman Dictionary of Psychology andPsychiatry (1984) defined rapport as “awarm, relaxed relationship that promotesmutual acceptance, e.g., between therapistand patient, or between teacher and student.Rapport implies that the confidenceinspired by the former produces trust andwilling cooperation in the latter” (p. 619).To paraphrase and apply this definitionto the testing situation, rapport refers tothe interactions between the assessor andthe person being assessed (client) thatpromote confidence and cooperation inthe assessment process. Rapport buildingis not something that is done at the outsetof testing and then forgotten. Instead,it is a process that evolves throughout theentire assessment endeavor (Barker, 1990;Sattler, 1988).The importance of rapport is not specificto psychological assessment; it is acritical concept in most clinical endeavors.There are several recommendations thatcan be drawn from other clinical situationsthat apply equally well to child testing.For example, Phares (1984) described thebasic elements of establishing rapport inpsychotherapeutic relationships as “havingan attitude of acceptance, understanding,and respect for the integrity of the client”(p. 195). Phares goes on to point that thisattitude is not synonymous with establishinga state of mutual liking but is morerelated to a clinician ’ s ability to convey tothe client a sincere desire to understand hisor her problems, and to help him or herto cope with them. This general attitudeof the assessor is the basic component ofestablishing rapport with the client. As aresult, our specific recommendations aredesigned to foster this attitude in testingsituations.While the importance of rapport is notconfined to the psychological assessmentof children and adolescents, there are severalunique aspects to the assessment ofyouth that make rapport building a complicatedprocess in this context. First, theclinical assessment of children typicallyinvolves many people (e.g., child, parent,and teacher) who have varying levels ofunderstanding of the assessment process,and who possess varying levels of motivationfor the assessment. Therefore, theassessor must be skilled in enlisting andfostering the cooperation of many differentparticipants. The issue of motivationis especially salient in the evaluation ofyouth because children and adolescentsare often not self-referred. Children areoften referred for evaluations becausetheir behavior causes problems for significantothers in their environment (Frick &Kimonis, 2008). Therefore, enlisting theircooperation and trust is a critical, but oftendifficult, process. Later in this chapter weprovide examples of how testing can bepresented to children and adolescents inways that foster the establishment of aworking relationship.A second factor that complicates thedevelopment of rapport in testing situationsis the presence of severe timelimitations. In many, if not most, testingsituations the assessor has limited timefor rapport building with all participants.Often testing is confined to one or severaldiscrete testing periods and testingstarts early in the first session. This isquite different from the many other clinicalcontexts, such as the psychotherapeuticcontext, in which there is likely to bemore flexibility in the time allowed for

90 chapter 5 Planning the Evaluation and Rapport Buildingdeveloping rapport prior to the initiationof some clinical intervention.Based on this discussion, it is evident thatestablishing rapport in the typical assessmentsituation for children and adolescentsinvolves enlisting the cooperation ofmultiple participants to divulge personal andsometimes distressing information, despitea potential lack of motivation and despitethe fact that the testing must be completedwithin a limited time frame. It is obvious fromthis description that rapport building is notalways an easy task in clinical assessments ofyouth. Therefore, it is important to outlinethe important considerations in the developmentof rapport in clinical assessments ofchildren and adolescents.Informed ConsentWe view informed consent in two ways inthis book. The first, which is the more traditionalway, is to view it as a legal and ethicalright of the recipients of any psychologicalservice. The assessor has the responsibilityof ensuring that informed consent is providedfor the assessment. However, we alsoview informed consent in a second way:as a basic element of rapport building. Asdiscussed previously, a fundamental elementin developing rapport is expressinga respect for the individual participatingin the evaluation. There is no more basicway of conveying respect than by placinggreat importance on the informed consentprocess.In Chap. 4, we discussed the legalrequirements of obtaining informed consentfrom a child ’ s legal guardian. However,the assessor can communicate asincere respect for the child ’ s guardianby spending a great deal of time reviewingall the testing procedures in very clearand specific terms, by discussing the limitsof confidentiality in sensitive terms, byclearly reviewing the intended uses of thetest results, and by allowing and encouragingthe parents to ask questions about theseissues. In essence, the assessor should conveyto the parent that the consent proceduresare not just a legal formality, but areintended as the first step in establishing acollaborative effort between the parent andassessor. Also, there is no greater damageto the development of rapport than a parents’perception that some procedures wereused without his or her full knowledge andconsent.The need to transcend legal requirementsis even more important with thechild. With the view that minors may notbe competent to make decisions regardingtheir need for certain medical or psychologicalprocedures, like psychologicaltesting, the right to informed consent generallyrests with a child’s parent or legalguardian. Unfortunately, many assessorstake this to mean that a child does not havethe right to have procedures explained tohim or her in understandable language.Although in some situations we agree thata child may not have the right to refuseparticipation in an evaluation, we feel thatin all situations, irrespective of a child ’ s age,the assessor should explain to the child allthe procedures that he or she will undergoas part of the testing. Clearly, the degreeof depth and sophistication of this explanationshould be made in recognition of thepossible fears about the evaluation that achild or adolescent might experience andwith recognition of his or her varying levelsof motivation. Boxes 5.3–5.5 provideexamples of how testing procedures canbe explained to children and adolescentsof various ages in ways that enhance theestablishment of rapport.Discussing testing with the child oradolescent is critical for conveying respecttowards the child, and helps enlist the childas a collaborative participant in the process.It reduces the feeling of the child that thetesting is being done to him or her ratherthan for or with him or her. Also, manychildren arrive for testing with substantial

chapter 5 Planning the Evaluation and Rapport Building91Box 5.3Explaining Testing to a 5-Year-Old BoyWe have argued that all children should havetesting procedures explained to them in termsthat are understandable given their developmentallevel. This is a crucial aspect of developingrapport with a child. However, manybeginning clinical assessors have difficultydescribing testing in terms comprehensible toyoung children and fail to recognize some ofthe fears and motivations that children bringto the evaluation. The following is an exampleof an explanation of procedures that is givento a 5-year-old boy referred to a private psychologistfor testing.“Hello, Johnny. My name is Dr. Test. I’mnot the type of doctor you come to whenyou’re sick, like with a stomachache or headache,but I’m the type of doctor who likes toget to know kids better, like how they feelabout some things and how they act sometimes.So what I’m going to do today is findout a lot more about you. I’m going to askyou to draw some pictures for me and tellme about them. I also have some picturesand I want you to make up stories aboutthem. And then, I have a bunch of questionsabout how you feel about certain things thatI’m going to help you answer. We will haveto work pretty hard together but I think itwill be fun, too. We’re going to take a lot ofbreaks and please let me know if you needto stop and go to the bathroom. Now, yourmom and dad have already been telling mea lot about you and I’m also going to betalking to your teacher at school. After I dothis, I’m going to take what you tell me, andwhat your parents and teacher tell me andtry to get a good picture of what you ’ re like,how you feel about things, all the thingsyou ’ re doing well, and anything you mightneed help in. And then I will talk to yourparents and to you about what I find andlet you know if there is anything that I cansuggest that might help you.”This explanation is designed to be anexample of the types of terms and phrasingthat can be used in explaining psychologicalprocedures to very young children. Ascan be seen from the content of the explanation,we feel that in this age group, oneof the most important sources of anxiety isthe fear of the unknown. Therefore, we tryto let the child know that the procedureswill be pretty innocuous (e.g., answeringquestions, drawing). Obviously the actualcontent of the description will depend onthe procedures that are planned. But wefeel strongly that all procedures to be usedshould be explained to the child, albeit in alanguage that is understandable.Also, to illustrate the level of explanation,the discourse was presented in a narrativeform. In actual practice it is helpful to involvethe child in the discussion by asking simplequestions (e.g., Do you like to draw?) andencouraging him or her to ask you questionsif there is anything he or she does not understand.This helps the child feel more respectedand valued in the assessment process. Finally,we often find it helpful in this age group topresent this information in the presence of thechild ’ s parent(s). When children see that theirparents are comfortable with the procedures,they often develop a greater sense of comfortthemselves.misconceptions about what the testing willentail (e.g., thinking that the psychologistis going to operate on their brain or thatthey will be punished for being bad). Simplyspending time to clearly review whythe child is being tested, what the childshould expect during testing, and whatwill happen with the test results helps toeliminate possible misconceptions andreduce unnecessary anxiety.

92 chapter 5 Planning the Evaluation and Rapport BuildingBox 5.4Explaining Testing to a 10-Year-Old GirlOlder, pre-adolescent children often have a betterunderstanding of the basic nature of the testingsituation than do younger children. However,the procedures should still be explained in veryclear and simple terms to ensure that there are nomisconceptions. In this age group, we find thatthe explanation must be sensitive to the potentialthreat to a child’s self-concept that the testing maypresent. One of the major emotional tasks duringthe pre-adolescent period is the development of asense of mastery and a sense of competence. Testingcan be a threat to a child in these areas forseveral reasons. First, just the term “testing” conveysthe possibility of failure. Secondly, the childmay have been implicitly or explicitly told thatthe reason for the testing is to see “what’s wrongwith you.” The explanation of testing in this agegroup should be sensitive to these issues. Here, weprovide a sample explanation to a 10-year-old girlreferred for a comprehensive evaluation.“Jessica, I want to explain exactly whatwe are going to be doing together today,and give you a chance to ask me any questionsyou may have. Your parents wereconcerned about some of the problemsyou have been having at school and theywanted to know if there was anything morethey could be doing to help you. In orderfor me to answer this question, I have tofind out a lot more about you-what you liketo do, how you feel about different things,what things you’re good at, what thingsyou might not be so good at. To do this,we are going to do a lot of different thingstogether. First, I am going to ask you to dosome reading and math problems with me.Then I will ask you to fill out some questionnairesthat will tell me how you feelabout different things, how you get alongwith kids in your class, and how you seeyour family. Finally, I am going to showyou some pictures and ask you to tell mesome stories about them. Before we starteach of these activities, I will tell you whatwe’re going to do and how to do eachthing. I promise to give you a chance to askme any questions you have about each task.I have already talked to your mother abouthow things go at home and I am going toask your teacher to fill out a questionnaireabout how she sees you at school. After Iget all the information, I should understandyou a little better and I will then talkabout what I found with you and your parents.Jessica, it is very important that youunderstand that I’m not looking for thingsthat are wrong with you. My guess is thatyou are like most kids. You have things thatyou ’ re good at and some things that you ’ renot so good at, and that there are thingsyou like and other things you don’t like. Iam just trying to get a good picture of allthese different parts of you.”Box 5.5Explaining Testing to an AdolescentThere are several crucial issues that one mustkeep in mind when explaining testing to anadolescent. First, adolescents spend a greatdeal of energy trying to convince people thatthey are no longer children. Therefore, onemust be very careful not to come across ascondescending to them. Secondly, because ofthe importance of peers in adolescence, adolescentsare very concerned with fitting in.Coming in for psychological testing may beviewed as a threat to this by making them feeldifferent from other adolescents. Therefore,the explanation should attempt to normalize thetesting as much as possible. Thirdly, privacy is(Continues)

chapter 5 Planning the Evaluation and Rapport Building93Box 5.5 (Continued)a major issue for adolescents. In testing, adolescentsmay be asked many personal questions.They must be warned of these questionsand informed as to how the information fromthe testing will be conveyed to other people.This is very threatening to most adolescents,and the explanation should be sensitive tothis issue. Fourth, a majority of adolescentsreferred for testing do not see the need forsuch testing and don ’ t want to be there. Amajor flaw we often see in presenting testingto adolescents is that the assessor tries to cajolethe adolescent into being happy to be thereand into appreciating the potential benefits oftesting. Clearly, the potential benefits of testingshould be discussed with the adolescent inan attempt to enhance motivation. However,this often has a minimal effect on motivation,and often one must simply acknowledge tothe adolescent that you understand that he orshe is not happy about being there but, if youwork together, you will get through it quicklyand relatively painlessly. The following is asample explanation of psychological testingprovided to a 16-year-old male.“Jeff, I want to explain what we will bedoing today and, please, feel free to ask meany questions about what I say. You probablyknow that your parents are concerned aboutyour behavior. They have seen some changesin you recently and they want to know if theycan do something more to help you. I understandthat you are not wild about being here,but if we work together, maybe we can see ifthere is anything that I can recommend tohelp you or at least put your parents’ mindsat ease. But if we’re going to get anything outof this we have to work together. I work witha lot of people of your age who don’t want tobe here at first, but end up getting a lot outof the experience. I will start by just askingyou about some of the things that have beengoing on with you lately to get your view onthings. I have already talked to your parentsabout their views of what’s going on. I alsohave some questionnaires for you to completeabout your feelings, your behaviors,and your attitudes. Some of these questionsare pretty personal, but they are importantfor me to get a better understanding of you.After the testing, I will summarize the resultsin a report and go over it with you and yourparents. At that time we can discuss anythingthat I think may help you.”Building Rapport with the ChildAs mentioned previously, the child is oftennot the one seeking an evaluation but isusually referred by some significant adultwho feels that the child or adolescent needsthe testing. Therefore, the motivation ofthe child for the evaluation is often low.Another reason for low motivation is thatthe child often realizes, or has been explicitlytold, that the evaluation is promptedby problems either at home or school. Asa result, the child is legitimately concernedabout the outcome of the evaluation (i.e.,getting into more trouble). In addition, thetesting situation is often unique in mostchildren’s experiences. Children have had afew similar experiences, and therefore theyoften have little idea of what to expect in thetesting situation. Finally, the many developmentalstages that characterize childhoodand adolescence imply that assessorsmust be familiar with development to beable to tailor their rapport-building strategiesto the unique needs of children at variousstages.We have already mentioned that rapportbuilding is a process that evolvesthroughout testing. It starts at the veryfirst contact between the assessor and the

94 chapter 5 Planning the Evaluation and Rapport Buildingchild. When an assessor greets a child, theassessor should (1) use a warm, friendly,and interesting tone, (2) be sure to greetthe child by name (don’t simply greet thechild’s parents), and (3) introduce him- orherself using his or her title (e.g., Dr., Ms.,Mr.). This last recommendation is a subjectof considerable debate by practicingpsychologists (Barker, 1990). However,we feel that using a title is important inthe time-limited, task-oriented assessmentsituation because it sets the stage that youare a professional (albeit a caring, friendly,and respectful one) who will be workingwith the child, and not a friend who willplay with the child.After informed consent, many authorsrecommend a period of time for discussinginnocuous and pleasant topics, suchas the children’s hobbies, pets, friends, orother interests (Barker, 1990). For youngerchildren, some authors even recommenda period of play to allow the children tobecome more accustomed to the examiner.In our experiences, such rapport-buildingstrategies should be used cautiously andsparingly. For many children, the assessormay be perceived as simply delayingthe inevitable by using these strategies.This could have the paradoxical effect ofincreasing their anticipatory anxiety. Inour experience, one of the best rapportbuildingstrategies is to begin the assessmenttasks quickly, so that the child beginsto realize that the procedures will not be asbad as they imagined.Periods of play before the evaluation areespecially problematic if structured testingis to follow. Young children often havedifficulty switching from unstructured tostructured tasks (Perry, 1990). Therefore, itis usually best when testing preadolescentchildren to start with the more structuredparts of the evaluation (e.g., rating scales,structured interviews) rather than startingwith less structured tasks (e.g., projectivedrawing tests). This is not only becauseof the greater difficulty in switching fromunstructured to structured tasks, but alsobecause the structured tasks have clearerdemand characteristics. That is, it is usuallyquite clear to children what is expected ofthem on these tasks and this, in turn, helpsthe children become more comfortable ina situation that is different from anythingthey have experienced in thepast.Box 5.6 provides a summary of someadditional rapport-building strategies foruse with children that were proposed byBarker (1990) in his book on interviewingchildrenBuilding Rapport with theParentThere are also some unique considerationsin building a working relationship with achild’s parents. Of course, the importanceof rapport with parents will depend on thedegree of their involvement in the testing.However, in most situations their involvementwill be substantial. Although manyevaluations are conducted at the request ofa parent, there are also many situations inwhich a child is referred by others (e.g.,school, court), and, in these situations,building rapport with the child’s parent iscritical. Under these circumstances, theassessor must allow the parent to expresshis/her views on the need for evaluationprior to the testing process. The assessorneed not necessarily agree with theseviews, but the assessor should convey tothe parents a sincere interest in understandingtheir views in order to build aworking relationship with them.Even for parents who have initiated thereferral for testing, the assessor should beaware of the potential threat to a parent ’ sself-esteem that many testing situationspresent. For many parents, acknowledgingthat their child might have sometype of disability is quite traumatic andcan evoke a sense of failure. Also, parentsoften struggle with guilt blame for

chapter 5 Planning the Evaluation and Rapport Building95Box 5.6Rapport-Building StrategiesBarker (1990), in his book on conducting clinicalinterviews with children and adolescents,discussed several helpful strategies for establishingrapport. These can be summarized asfollows:1. A critical basis for rapport building isan assessor’s communication style. Theassessor who is able to adopt a warm,friendly, respectful, and interested communicationstyle is more likely to developa good working alliance with a child.2. The assessor’s physical appearance can alsoenhance rapport. Overly formal dress canmake a child feel ill at ease.3. Assessors should attempt to conform his orher posture, movements, speed of speech,voice tone and volume, etc. to the style ofthe person being tested. This should bedone sensitively and unobtrusively.4. Assessors should tailor their vocabulariesto match the vocabularies of theperson being tested. Few things impedethe establishment of rapport as much asrepeatedly using words and expressionsthat are unfamiliar to those with whomyou are speaking.5. Respect the views of those you are testing.This does not necessarily mean agreeingwith or approving of the views expressed.6. Occasionally the assessor should adopt aone-down position. To reduce the intimidationthat children sometimes feel withexperts, the assessor can sometimes aska child, from a position of ignorance,about something with which a child hasexpertise, such as video games, televisionshows, or soccer.7. Taking time during the testing to talk ofexperiences and interests that the assessorand child have in common can alsoincrease the trust between the assessor andthe child.Barker (1990) also emphasizes that thedevelopment of rapport is continuousthroughout the testing process. “Rapportcan always be developed further; the reverseis also possible. Although it is certainly truethat once it is well established, rapport canwithstand a lot of stress, it nevertheless canbe damaged or even destroyed at any timeif continuing attention is not paid to maintainingit” (p. 35).Source: Barker (1990). Clinical Interviews with Children and Adolescents. New York: Norton.their child ’ s problems and may be concernedthat testing will confirm theirpotential role in their child’s difficulties.An assessor should be sensitive to thesedynamics and allow the parents to expresstheir concerns at some point during thetesting. Additionally, the parents should besupported in their role of getting help fortheir child. For example, an assessor mighttell the parents how lucky their child is tohave parents who care enough to obtainhelp for him or her, and not just let thingsget worse. This helps to reframe the testingsituation as one that could increase theparents’ self-esteem, rather than one that isa threat to their self-concept.Several reasons were given for startingwith structured tasks in testing childrenin an effort to enhance rapport. In ourexperience, the opposite is true in rapportbuilding with parents. Even prior toobtaining specific background informationfrom a parent, it is important to let theparent discuss his or her concerns aboutthe child in an unstructured format. Theunstructured clinical interview is discussedin more detail in a later chapter. However,having such an interview at the start of the

96 chapter 5 Planning the Evaluation and Rapport Buildingevaluation conveys to the parent (1) a genuineconcern with his or her perceptionsof their child ’ s adjustment and (2) that theevaluation will be personalized for theindividual child. If parents are immediatelyasked to fill the rating scales or administera structured interview as part of a standardevaluation, they often develop the impressionthat the assessor is more interested inadministering tests than in actually understandingtheir child ’ s needs. As one wouldexpect, such an impression is very damagingto the development of rapport.personally request the teachers’ participationin the evaluation, acknowledging andthanking the teacher for his or her efforts,rather than simply sending the material tothe teacher via the child, parent, or mail.Such a call is a professional courtesy thatgreatly enhances the collaborative effort. Itsets the tone for the teacher being involvedin the evaluation as a valued professionalwho has much to offer in the assessmentof the child.ConclusionsBuilding Rapport with TeachersIt is becoming increasingly clear that evaluationsof children must involve informationfrom teachers (Loeber, Green, & Lahey,1990). The degree of teachers’ involvementvaries considerably depending on the focusof the evaluation. However, many assessorswho are not used to working in schoolsettings find themselves ill-equipped tocollaborate with teachers to conduct psychologicalevaluations (Conoley & Conoley,1991).In the introduction to the concept ofrapport, we defined the basic ingredient torapport building as exhibiting an attitudeof respect towards the client or informant.Although many psychologists work hard inrespecting and developing rapport with parentsand children, often this respect is lostwhen dealing with other professionals, suchas teachers. A key to demonstrate this attitudeis by respecting the importance of teachers’time. Scheduling phone calls during teacher’splanning times, eliminating all but the mostessential work for the teacher, and alwayspersonally thanking the teacher for his or herefforts in the evaluation are very simple, yetimportant, rapport-building strategies.If a teacher is sent assessment material forcompletion (e.g., rating scales), it is importantfor the assessor to call the teacher andIn this chapter, some non-specifics of theclinical assessment of children were discussed.That is, a successful evaluationis not simply a matter of appropriatelyadministering and interpreting psychologicaltests. It is also dependent on anassessor’s ability to provide an appropriatecontext in which the testing takes place.The first major issue discussed was theimportance of good planning. A good evaluationis focused and goal-oriented. The purposeof the evaluation and the intended usesof the assessment results will have a majorimpact on how the assessment is structured.Enough information should be availableprior to actual testing so that the assessor hassome initial hypotheses to be tested in theevaluation.The second part of the chapter isfocused on rapport-building strategies withall participants in the evaluation. Developinga collaborative, respectful, and trustingworking relationship is crucial to a successfulevaluation. Being able to developrapport is a skill that often takes yearsof practical experience to develop fully.However, in this chapter we have tried tohighlight some of the important issues inrapport building with children and adolescentsof various ages. We have also triedto make some practical recommendationsthat address these issues.

chapter 5 Planning the Evaluation and Rapport Building97Chapter Summary1. The first step in planning an evaluationis to clarify the reason for referral,both in terms of the purpose of testingand the types of behavior that led to thereferral.2. Two important decisions that a clinicalassessor should make prior to starting anyevaluation is whether or not a formalevaluation is warranted and whether heor she is the most appropriate person toconduct the evaluation.3. In addition to competently administeringtests, clinical assessors must createan appropriate environment withinwhich the evaluation can take place.4. Building rapport with a child refers todeveloping a collaborative and supportiverelationship with the child for thepurpose of conducting the evaluation.5. Building rapport with other importantpeople who will be involved in the evaluation(e.g., parents, teachers) is alsocritical to the assessment process.6. A thorough and sensitive informed consentprocedure can play a major rolein showing respect to the child client,and his or her parents and thereby cangreatly aid in the establishment of rapport.7. An explanation of the testing procedureswith a child must be sensitive to alarge number of motivational and developmentalissues.

P a r t I IAssessment Methods

C h a p t e r 6Self-Report InventoriesChapter Questionsl What are the strengths and limitationsof self-report measures for child andadolescent assessments?l What are some of the key differencesbetween omnibus measures and singledomain measures?l How does the MMPI-A differ from itsadult counterpart?l Which of the self-report measures possessesgood evidence of content validity?l What are validity scales and how canthey be used to interpret self-reportmeasures?Omnibus PersonalityInventoriesThe use of self-report inventories withchildren is a relatively new phenomenon.It was heretofore commonly believed thatchildren could not accurately report ontheir own feelings, perceptions, and behaviors.As a result, parent and teacher reportshave routinely been preferred over the useof self-report inventories in child personalityassessment. One of the first popular childassessment instruments, for example, thePersonality Inventory for Children (PIC;Wirt et al., 1984), resembled a “junior”MMPI. It included a large item set similar to101P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_6, © Springer Science+Business Media, LLC 2010

102 CHAPTER 6 SELF-REPORT INVENTORIESthe MMPI, and the name conveys similaritiesto omnibus personality inventories. Yet, thePIC was, and is (PIC-2; Lachar & Gruber,2001) a parent rating scale.As a result, many of the omnibus scalesdescribed in this chapter are relatively new.This is a result of the growing consensus thatat least older children and adolescents can provideuseful information about their feelingsand behavior. However, as with any assessmenttool, there are limitations to the reliability,validity, and usefulness of any self-reportmeasure, and these limitations should betaken into account when designing an assessmentbattery and interpreting its results.Unfortunately, all of the complexityof using and interpreting omnibus selfreportinventories cannot be conveyed inone chapter. Most of the inventories discussedherein have entire volumes devotedto their interpretation. The following discussionserves primarily as an introductoryguide for studying the larger literature thatis available for an instrument. The eventualuser of any of the omnibus inventoriesdiscussed will have to spend considerabletime with the test manuals and additionalreadings and seek supervised interpretationpractice. What has become more clear inrecent research is that despite their relativelack of validity for the assessment of veryyoung children, self-report inventories canprovide invaluable information as to theyouth’s perception of his/her functioningand the factors that ameliorate or exacerbatehis/her problems.Whether or not a self-report inventoryhas a place in a psychological assessment willdepend on many factors, including the client’sdevelopmental level, presenting problem, andthe purpose of the assessment. This chapterin no way represents an exhaustive review ofthe self-report inventories available. Instead,we have reviewed what appear to be the mostwidely used and/or well-researched measures.In addition, although the appropriate uses ofself-report measures are highlighted, thereare limitations to any assessment technique.For self-report rating scales in particular,one must consider the client’s comprehensionof items and the potential for responsesets including socially desirable responsetendencies. The instruments reviewedbelow vary in the degree to which they haveappeared to take these factors into account.Therefore, the burden rests with the clinicianto select tools that will answer the referralquestion in a legitimate, comprehensive, andcost-effective manner (Table 6.1).Behavior Assessment Systemfor Children Self-Report ofPersonality (BASC-2-SRP;Reynolds & Kamphaus, 2004)The Behavior Assessment System for Children,2nd edition-Self-Report of Personality (BASC-2-SRP; Reynolds & Kamphaus, 2004) isbased closely on its predecessor, albeit withsome unique features. For example, theBASC-2 includes a college report form forstudents aged 18–25. Our discussion, however,will focus on the child version (SRP-C) for ages 8–11 and the adolescent version(SRP-A), which is normed for ages 12–21.Interested readers should consult theBASC-2 manual (Reynolds & Kamphaus)for a discussion of the development of theunique college student report version. Aswith the previous version of this instrument,the BASC-2-SRP is the componentof the BASC-2 that attempts to gaugethe child’s perceptions and feelings aboutschool, parents, peers, and his or her ownbehavioral problems. The estimated minimumreading level of the SRP is the thirdgrade. The SRP may also be read to childrenin order to ensure comprehension ofthe items. According to its authors, it takesapproximately 20–30 min to administer.Scale ContentThe SRP includes 16 scales: 12 clinical and4 adaptive. The scales were developedusing a combination of rational, theoretical,

CHAPTER 6 SELF-REPORT INVENTORIES103Table 6.1 Overview of Self-Report InventoriesInventory Ages ContentBehaviorAssessmentSystem forChildren,2nd edition(BASC-2;Reynolds &Kamphaus,2004)AchenbachYouth Self-Report (YSR;Achenbach& Rescorla,2001)8–11;12–21;18–25139–185 items; Emotionalsymptoms index, inattention/hyperactivity, internalizingproblems, school maladjustment,personal adjustment;3 validity scales; 12 clinicalscales, 4 adaptive scales, 4optional content scales11–18 112 items; Internalizing,Externalizing, Competence;8 scalesMinnesotaMultiphasicPersonalityInventory-Adolescent(MMPI-A;Butcher et al.,1992)14–18 478 items; 10 clinical scales;6 validity scales; 15 contentscales; 6 supplementaryscales; 28 Harris-Lingoesscales; 3 Si scalesReadingLevelTime toAdministerNormSample Reliability Validity MI3rd grade 20–30 min E Internal consistency(good);4–8 week test–retest (good)Good relationswith similartools; supportfor factorsE5th grade 15–20 min G Internalconsistency(good); 7 monthtest–retest (moderate);interrater(low to moderate)5th–6thgrade90 min G Internal consistency;test–retest; contentscales betterreliability thanclinical scalesGood validity(clinical vs. nonclinicalgroups);limited info. onrelations withother toolsItem overlapdetracts fromcontent validity;scores do notmatch 4 factorsolution foundin researchEP(Continues)

104 CHAPTER 6 SELF-REPORT INVENTORIESTable 6.1 (Continued)Inventory Ages ContentReadingLevelTime toAdministerNormSample Reliability Validity MIConners3rd Edition,Self-Report(Conners-3;Conners,2008a)PersonalityInventory forYouth (PIY;Lachar &Gruber, 1994)8–18 59 items; 39-item shortform; 3 validity scales, 6empirical scales, 1 rationalscale, 5 DSM-IV-TR symptomscales9–18 270 items, 80 item screener;9 clinical scales (several subscales);4 validity scales3rd grade 20 min E Internal consistencyand test–retest (good)3rd–4thgrade30–60 min G Internalconsistency;test–retest;good reliabilitybut lower forsubscalesGood evidenceof criterionrelatedanddiscriminantvalidityGood validity(clinical vs. nonclinicalgroups);some supportfor criterionrelatedvalidityE aGNote: Norm sample evaluated as E = excellent, G = good; MI = degree of enhancement of muti-informant assessment; E = Excellent, G = Good, P = Poora We view the Conners-3 as providing an excellent means of assessment of externalizing problems; however, if an extensive assessment of internalizing or learningdifficulties is desired, the Conners Comprehensive Behavior Rating Scales (CBRS; Conners, 2008b) is recommended.

CHAPTER 6 SELF-REPORT INVENTORIES105and empirical approaches to test development.Based on relatively low reliabilitiesof some SRP scales on the original BASCand preliminary analyses comparing SRPversions with different response formats,the authors developed the BASC-2-SRPto include both True-False items andFrequency-based items (i.e., Never, Sometimes,Often, Almost Always). Items fromthe original BASC were reviewed, particularlyfor reliability and developmentalappropriateness. Scales were definedbased largely on the original version, itemswere retained or new items developedbased on these definitions, and covariancestructure analysis was used to enhance thehomogeneity of scale content (Reynolds& Kamphaus, 2004). The SRP includesscales that are typical of self-report measures(e.g., anxiety, depression), as well assome that are relatively unique (e.g., selfreliance,locus of control; see scale definitionsin Table 6.2). The BASC-2 systemalso includes critical items that indicateclinically significant problems that warrantfurther follow-up assessment.Five composites were constructed for theSRP using factor analysis: Emotional SymptomsIndex; Inattention/Hyperactivity;Internalizing Problems; Personal Adjustment;and School Problems. The Inattention/Hyperactivitycomposite includesthe Attention Problems and Hyperactivityscales. The School Maladjustment compositeincludes the Attitude Toward School andAttitude Toward Teacher scales. The PersonalAdjustment composite assesses selfperceivedpersonal strengths and includesthe four adaptive scales (i.e., Relationswith Parents, Interpersonal Relations, Selfesteem,and Self-reliance). The Internalizingcomposite consists of the remainingsix scales (i.e., Atypicality, Locus of Control,Social Stress, Anxiety, Depression, andSense of Inadequacy) The ESI representsthe scores for the six scales with the highestloadings on an unrotated first factor (i.e.,Social Stress, Anxiety, Depression, Senseof Inadequacy, Self-esteem reversed scored,and Self-reliance reverse scored). Accordingto the authors, this index is “the mostglobal indicator of serious emotional disturbance”(p. 81).An additional feature, new to the adolescentversion of the BASC-2-SRP, is theinclusion of four content scales: AngerControl, Ego Strength, Mania, and TestAnxiety. These scales were formed viaa combination of rational and empiricalmethods and include some items that arepart of other SRP scales and other itemsthat are not part of another scale. Thesescales are considered optional and are notpart of the hand scoring of the BASC-2(Reynolds & Kamphaus, 2004). Obviously,research is needed regarding the usefulnessof these scales in clinical assessments, butthey may very well represent importantconstructs heretofore not assessed.Administration and ScoringThe SRP is available in three administrationformats: a scannable form, a carbonlesshand-scored form, and a computer-scorableform. The SRP has a reasonable administrationtime, given that there are 139 items onthe SRP-C and 176 items on the SRP-A.Although efforts were made to keep the itemstems brief and at a third grade reading level,some young people may still have difficultyusing the SRP. This and other self-reportinventories are not recommended for usewith children with significant cognitive deficitsor severe reading problems. Oral administrationor the use of the available audiotapefor the SRP is possible, but the child withcomprehension problems may still have agreat deal of difficulty with this procedure.Validity ScalesThe BASC-2-SRP includes three validityscales, as well as two indexes that alert theclinician to response patterns or inconsistencyin responses (i.e., the ResponsePattern Index and Consistency Index,respectively). The three validity scales are

106 CHAPTER 6 SELF-REPORT INVENTORIESTable 6.2 BASC-2-SRP Scale DefinitionsScaleAnxietyAttention ProblemsAttitude to SchoolAttitude to TeachersAtypicalityDepressionHyperactivityInterpersonal RelationsLocus of ControlRelations with ParentsSelf-EsteemSelf-RelianceSensation SeekingSense of InadequacySocial StressSomatizationDefinitionFeelings of nervousness, worry, and fear; the tendency to be overwhelmedby problemsTendency to be easily distracted and unable to concentrate more thanmomentarilyFeelings of alienation, hostility, and dissatisfaction regarding schoolFeelings of resentment and dislike of teachers; beliefs that teachers areunfair, uncaring, or overly demandingThe tendency toward bizarre thoughts or other thoughts and behaviorsconsidered “odd”Feelings of unhappiness, sadness, and dejection; a belief that nothing goesrightTendency to report being overly active, rushing through work or activities,and acting without thinkingThe perception of having good social relationships and friendships withpeersThe belief that rewards and punishments are controlled by externalevents or other peopleA positive regard for parents and a feeling of being accepted by themFeelings of self-esteem, self-respect, and self-acceptanceConfidence in one’s ability to solve problems; a belief in one’s personaldependability and decisivenessThe tendency to take risks and seek excitementPerceptions of being unsuccessful in school, unable to achieve one’s goals,and generally inadequateFeelings of stress and tension in personal relationships; a feeling of beingexcluded from social activitiesThe tendency to be overly sensitive to, experience, or complain aboutrelatively minor physical problems and discomfortsNote: From Reynolds & Kamphaus (2004).patterned after those found in the MinnesotaMultiphasic Personality Inventory(MMPI) tradition (see below). These scalesprovide a variety of checks on the validity ofa child’s results. The F Index is designed toindicate if a respondent may have answeredin an overly negative (i.e., “fake bad”)manner. The L Index does the opposite inthat it evaluates a tendency to respond in anoverly positive manner. The V Index onthe SRP consists of nonsensical items(e.g., “I have never been to sleep”) that, ifendorsed, would likely indicate carelessnessor lack of cooperation. The ResponsePattern Index (available in the computerscoring program) assesses the degree towhich an item response is the same as theresponse to the previous item. For example,marking “True” for 15 items in a row wouldyield a higher number in this Index andwould suggest that the respondent indicated“True” regardless of item content.The Consistency Index (also availablein the computer scoring program)is based on response patterns to itemsthat should be answered similarly. Theauthors also point out that the change toa mixture of True-False and Likert-type

CHAPTER 6 SELF-REPORT INVENTORIES107response formats on the new version ofthe SRP may also help safeguard againstresponse sets (Reynolds & Kamphaus,2004). It is still quite possible, however,to have the SRP invalidated by a response setor lack of cooperation. The validity indexesshould be examined at the outset of interpretation;further, the clinician should examinefor himself/herself response patterns thatdo not seem to fit the other evidence providedabout a case. In particular, a child oradolescent might present himself/herselfin an unrealistically favorable light, givenreferral concerns (i.e., that is, nearly allT-scores on the clinical scales are below 50;see Box 6.1).Box 6.1An Example of Fake Good Response SetSelf-report inventories, despite the best effortsof test developers, always remain susceptibleto response sets. In the following case example,the BASC-2-SRP was utilized.Bethany is a 14-year-old girl who wasadmitted to the inpatient psychiatric unit ofa general hospital with a diagnosis of MajorDepression following threats of suicide. Bethany’sparents reportedly divorced two yearsago. She currently lives with her mother andsees her father about once a month. Bethanydoes not have any siblings. Her motherreported that Bethany tends to isolate herselfat home and seems sad and irritable most ofthe time. Bethany is currently in the eighthgrade, and her school attendance during thisyear has been poor. She has been suspendedfrom school on numerous occasions this year,including for getting into fights with peers,refusing to follow teachers’ directions, anddamaging school property. Her grades arereportedly poor, although her mother indicatedthat she used to make mostly “As”and “Bs” until this year. In addition, it wasreported that Bethany has a history of problemsconcentrating at school.Bethany was talkative during the diagnosticinterview, yet it appeared that she was tryingto portray herself in a favorable light, as sheendorsed very few symptoms. When asked abouthobbies, for example, she said that she liked toread. When questioned further, however, shecould not name a book that she had read.According to her mother, Bethany’s familyhistory is significant for depression on bothher maternal and paternal side. Bethany’sfather reportedly had difficulty in school.Bethany’s results show evidence of a socialdesirability response set. On the L-scale of theSRP, Bethany obtained a raw score of 12 whichis in the Extreme Caution range. Furthermore,all but one of her clinical scale scores were lowerthan the normative T-score mean of 50, and allof her adaptive scale scores were above the normativemean of 50. In other words, the SRPresults suggest that Bethany is well-adjustedwhich is inconsistent with her reported currentfunctioning and background information.This example clearly indicates the needto consider all evidence gathered rather thanstrictly relying on the results of any one assessmentstrategy. Her SRP scores were:ScaleClinical ScalesT-ScoreAttitude to School 43Attitude to Teachers 39Attention Problems 44Hyperactivity 40Sensation Seeking 39Atypicality 40Locus of Control 38Somatization 37Social Stress 40Anxiety 32Depression 42Sensation Seeking 51Sense of Inadequacy 42Adaptive ScalesRelations with Parents 51Interpersonal Relations 56Self-Esteem 55Self-Reliance 59

108 CHAPTER 6 SELF-REPORT INVENTORIESNormingThe SRP was normed on a national sampleof 1,500 children aged 8–11 and 1,900 adolescentsaged 15–18. Equal numbers of boysand girls were included within each agegroup (i.e., 8–11; 12–14; 15–18). Based onthe occurrence of age group differences onmany subscales, these age groups were usedin the development of T-scores (i.e., separateT-score distributions were used for 8–11-year olds from 12–14-year olds). Reynoldsand Kamphaus (2004) report that “Withineach sex at each age grouping, the Generalnorm samples were matched to targeted USpopulation estimates taken from the March2001 Current Population Survey (CurrentPopulation Survey, 2001)” (p. 116–117). Thematched variables were socioeconomic status,race/ethnicity variable and geographicregion. The authors also considered thepresence of emotional and behavioral problemsin the General norm sample relative tothe proportion of youth with such problemsin the general population.The SRP offers clinical norms for asample of 577 children and 950 adolescentsselected from an unspecified numberof special education classrooms and mentalhealth settings throughout the UnitedStates. Data presented by the authorsshow good correspondence to the generalUnited States population for the Generalnorm group on the demographic variablesconsidered. In addition, the Clinical normgroup showed variability in race/ethnicity,geographic region, and age for eachcategory (i.e., Learned Disability, ADHD,other; Reynolds & Kamphaus, 2004). Aswith the previous version of the SRP, separategender norms were also devised. Thepurpose of this procedure was to reduceany sex differences in T-scores on the variousscales and composites. Such norms maybe of interest if the clinician wants to considerthe severity of a child’s problems relativeto others of the same sex. However, inmost assessment situations, the question isregarding the degree of problems relativeto the overall population of same-aged children,which would suggest the use of normsthat include both boys and girls (Reynolds& Kamphaus, 2004).ReliabilityThe reliability of the SRP scales is good asindicated by a variety of methods. Medianinternal consistency coefficients are generallyin the .80s (see Table 6.3). Test-retestcoefficients taken between 4 and 8 weekslater are generally in the .70s. As shown inTable 6.3, the lowest internal consistencyTable 6.3 BASC-2-SRP Median InternalConsistency CoefficientsScaleCoefficientAnxiety 0.86Attention Problems 0.78Attitude to School 0.82Attitude to Teachers 0.79Atypicality 0.83Depression 0.86Hyperactivity 0.76Interpersonal Relations 0.79Locus of Control 0.78Relations with Parents 0.87Self-Esteem 0.82Self-Reliance 0.70Sensation Seeking 0.70Sense of Inadequacy 0.79Social Stress 0.83Somatization 0.67Median 0.80Composites School Problems 0.85Inattention/Hyperactivity 0.84Internalizing Problems 0.96Personal Adjustment 0.89Emotional Symptoms Index 0.94

CHAPTER 6 SELF-REPORT INVENTORIES109was for the Somatization scale, whichmay make sense in that this scale consistsof physical symptoms which may not behighly interrelated.ValidityThe BASC-2 manual provides an extensivereport of exploratory and confirmatoryfactor analyses of the SRP items.Three factors were initially found, withAttention Problems and Hyperactivityloading on the Internalizing Problemsfactor; however, a four-factor solutionwith these two scales included as a separatefactor was judged superior. This factorstructure is shown in Table 6.4. Theschool maladjustment factor remainsa relatively unique contribution of theBASC self-report. The personal adjustmentfactor is also unique in that it providesa multidimensional assessment ofTable 6.4 BASC-2-SRP Factors andScale MembersSchool MaladjustmentAttitude to SchoolAttitude to TeachersSensation Seeking (adolescent only)Internalizing ProblemsAnxietyDepressionLocus of ControlSense of InadequacySomatizationSocial StressPersonal AdjustmentRelations with ParentsInterpersonal RelationsSelf-EsteemSelf-RelianceInattention/HyperactivityAttention ProblemsHyperactivityadaptation or potential strengths. Theimplications of this composite for longtermprognosis are unclear, however.The criterion-related validity of theSRP was evaluated by correlating it withthe Achenbach Youth Self-Report, theConners-Wells Adolescent Self-reportScale, Children’s Depression Inventory,Revised Children’s Manifest AnxietyScale. Notably, research on the correlationsbetween the adolescent version ofthe BASC-2 and the MMPI-A has not yetbeen conducted.In general, the SRP scales correlatedhighly (i.e., r = .65 and higher) with analogousscales from the Achenbach YouthSelf-report, whereas the correlations weregenerally moderate between the SRP-A andthe Conners-Wells Adolescent Self-reportScale. Interestingly, the Depression scale ofthe SRP showed only modest correlationswith the Children’s Depression Inventorysubscales and total score (i.e., r = .09–.42) forthe child version of the SRP. However, theadolescent version of the SRP demonstratedmuch more congruence with thecorrelation coefficients ranging from .39to .69. The Anxiety scale of the SRP demonstratedgenerally moderate correlationswith the Revised Children’s Manifest AnxietyScale (i.e., r = .31 to .60 for the SRP-C;r = .33 to .49).The original BASC was the focusof a number of investigations of the correlatesof its scales in varied populations andfor varied purposes (e.g., basic researchquestions on child functioning; treatmentoutcome research; see Kamphaus & Frick,2005 for a review). However, to date, veryfew such investigations have used theBASC-2 system. The BASC-2 manualprovides initial results on the correlatesof the SRP scales; however, much moreresearch is needed, particularly with clinicalpopulations to help provide further understandingof the information garnered fromthese scales.

110 CHAPTER 6 SELF-REPORT INVENTORIESInterpretationAlthough the composites of the SRP havesome factor-analytic evidence to supporttheir validity, there is little evidence as tothe validity and utility of the composites.The composites may very well provide parsimoniousand clinically relevant information,but such studies simply have not beenconducted yet. In the BASC-2 manual, theauthors provide results indicating that theInternalizing Composite and EmotionalSymptoms Index show morderate tostrong correlations with virtually all scalesand composites from the Achenbach YouthSelf-report. The Inattention-Hyperactivitycomposite showed correlations withthe ADHD-oriented scales on the Achenbachand on the Conners-Wells AdolescentSelf-report Scale that were r = .54 andhigher. However, given the relative lack ofresearch to date and the difficulty in interpretingthese composites given their variedcontent (see Table 6.4), until furtherresearch studies are available, initial effortsat SRP interpretation should focus on thescale level. Examples of such interpretationsare provided in Box 6.2. The scales ofthe BASC-2, and by extension, the originalBASC, are better understood because thescale contents have some rational, theoretical,and research basis.Many items from the original BASC-SRP were retained or slightly altered forthe BASC-2-SRP (see Reynolds & Kamphaus,2004), which lends some confidencein scale interpretation for the clinicianwith experience using the previous version.Item-level factor-analytic results provideone empirical clue to the meaning of SRPscales. Some of the items with the mostsubstantial factor loadings on each scaleare identified in Table 6.5. These symptomscan be linked to background informationin order to interpret SRP results withgreater certainty. However, the elevationsof SRP scales may also lead the clinicianto follow-up on issues that might not havebeen raised during the gathering of backgroundinformation.We propose the following steps in interpretingthe SRP (which can be appliedto other omnibus rating scales as well:(1) check validity scales; (2) check criticalitems if available; (3) determine whichscales are elevated; (4) examine the itemsthat appeared to lead to scale elevations.Some caution is needed in keeping afocus on interpretation at the scale level. Inparticular, scales may have intuitive appealand some relevant content, but questionablereliability would hamper the validityof interpretations garnered based onunreliable scales. The reliability estimatesshown in Table 6.3 can help a cliniciangauge the level of confidence in scale-levelinterpretations for the BASC-2-SRP.Strengths and WeaknessesThe BASC-2-SRP is a potentially usefultool in child and adolescent assessment.The BASC-2-SRP has numerousstrengths that make it a viable option foruse with children and adolescents. Notablestrengths include:1. A broader age range than is typicallyavailable for omnibus inventoriesdesigned to obtain self-report fromchildren.2. Unique scales that are relevant to themilieus of children, such as attitudetoward teachers and school and parentchildrelations3. A normative sample that is well-describedand seems appropriately reflective of theUS population as of 2001.4. Good reliability estimates (see Reynolds& Kamphaus, 2004)5. Ease of administration and scoring.6. Items account for heterogeneityof behaviors and symptoms withindomains (McMahon & Frick, 2005).

CHAPTER 6 SELF-REPORT INVENTORIES111Box 6.2A Sample Case Illustrating the Interpretation of the BASC-2-SRPFarrah is a 14-year-old high school freshmanwho was recently discharged from a psychiatricinpatient unit, where she was hospitalizedfor suicidal statements and severe cuttingbehaviors.Farrah presented for the evaluation withdepressed mood and flat affect. She criedseveral times during the evaluation. She complainedabout her difficulty getting alongwith her mother and the fact that she rarelyhears from her father. Farrah reportedly hadbehavioral problems at school since a veryyoung age and tends to interact with peers ofwhom her mother does not approve. She hashad a recent increase in physical complaintsand has missed several days of school this yearbecause of those complaints.Farrah’s BASC-2-SRP results were consistentwith background information: T-scoresand brief descriptions of her results for elevatedscales are belowDepression 80This finding is consistent with observationsduring the evaluation and backgroundinformation suggesting recent suicidality andincreased anhedonia.Sense of Inadequacy 74Farrah expressed a lack of confidence abouther ability to do well in school and succeed inother tasks. Her grades this year have beenmostly “Ds” and “Fs,” and she has a history ofgetting into trouble at school. However, hermeasured intellectual functioning and academicachievement are in the High Averagerange, suggesting that her lack of confidenceregarding schoolwork may be unrealistic.Attention Problems 68Farrah’s reports of difficulty concentrating onthe BASC-2 are consistent with her reportedproblems concentrating in general and herrelative inattentiveness during test.Attitude to School 74Farrah appears to have low self-efficacyregarding school, and she reports that schoolis boring.Attitude to Teacher 71During interview, Farrah described herteachers as uncaring and as having unrealisticexpectations.Locus of Control 76During an interview, she reported that sheis unfairly blamed for things at home and atschool and that her mother’s expectations aretoo high – reports that are consistent with herreports on the BASC-2.Somatization 86Farrah reportedly complains of nausea andheadaches frequently. These complaints appearto be independent of her anti-depressant medicationregimen and also seem to be a strategy forher to avoid going to school.Social Stress 75Farrah reported that she usually feelsuncomfortable around others, includingpeople her own age. She also reported feelinglonely most of the time for the last twoyears.Relations with Parents 28This adaptive scale score is low and fits withFarrah’s long history of parent-child enmity.Self-Esteem 30Farrah views herself as less attractive thanothers and does not see herself as having anyparticular strengths.

112 CHAPTER 6 SELF-REPORT INVENTORIESTable 6.5 SRP Key Symptoms as Indicated by Items with the Highest Factor Loadings perScaleScaleAnxietyAttention ProblemsAttitude to SchoolAttitude to TeachersAtypicalityDepressionHyperactivityInterpersonal RelationsLocus of ControlRelations with ParentsSelf-EsteemSelf-RelianceSensation SeekingSense of InadequacySocial StressSomatizationKey SymptomsNervousness, being bothered by little things, worry, fearTrouble paying attention, getting in trouble for not paying attentionNot caring about school, feelings of wanting to quit school, expressions ofschool hatred, expressions of boredom at schoolFeeling that teachers are unfair, feeling that teacher is not proud of him/her, report that teacher gets mad at him/her for no reasonHearing voices/things that others cannot, feeling like someone is watching,seeing things.Feeling that does nothing right and that life is getting worse, feeling thatnothing goes his/her way.Is told by others to be still, having trouble sitting still, feeling like has tomove aroundFeeling that nobody likes respondent (reverse scored), feeling that othershate to be with him/her (reverse scored), being made fun of by others(reverse scored)Complains of being blamed for things that he or she can’t help or didn’tdo, people get mad at respondent for no reasonParents listen to what respondent says, parents are proud of him/her,parents trust him/herWishes he/she were different (reverse scored), respondent likes the way heor she looksSays that he or she is dependable and is good at making decisionsLikes it when dared to do something, likes to ride in a car going fastFails even when trying hard, wants to do be better but is unable, has difficultykeeping mind on school workFeels left out, feels that others find things wrong with him/her, people findthings wrong with him/herComplains of stomach upsets, nausea, and dizziness7. Availability of a range of derived scoresand norms for general, clinical, andgender-referenced samples.8. Inclusion of validity scales that areintuitive.9. Expanded, user-friendly manual.10. A clear link between teacher and parentrating scales (BASC-2-PRS andBASC-2-TRS) which enhances its usein multi-informant assessment.To date, notable weaknesses of the BASC-2-SRP are:1. Lack of research regarding this versionof the BASC and the improvementsmade from the original BASC-SRP2. Limited criterion-related validity evidence,particularly for the adaptive scales3. Lack of case studies in the manual4. Lack of validity evidence for the validityscales5. An excessive number of Self-Esteem scaleitems that deal with self-perceptions ofpersonal appearance which is also a concernraised about the previous version ofthe SRP (Hoza, 1994)

CHAPTER 6 SELF-REPORT INVENTORIES113Achenbach System ofEmpirically Based AssessmentYouth Self-Report (YSR;Achenbach & Rescorla, 2001)The Youth Self-Report (SRP; Achenbach& Rescorla, 2001) is one component of thelarger set of assessment instruments offeredwithin the Achenbach System of EmpiricallyBased Assessment that includes a parentrating scale, a teacher rating scale, anobservation scale, and other measures. TheYSR is designed for ages 11 through 18,and obtains adolescents’ reports about theirown competencies and problems in a formatsimilar to that of the parent CBCL andteacher TRF, discussed in the next chapter.As with their predecessors, the most recentmeasures in the Achenbach system haveconsiderable content overlap, which can beboth an advantage and disadvantage of thissystem, depending on what the clinician isseeking from behavioral rating scales.An additional form covers the ages of18–30 entitled the Young Adult Self-Report(YASR; Achenbach, 1997). The YASRcontent is similar to that of the YSR withnoteworthy differences such as scales forsubstance use and some unique adaptivefunctioning scales (e.g., Friends, Education,Job, etc.). In light of the developmentalperiods of focus in this text, the YASR willnot be discussed in detail here, but an introductionmay be found at http://www.aseba.org/products/yasr.html.Scale ContentComposite scores reflecting externalizingand internalizing dimensions and a totalcomposite are offered. The following clinicalscales contribute to these composites.Withdrawn/Depressed- preferring to bealone, shy, sulks, sad, lacking energy, etc.Somatic Complaints- nausea, headaches,dizziness, etc.Anxious/Depressed- crying, fears, nervous,suicidal ideation, etc.Rule-Breaking Behavior- lying, substanceabuse, truancy, stealing, etc.Aggressive Behavior- teasing others,arguing, fighting, destruction of property,etc.Included in the Total Composite but notthe Internalizing or Externalizing compositesare the following:Social Problems- jealous of others,teased by others, clumsy, etc.Thought Problems- strange behaviors,hoarding objects, sleeping less, hallucinatoryexperiences, etc.Attention Problems- failing to finishassignments, immature, impulsivity, daydreaming,etc.Scales referred to as Social Competenceare also included that assess participationin a variety of activities (e.g., sports) andsocial interactions (e.g., friendships).The clinical scales are empiricallyderivedvia factor analysis, and the competencescales are rationally derived.Critical items (e.g., harming self, settingfires, etc.) are also available on this versionof the YSR. In addition, six DSM-Oriented scales are available for theYSR. These scales were formed based onpsychiatrists’ impressions of items (seeAchenbach, Dumenci, & Rescorla, 2001)that theoretically map on to the DSMrelateddomains being assessed (i.e.,Affective Problems, Anxiety Problems,Somatic Problems, Attention/HyperactivityProblems, Oppositional DefiantProblems, Conduct Problems). Researchhas supported the structure of the syndromescales of the YSR across over 20cultures (Ivanova et al., 2007).Administration and ScoringThe YSR is designed to be self-administeredand requires approximately 15–20min (Achenbach & Rescorla, 2001). TheYSR uses a three choice response format:“Not True, Somewhat/Sometimes True,and Very True or Often True.” Some items

114 CHAPTER 6 SELF-REPORT INVENTORIESalso provide space for additional informationsuch as, for example, “Describe:” following“I store up too many things I don’tneed.”Templates are used for hand-scoring,and computer scoring is also available. Anintegrative computer program can be usedto compare results for several raters (e.g.,a parent, two teachers, and the YSR), andthis option is discussed in great detail byAchenbach and Rescorla (2001). The levelof comparability facilitates the study ofinter-rater agreement in clinical or researchsettings and aids in clinical interpretationof converging evidence and discrepanciesin reports of the child’s functioning.the range of low scores on the clinical scalesand high scores on the competence scales.For example, T-scores for the clinical scaleswere not allowed to be articulated below avalue of 50. The transformation to reduceskewness and truncated score range bothserve to make the T-score distribution forthe YSR different from original sample results.The intent of this approach is to aid in interpretationof strengths and difficulties acrossdomains. However, this lack of reflectionof sample characteristics in the T-scoresmakes them of dubious value for researchpurposes in particular. For most researchquestions, raw scores would likely be moreappropriate than normalized T-scores.NormingThe design of the YSR norming sampleattempted to mimic the national populationof school children for ages 11 through18 in terms of SES, geographic region,and ethnicity. The sample included 1,057youth, 52% of whom were boys. Childrenwho had reports of mental health, substanceabuse, or special education serviceswere excluded, thus making this a “normal”sample rather than a “normative sample.Most participants (i.e., 53%) were froma middle SES background, whereas 16%were from lower SES homes. The samplewas 60% White, 20% African American,8% Latino, and 11% Mixed or Other. Fromthese statistics, both African Americans andindividuals identifying as Latino(a) appearto be underrepresented (Current PopulationSurvey, 2001). Approximately 40% ofparticipants were from the southern partof the United States with the Northeast,Midwest, West each being represented byapproximately 20% of the sample participants(see Achenbach & Rescorla, 2001).The derived T-scores for the YSR arenormalized, which results in changing theshape of the raw score distribution (i.e., reducingskewness). Furthermore, the T-scoredistributions are truncated, which limitsReliabilityInternal consistency estimates are reasonablefor the clinical scales falling between.71 and .89. The internal consistencies ofthe composites are all above .90. Internalconsistency estimates are somewhat lowerfor the competence scales (see Achenbach& Rescorla, 2001).Short-term (i.e., approximately oneweekinterval) test-retest coefficients aregenerally good, with only the Withdrawn/Depressed scale having a coefficient below.70. Seven-month test-retest coefficientswere adequate with coefficients generallyin the .50 range. Coefficients for the Withdrawn/Depressedscale and Somatic Complaintsscale were somewhat lower.ValidityThe YSR manual does not report evidenceof criterion-related validity, particularly inregards to the correspondence between theYSR and other measures of emotional andbehavioral functioning.Some differential validity data arepresented, with the scales of the YSRconsistently differentiating between clinicreferredand non-referred youth. Exceptionsto this differential validity were the

CHAPTER 6 SELF-REPORT INVENTORIES115DSM-Oriented Anxiety Problems andSomatic Problems scales. The ability of theYSR to differentiate among clinical groupsis not addressed. Achenbach and Rescorla(2001) indicate that differential validity isthe driving force behind content selectionfor the current YSR and its predecessors.The most recent YSR has six items thatdiffer from the items in the previous YSR.Generally speaking, the validity evidencereported in the manual concerning the YSRis minimal. However, the previous versionof the YSR enjoys a great deal of validityevidence from independent researchers. Inaddition, in light of the close item correspondencebetween the two versions, onecan surmise that support for the validity ofthe earlier YSR can be taken as providingsome support for the current YSR, particularlythe problem scales.A study by Thurber and Hollingsworth(1992) compared YSR results withthe results of several other measures (e.g.,California Personality Inventory and BeckDepression Inventory) in a factor-analyticinvestigation. The sample for this studyincluded 102 adolescent inpatients. Supportfor the existence of the internalizingand externalizing dimensions was found,as these factors converged with measuresof similar constructs to form recognizablefactors. Of interest was an additionalfinding that the Externalizing Scale maybe affected by a tendency to respond in asocially desirable way and deny problems.The Internalizing Scale also showed somesensitivity to response sets in that it wasaffected somewhat by minimizing symptoms(Thurber & Hollingsworth, 1992).Brown (1999) likewise found that “highrisk”adolescents tended to underreportbehavior problems when school recordsand police reports were used as externalcriteria. Adolescent reports tended to agreewith other reports for “more positivelyoriented items.” These findings shouldbe taken into account when interpretingself-report results and should be combinedwith corroborating evidence in drawingconclusions.In contrast, Sourander, Helstelae, andHelenius (1999) found that Finnish adolescentsreported significantly more problemsthan their parents, and girls reported moredistress, especially internalizing problems,than boys. These authors concluded thatmany adolescents may not be receivingappropriate mental health services becausetheir problems go unrecognized by theirparents.A criterion-related validity study byHandwerk, Friman, and Larzelere (2000)compared the YSR to the NIMH DiagnosticInterview Schedule for Children(DISC). They compared DISC and YSRresults to behavior in a treatment programand, generally speaking, found no differencesbetween diagnostic groups formedby using either instrument. Similar resultshave also been found for the YSR in comparisonto the DISC Version 2.1 (Morgan& Cauce, 1999).More validity studies exist for variouscultural groups on the previous version ofthe YSR. Reliability and factorial validity ofthe YSR have been found to be comparableto North American findings in the Netherlands(de Groot, Koot, & Verhulst, 1996),Switzerland (Steinhausen & Metzke, 1998),Japan (Kuramoto et al. 1999), and Spain(Abad, Forns, Amador, & Martorell, 2000).Research has led to the conclusion, giventhat sex differences appear very consistentlyon the YSR, that sex is a more importantconsideration in predicting psychopathologythan demographic factors such as age ornationality (Steinhausen & Metzke, 1998).However, unlike the teacher and parentreport measures of the Achenbach system,the YSR does not include gender-specificnorms.A predictive validity study of the previousYSR was conducted in Finland, where121 adolescents were administered theYSR at age 14 or 15 and followed up toages 20 and 21 (Aronen, Teerikangas, &

116 CHAPTER 6 SELF-REPORT INVENTORIESKurkela, 1999). YSR problems were goodpredictors of adult symptomatology. Ofequal interest are the findings that internalizingsymptoms were better predictorsand that self-report was more predictivethan parent report. Results such as thesespecifically support the practical utility ofthe YSR, and they suggest that self-reporttechnology should be used with all childrenand adolescents with adequate readingcomprehension.InterpretationThe Achenbach manual provides some casestudies, yet it does not provide interpretiveguidance. This omission is remediedto some extent by the existence of otherarticles on the YSR and by its amenabilityto general interpretive approaches such asthose described earlier in this chapter.A strength of the Achenbach approachto scale construction is the ease withwhich interpretations can be made acrossinformants because of the close item correspondenceon the different forms. Asmentioned above, Achenbach and Rescorla(2001) detail the approach by which statisticalcomparisons across informants may bemade. To the extent that informants agreeon the presence of a problem, the clinicianmay be more confident that a problem warrantingattention exists.However, it is possible to be too heavilyinfluenced by indexes of agreement inthe interpretive process. For example, onemight require agreement across raters tomake diagnostic or other decisions. Weprefer to consider each rater as a valuablesource of information that may bediagnostically or otherwise valuable forcase conceptualization in its own right,when combined with other information.To illustrate, if a clinician requires parentalagreement for a self-report finding, achild or adolescent may be denied neededservices (Sourander et al., 1999). Youthmay be rich and valid sources of informationabout their emotional and behavioralfunctioning (Aronen et al., 1999; Barryet al., in press).Convergence across informants hasmerits for answering a referral questionand making recommendations. However,we also hold to the philosophy that differentraters make unique contributions tothe understanding of a child’s referral difficulties(see Chap. 15) and that disagreementamong informants is not necessarilyindicative of a measurement problem.Strengths and WeaknessesThe YSR has several strengths:1. Brief administration time2. A large research base on its closelyrelatedpredecessor3. Research conducted with individualsrepresenting many cultures and normsfrom a number of cultural groups4. A large base of experienced users5. Considerable item overlap with its parentand teacher report counterparts, whichaids in cross-informant interpretation6. A helpful Web site with information foradministration, purchasing, interpretation,and user discussion is available athttp://www.aseba.org/products/ysr.htmlWeaknesses of the YSR include:1. Little assessment of school-relatedproblems2. Limited assessment of adaptive competencies3. The absence of validity scales4. Limited construct validity evidence todate for the current YSR.5. Direct comparisons of norm sampledemographics to US population are notprovided; however, African Americansand Latino(a)s appear to be under-represented.

CHAPTER 6 SELF-REPORT INVENTORIES117Minnesota MultiphasicPersonality Inventory-Adolescent (MMPI-A; Butcheret al., 1992)The Minnesota Multiphasic PersonalityInventory-Adolescent (MMPI-A; Butcheret al.) has strong roots in the original workof Hathaway and McKinley (1940), andthe authors of MMPI-A tried to maintaincontinuity with the original work (Butcheret al.). This objective was achieved, as all ofthe clinical and validity scales were retained.In addition, the MMPI-A includes contentscales (discussed below) which may be usefulfor addressing many clinical questionsof interest. The MMPI-A is designed foryouth aged 14–18.Scale ContentKnowledge of the history, rationale, psychometricproperties, and item content ofthe MMPI-A clinical scales is importantfor proper interpretation. Consequently,each of the featured clinical scales is discussedin turn.Scale 1, Hs: Hypochondriasis. Items forthis scale were originally developed toidentify respondents with a history ofsymptomatology characteristic of hypochondriasis(Butcher et al., 1992). Items ofthis scale assess topics such as nausea, vomiting,upset stomach, sleep problems, chestpain, numbness, muscle twitching, bodilytenderness, dizziness, weakness, and lack ofgeneral feeling of wellness.Scale 2, D: Depression. Hathaway andMcKinley (1942) described this measureas an index of general dissatisfaction withone’s life, including feelings of discouragement,hopelessness, and low morale. Itemcontent includes appetite changes, healthworries, anhedonia, work problems, tension,constipation, increased swearing,concentration problems, sleep problems,withdrawal, teasing animals, low self-confidence,low self-esteem, worry at bedtime,crying easily, decreased reading comprehension,weight change, and impairedmemory, among other things. The itemcontent of this scale is diverse, reachingfar beyond diagnostic criteria for depressionsuch as those included in the DSMand likely contributes to mediocre internalconsistency reliability coefficients.Scale 3, Hy: Hysteria. According toButcher et al. (1992), this scale consists of60 items that were originally selected toidentify individuals who respond to stresswith hysterical reactions that include sensoryor motor disorders without an organicbasis. Some of the item content includespoor appetite, fatigue, cold extremities,decreased work productivity, nausea andvomiting, urges to curse, poor concentration,disturbed sleep, health concerns,chest pain, unhappiness, difficulty persuadingothers, muscle twitching, irritability,worry about contracting diseases, dysfunctionalrelationships with family members,concern about others’ opinion, and dislikeof school. Like D (scale 2), this item poolis very diverse, resulting in mediocre internalconsistency coefficients. It should alsobe noted that the first three scales of theMMPI-A share considerable item overlap.Scale 4, Pd: Psychopathic Deviate. Thisscale was originally constructed based onthe responses of individuals with historiesof lying, stealing, sexual promiscuity, andalcohol abuse (Butcher et al., 1992). Furthermore,high scores on this scale areassociated with family, legal, and schooldifficulties (Butcher et al.). Item contentincludes loss of interest in daily activities,a desire to leave home, feeling misunderstood,feeling used, poor concentration,having unusual experiences, history oftrouble because of sexual behavior, historyof stealing, unhappiness, disapprovalby family members, winning arguments,inability to tolerate ridicule, regrettingactions, admissions of misbehavior, historyof school disciplinary action, feeling likesomeone is out to get him/her, and weight

118 CHAPTER 6 SELF-REPORT INVENTORIESchanges. For the most part, this is the firstscale with item content that is distinctivefrom the first three.Scale 5, Mf Masculinity-Femininity. Thisscale was originally developed on a sampleof adult men, described by Hathaway(1956) as “male sexual inverts” (Butcheret al., 1992). Presumably, males with highscores are more feminine, and women withclinically significant scores are thought tohave more masculine interests. Item contentincludes topics such as lack of interestin mechanics magazines, reluctance toincriminate oneself, a desire to be of theopposite gender, interest in love storiesand poetry, sensitive feelings, lack of interestin forest ranger work, being a soldieror hunter, expressing the need to argue tomake a point, attending few parties, dislikefor wagering, interest in gardening andcooking, maintaining a diary, fear of snakes,worry, and talks about sex. This scale is, tosay the least, unique, and, by most currentunderstanding, out of step with the times.Therefore, it is not particularly useful forclinical interpretation.Scale 6, Pa: Paranoia. This scale isdesigned to assess paranoid symptomatology.The scale includes item contentsuch as feelings of persecution, havingevil thoughts, feeling misunderstood,emotional lability, feeling possessed byevil spirits, unhappiness, sensation seeking,distrust of others, crying easily, feelingas though one is being followed orpoisoned by someone, ideas of reference,and history of legal trouble.Scale 7, Pt: Psychasthenia. This scaleassesses anxiety, particularly a tendencyto worry obsessively. The item content ofScale 7 includes health worries, loss of interestin activities, having shameful thoughts,emotional lability, poor concentration,fatigue, unhappiness, low self-esteem,feelings of regret, guilt, impaired readingcomprehension, impaired memory, worry,restlessness, excitability, fear of speaking infront of others, being easily embarrassed,impatience, counting unimportant things,rumination, and overreaction to failure.This scale has many items in common withscales 2, 3, and 8.Scale 8, Sc: Schizophrenia. This scale wasdesigned to identify patients with diagnosesof various forms of psychosis (Butcheret al., 1992). Scale 8 items sample content suchas lack of interest in daily activities, havingunwanted thoughts, desire to leave home,poor concentration, bizarre experiences,stealing, feelings of persecution, avoidanceof others, day-dreaming, muscle twitching,urges to do something socially unacceptable,changes in speech pattern, decreasedreading comprehension, impaired memory,blackout spells, fear of losing control,impaired balance, restlessness, difficultyinitiating activity, excitability, numbness,de-creased taste sensitivity, sexual preoccupations,impaired relationships with parentsand other family members, loneliness,lack of intimacy, impatience, and feelingsof unreality. This scale shares many itemswith scales 2, 3, and 7. It is a long scale with77 items, making item overlap with otherscales a central characteristic. Its high correlationswith other scales are discussed ina later section. Some of the items that differentiatethis scale from others have to dowith impaired social relationships and poorreality contact.Scale 9, Ma: Hypomania. This scaleassesses a tendency toward excitability andincludes items assessing tension, desire toleave home, crying spells, urges to do somethingsocially unacceptable, indecision,sensation seeking, racing thoughts, feelingsof persecution, lack of fear of heights,blackout spells, occasional ability to makedecisions very easily, self-righteousness,restlessness, satisfaction with personalappearance, sweatiness, excitability, excessivethirst, and admiration for clevernesseven if it is criminal.Scale 10, Si: Social Introversion. Si scoreelevations are produced by content suchas failure to face crises or problems, poor

CHAPTER 6 SELF-REPORT INVENTORIES119concentration, poor sociability, unhappiness,fear of ridicule, lack of interest inparties, easily losing arguments, low sensationseeking, change in speech pattern,distrust of others, indecision, shyness, difficultywith small talk, brooding, concernsabout personal appearance, embarrassmentin front of groups, failure to initiate conversation,difficulty making friends, loneliness,envy of others’ successes, and lowself-esteem.Content scales are a relatively uniquefeature of the MMPI that were developeddifferently from the original clinical scales.Whereas empirical approaches, includingempirical criterion keying, were usedfor the development of the original scales,content scales depend more on a rational/theoretical approach to test development(Williams et al., 1992) in which scale contentconsiders both empirical factor loadingsand homogeneity of content withinscales.The first step in the content scale developmentprocess was to select 22 contentcategories based on a review of the adultexperimental version (the predecessor tothe MMPI-2). In the second step, a total ofthree raters assigned items from the adultexperimental form to the 22 categories(Williams et al., 1992). A group consensuswas reached on the assignments, and someitems were discarded. A total of 21 contentscales remained after this step. In the nextstep, correlations and reliability indiceswere used to enhance the reliability andhomogeneity of each scale. The fourthstage involved another “rational review”of the items in response to the aforementionedstatistical data. Some scales wererenamed and some dropped at this stage.In the fifth and final step, items that correlatedhigher with a scale of which theywere not a member were removed. Theresult was 15 MMPI-2 content scales forthe adult measure (Williams et al., 1992).These same procedures were applied tothe development of the MMPI-A contentscales, with the MMPI-2 content scalesserving as the foundation. Items wereadded and removed, and some new scaleswere developed (Williams et al., 1992).This step resulted in the retention of themajority of the MMPI-2 content scales foradolescents and the addition of three scales-Alienation, Low Aspirations, and SchoolProblems. Descriptions of the MMPI-Acontent scales are shown in Table 6.6.There are also six supplementary scaleson the MMPI-A (Butcher et al., 1992):Anxiety (A): distress, discomfort, conformity,being upset by social situations.Repression (R): tendency toward submissivenessand conventionality, avoidanceof conflictMAC-R, MacAndrew AlcoholismScale-Revised: substance abuse problems,willingness to take risks, extraversionAlcohol/Drug Problem Acknowledgement(ACK): items that directly refer to drugand alcohol useAlcohol/Drug Problem Proneness(PRO): stimulus seeking, negative peergroup influence, rule-breaking, negativeattitudes toward achievementImmaturity (IMM): orientation to thepresent instead of future, lack of insight,hostility, self-centerednessAdministration and ScoringThe MMPI-A is unusually long (i.e., 478items) compared to other self-reportinventories designed for children and adolescents,which calls for special administrationguidelines. In total, the MMPI-Atakes approximately 90 min to administer,and some adolescents may have to takethe test in more than one session (Butcheret al., 1992). Furthermore, because manyadolescents require supervision duringthe administration of these scales, considerablymore examiner time may berequired. Substantial administration timesavings, however, can be gained by using

120 CHAPTER 6 SELF-REPORT INVENTORIESTable 6.6 MMPI-A Content ScalesScaleAnxiety (A-anx)Obsessiveness (A-obs)Depression (A-dep)Health Concerns (A-hea)Alienation (A-aln)Bizarre Mentation (A-biz)Anger (A-ang)Cynicism (A-cyn)Conduct Problems (A-con)Low self-esteem (A-lse)Low Aspirations (A-las)Social Discomfort (A-sod)Family Problems (A-fam)School Problems (A-sch)Negative Treatment Indicators (A-trt)DescriptionIncludes excessive worry, problems sleeping, problemsconcentrating, tensionUnreasonable worry, rumination, difficulty making decisions,reports that others are impatient with them, regretIncludes frequent crying, fatigue, self-deprecatingthoughts, hopelessnessPhysical complaints including nausea, dizziness, constipation,difficulty hearing, headachesFeeling disliked and misunderstood by others, feeling thatothers are out to get them, preferring to be aloneStrange thoughts and experiences, hallucinations, paranoiaStarting fights, cursing, destroying things, irritability,impatience with othersMistrust of others, feeling that others are unfair, feelingthatothers are jealousStealing, lying, disobeying rules, shoplifting, being disrespectfultoward othersFeeling unattractive, lacking self-confidence, feelings ofuselessnessDislike of studying and reading, giving up quickly, difficultystarting tasksShyness, avoidance of others, dislike of crowds or socialgatheringsFamily discord, feeling that one cannot depend on familymembers, jealousy, limited family communicationPoor grades, negative attitudes toward teachers, suspensions,truancy, belief that school is a waste of time.Negative attitudes toward doctors and mental healthprofessionals, feeling that faults and bad habits cannot beovercome, unwillingness to face problemsFrom Butcher et al. (1992).a computerized adaptive administrationformat (Forhey, Handel, & Ben-Porath,2000). In fact, time savings of 50–123items may be possible, with research suggestingno significant differences in averagescale scores or in the distributions ofscale scores (Hays & McCallum, 2005).Checks on the adolescent’s readingcomprehension level are also required.Readability analyses of individual itemsshow readability at approximately the fifthto sixth grade level in most cases. However,when in doubt, an examiner may ask thechild to read some items aloud to get somesense of the child’s ability to comprehendthe item content. The validity checks provideanother useful alert to possible readabilityproblems.

CHAPTER 6 SELF-REPORT INVENTORIES121Validity ScalesThe MMPI series has a long tradition ofthe use of validity indexes which is reflectedin the adolescent version. Brief descriptionsof the validity scales follow (Butcheret al., 1992):Cannot Say (?). This scale is comprised ofthe total number of items that the respondenteither failed to answer or endorsedas both true and false. If there are a largenumber of items fitting this description,the clinician should attempt to ascertainthe reason (e.g., carelessness, discomfort,difficulty with comprehension, defiance).LIE (L). This scale is intended to detectnaive attempts by adolescents to put themselvesin an overly favorable light.F, Fl, and F2 (Infrequency). The F scaleis the antithesis of the L scale in that itassesses the tendency of individuals toplace themselves in an unfavorable light,or “fake bad.” Items were selected for thisscale if they were endorsed in their deviantdirection by less than 20% of the normativesample.K (Defensiveness). According to Butcheret al. (1992), “This scale was designedoriginally to identify adults in psychiatricsettings who displayed significant degreesof psychopathology, but produced profilesthat were within normal limits” (Meehl& Hathaway, 1946, p. 40). Butcher et al.(1992), however, suggest that an MMPI-Aprofile should not be invalidated solely onthe basis of an elevated K score, particularlyif used with individuals who are notin a restrictive mental health or psychiatricsetting.VRIN (Variable Response Inconsistency).The VRN scale consists of pairs of itemsthat have either similar or opposing itemcontent. The score yielded by the VRINscale reflects the number of item pairsanswered inconsistently. A high score mayreveal a careless response style on the partof the client.TRIN (True Response Inconsistency). Thisscale is analogous to the VRIN scale inthat it is made up of pairs of items. It differsin that the TRIN scale consists solelyof items with opposite content. An elevatedscore may reveal an acquiescence responseset, or the tendency for the test subject toindiscriminately answer True to the items.Conversely, a low TRIN score may revealnon-acquiescence.Some validity evidence exists to supportthe use of MMPI-A validity scales; astrength of these scales relative to similarscales from other measures. The L and Kscales have been shown to be reasonablygood at assessing symptom underreporting(Baer, Ballenger, & Kroll, 1998; Stein& Graham, 2005). Validity scale cut scoreshad to be lowered somewhat to detecta fake good response set when evaluatingadolescents in a correctional facility(Stein & Graham, 1999). The F, Fl, F2,and VRIN scales were best for assessingrandom responding for a sample of 354adolescents (Archer & Elkins, 1999).NormingThe MMPI-A was normed in eight statesin the continental United States on 1,620adolescents. One state, however (Washington),contributed only 14 cases to thenorming.The distribution of the sample byvariables such as gender, age, grade, andparental education and occupation aregiven in the manual. These variables werenot, however, used as stratification variablesin order to match US Census orother criteria as is common for clinical testinstruments. The Hispanic population, forexample, is clearly under-sampled, constitutingonly 2.2% of the female sample and2.0% of the male sample, which is smallerthan the sample of Native American children.However, a great deal of subsequentresearch on the MMPI-A has been conductedwith Hispanic individuals, increasingthe confidence one can have in usingthis instrument with Hispanic clients (see

122 CHAPTER 6 SELF-REPORT INVENTORIESButcher, Cabiya, Lucio, & Garrido, 2007).Similarly, the SES distribution may beskewed toward higher levels of SES thanthe national population. The authors notedthat, “This rough classification of occupationssuggests that mothers and fathers aredescribed by many children as having professionaland managerial occupations, whilerelatively low percentages are recorded forthe homemaker and unskilled” (Butcheret al., 1992, p. 13). The age distribution ofthe sample is also highly variable. At age18, only 42 male cases and 45 female caseswere collected.The small sample at age 18 may contributeto flawed estimates of psychopathology.A study by Shaevel and Archer(1996) revealed that 18-year olds scoredsubstantially differently on the MMPI-Aand MMPI-2. More evidence of pathologywas obtained on the MMPI-2 with correspondinglylower validity scale values.Differences in T-scores between the twoinstruments were sometimes as high as 15points for the same scales.At the opposite end of the age range, onestudy of an inpatient population of 13-year oldsfound little difference in scores in comparisonto a matched group of 14-year olds (Janus, deGroot, & Toepfer, 1998). Another investigationfound little effect of demographic variableson MMPI-A T-scores (Schinka, Elkins,& Archer, 1998). However, Archer (2005) discussedthe tendency across several samples forsymptom endorsement on the MMPI-A to beinversely correlated with age.The norm sample also included 193individuals who had received mental healthservices, leading to a relatively large proportionof adolescents who do not appearelevated on the MMPI-A (Archer, 2005).However, removing these individuals andrecalculating norms does not appear tochange the pattern of results a great deal(Hand, Archer, Handel, & Forbey, 2007).The US clinical normative sample consistedof 420 boys and 293 girls. All of theclinical cases were taken from the Minneapolisarea (Butcher et al., 1992). Furtherdetails regarding the clinical sample can befound in Williams et al. (1992). The majorityof cases (i.e., 71% of the boys and 56%of the girls) were undergoing treatment inalcohol/drug units (Williams et al., 1992),suggesting that the clinical sample couldbe reconceptualzied to more accuratelyreflect the preponderance of substanceabuse cases.ReliabilityThere are distinct scale differences inthe internal consistency estimates for theMMPI-A (see Table 6.7). Some of the clinicalscales (e.g., Hs, Pt, Si) have respectableestimates. In direct contrast, some of thescales have internal consistency estimatesthat raise questions about their content.The desirability of including a scale thatpossesses more error than reliable varianceis not clear. The Mf coefficients of.43 (boys) and .40 (girls) are the worst ofthose reported. The Pa, D, and Ma scalesare also less reliable than most of the scalesdescribed in this chapter.Internal consistency estimates for thevalidity scales range from unacceptably lowto impressively high, with most being moderate(.70s and .80s). According to Butcheret al. (1992), the lowest coefficients wereobtained for the L scale, where coefficientsranged from .53 in the female clinical sampleto .64 in the male normative sample. Incontrast, the F scale produced coefficientsranging from .81 (female clinical sample)to .90 (male normative sample).The internal consistency estimates forthe “content” scales of the MMPI-A aregenerally better than those for the originalclinical scales (see Table 6.7). The A-depscale coefficients are considerably betterthan those of the original D scale, rangingfrom a low of .80 for the normative sampleof boys to a high of .89 for the clinical sampleof girls (see Butcher et al., 1992).The lowest internal consistencies of thecontent scales are produced by the A-las

CHAPTER 6 SELF-REPORT INVENTORIES123Table 6.7 MMPI-A Median Internal Consistency Reliability EstimatesClinical Scale Boys (N = 805) Girls (N = 815) Content Scale Girls and BoysScale 1, Hs .78 .79 A-anx .80Scale 2, D .65 .66 A-ohs .74Scale 3, Hy .63 .55 A-dep .83Scale 4, Pd .63 .68 A-hea .82Scale 5, Mf .43 .40 A-aln .74Scale 6, Pa .57 .59 A-biz .75Scale 7, Pt .84 .86 A-ang .72Scale 8, Sc .88 .89 A-cyn .80Scale 9, Ma .61 .61 A-con .73Scale 10, Si .79 .80 A-lse .74A-las .61A-sod .78A-fam .82A-sch .70A-trt .76A .89R .53MAC-R .48ACK .66PRO .69IMM .82Note: From Butcher et al. (1992).scale, which has coefficients ranging from.55 to .66. These coefficients, however, arebetter than those of the MF clinical scale.Some of the supplementary scales arealso plagued by poor reliability estimates.The revised MacAndrew (MAC-R) scaleyields a median coefficient of .48, whichis, again, lower than most of the scalescited in this chapter. This lack of reliabilityalso makes the MAC-R scale difficultto validate because reliability is a necessarycondition for validity. The MMPI-A manualcautions that a cut-off raw score of 28 onthe MAC-R may result in false positives;the existence of such poor reliability estimatesmakes one question the reliability ofany cut score or, for that matter, the inclusionof the scale. Reliability coefficients inthe .40s are typically not seen as adequatefor clinical decision making.If one orders all of the MMPI-A clinicaland content scales by their reliabilityestimates, some implications for interpretationbecome clear. Scales can be groupedby reliability coefficients with guidancefor interpretation as shown:This reliability-based interpretivehierarchy is, of course, overly simplisticbecause the validity of these scales is notequivalent for all purposes. The hierarchy,however, is useful in that there is a relationbetween reliability and validity. The fourscales with median coefficients below .60are less likely to be the beneficiaries of substantialvalidity evidence, as will be notedin the next section.

124 CHAPTER 6 SELF-REPORT INVENTORIESGo o d Re l i a b i l i t y (median coefficient .80)Scale 7, PtScale 8, ScA-anxA-depA-cynA-heaA-famAAd e q u at e Re l i a b i l i t y(median coefficient = .70 to .79)Scale 1, HsScale 10, SiA-obsA-alnA-bizA-angA-conA-lseA-sodA-schA-trtPo o r Re l i a b i l i t y (median coefficient = .60to .69)Scale 2, DScale 3, HyScale 4, PdScale 9, MaA-1asACK PROIn a d e q u at e Re l i a b i l i t y (median coefficient.59)Scale 5, MfScale 6, PaRMAC-RFor the MMPI-A, the preceding chartssuggest that the clinician could have moreconfidence in the information obtainedfrom the content scales than the clinicalscales. The higher internal consistency reliabilityof the content scales may very wellbe an artifact of the sometimes substantialitem overlap in the clinical scales. In lightof this issue, for clinical scale elevations,caution must be taken to determine whatsorts of symptoms led to the elevations.As is typical for such scales, test-retestcoefficients differ from internal consistencyestimates. Test-retest coefficientsare somewhat more difficult to interpret,however, because it is unclear whether ornot some scales measure traits that theoreticallyshould be stable over at least shortperiods of time. Regardless, test-retest datacan be of value when gauging changes fromone evaluation to another.One scenario might involve an adolescentwho was hospitalized with paranoidideation that was reflected by a highT-score (78) on the Pa scale. It is conceivablethat this individual would obtain alower score of 61 on re-test prior to dischargetwo weeks after the initial assessment.One interpretation of these results isthat treatment has been effective. Anotherinterpretation is that Pa scale results arerelatively unstable (r = .65) and that theT-score of 78 was spuriously high or the61 was erroneously low. These test-retestdata do provide an alternative hypothesisfor this score difference that, in this case,may have implications for discharge planning.In such a scenario, when MMPI-Aresults may not be well-corroborated byother clinical findings, more careful outpatientfollow-up may be warranted to ensurethat paraniod ideation has abated significantlyenough so as to not adversely affectfunctioning in school or other settings.Overall, the reliability estimates for theMMPI-A are more variable than might beexpected. Such variability requires a morediscerning user who evaluates the reliability

CHAPTER 6 SELF-REPORT INVENTORIES125of results on a scale-by-scale basis, whichwould not be necessary with more uniformreliability coefficients. It is also noteworthythat the new content scales appear to bemore reliable on average than the originalclinical scales. The user may more confidentlyassume that the content scales possessadequate reliability.ValidityAn important fact to keep in mind wheninterpreting the factor structure of theMMPI-A is the extent of item overlap(Archer, Belevich, & Elkins, 1994). Theclinical scales were designed with manyoverlapping items that serve to strengthenthe correlation between the scales. Item31, for example, is included on six scales(2, 3, 4, 7, 8, 10). An analogous situationwould be to have some WISC-IV itemsincluded on several subtests or composites.It is difficult to imagine, but what if severalBlock Design items were allowed, becauseof their correlations with the Verbal Comprehensionsubtests, to be included in calculationsof the Verbal ComprehensionIndex? Such a move would probably begreeted by skepticism, causing cliniciansto wonder about the distinction betweenthe Verbal Comprehension and PerceptualReasoning Indexes.Analogously, scales 7 (Pt) and 8 (Sc)correlate highly with one another at .85for females and .83 for males (Butcheret al., 1992). As might be expected, thesescales both “load” highly on the first factor.However, these scales have 17 items incommon, which parsimoniously explainsthe similar factor loadings of these scales.This validity evidence is potentially importantin that it warns against routinely interpretingthese scales as measures of distinctconstructs, traits, or symptom clusters.An early factor analysis of the MMPI-A revealed four factors: general anxiety,overcontrol or repression, the Si (thirdfactor) and Mf (fourth factor) scales(Butcher et al., 1992). This factor solutionreported in the manual is highly similarfor both males and females.The general anxiety factor accountsfor the vast majority of the variance in thecorrelation matrix. Factor 1 is marked byloadings for the Hs, D, Hy, Pd, Pa, Pt, andSc scales. Based on these results, this factorlooks like a measure of general distress.The second (overcontrol) factor is identifiedby loadings for L and K for males andMa for females. The Si factor is clear-cut forboth genders, whereas the Mf factor is clearcutfor males only.Another factor analysis of the same normativesample produced somewhat differentfindings, with 14 factors being yieldedfrom an exploratory factor analysis at theitem level and 8 factors when conductedat the scale level (Archer et al., 1994). Afactor analytic study of the MMPI-A contentscales, based on normative samples,suggested that the 15 content scales couldrepresent two latent variables-“generalmaladjustment” and “externalizing tendencies”(McCarthy & Archer, 1998). Asingle factor solution, on the other hand,may be more appropriate for girls.Other Validity StudiesThere is a wealth of external validity evidencefor the MMPI-A clinical and contentscales. The reader will note, however, thatthis evidence re-quires cross-validation;therefore, much of it is difficult to interpret.For example, the A-dep (Depressioncontent scale) was correlated with severalcriteria/variables for the normative andclinical samples. Correlations with thesecriteria ranged from a low (consideringthe absolute magnitude of the correlation)of −.18 with grades in school and outstandingpersonal achievement to a highof .24 for increase in disagreements withparent(s). The correlations between a highA-dep score and suicidal ideation/gestures

126 CHAPTER 6 SELF-REPORT INVENTORIESand history of depression were .23 and.22 (Williams et al., 1992). By comparison,the A-anx (Anxiety) scale correlated.23 with a history of depression, and theA-Ise (Low Self-Esteem) also correlated.26 with depression history. The majorityof the external validity coefficients seem tobe in the range of .20 to .30. The authorssuggest several reasons for these resultsand numerous methodological caveats,including the appropriateness of the criterionmeasures, sample sizes, compositionof the clinical samples, and other factors(Williams et al., 1992). It has also beensuggested that the A-dep and A-anx contentscales simply do not discriminate wellbetween the anxiety and depression constructs(Arita & Baer, 1998).The A-cyn (Cynicism) and the A-trt(Negative Treatment Indicators) scales producedlittle external validity data to supporttheir use (Butcher et al., 1992; Williamset al., 1992). The A-sch (School Problems)fared better than most by producing 44 significantcorrelations with meaningful externalcriteria in the normative sample.Over the years, several studies havebeen conducted to assess the ability of theMMPI-A to discriminate between levelsof symptomatology, diagnostic categories,and so forth. One such study has evaluatedthe ability of the MMPI-A to differentiatethree patterns of substance abuse (behavioralundercontrol, absence of behavioralundercontrol, and behavioral undercontroland overcontrol) for a sample of 180“substance abusers” (Gallucci, 1997). Severalscales were needed to predict groupmembership at a 79% correct classificationrate including MAC-R, D, Pd, Ma,Hy, Alcohol/Drug Problem Proneness,and Alcohol/Drug Problem Acknowledgment.Still another investigation used theMMPI-A to differentiate adolescent criminaloffenders who had violent infractions(Hicks, Rogers, & Cashel, 2000). TheMMPI-A was predictive of total numberof infractions, and Pd was associated morewith violent infractions. Archer and Slesinger(1999) investigated the relationshipbetween three MMPI-A suicidal ideationitems and score profiles. The three itemswere associated with higher clinical scaleT-scores in general. Finally, with regardto group differentiation validity, Cumella,Wall, and Kerr-Almeida (1999) used theMMPI-A to discriminate between casesof anorexia and bulimia. There are twofindings of interest: (1) the MMPI-A didnot differentiate the groups as well as theolder MMPI, and (2) bulimia patients differedfrom anorexia patients across content,supplemental, clinical, and validityscales suggesting different symptomsunderlying these disorders.An additional two studies have madedirect comparisons between the MMPI-Aand projective (i.e., Rorschach) and ratingscale (i.e., Achenbach CBCL and TRF)measures. Archer and Krishnamurthy(1997) compared the utility of the MMPI-A and Rorschach (Exner’s ComprehensiveSystem) for distinguishing between ConductDisorder and depression. In short, theMMPI-A did a better job of predicting diagnosticstatus with the Rorschach contributingonly two variables that accounted for asmall proportion of variance in a depressiondiagnosis, and none to a Conduct Disorderdiagnosis. The Rorschach results contributedno significant variance to the predictionbeyond that contributed by eitherthe D or A-dep scales. The MMPI-AA-con, A-cyn, and IMM scales were theonly predictors of Conduct Disorder diagnosis.Despite decades of research and useby clinicians, the MMPI-A as a validand useful measure in adolescent assessmentis not completely understood.Archer (2005) has provided a summaryof much of the work on the MMPI-A.In his summary, which is highlighted in Box6.3, Archer promotes an understanding of

CHAPTER 6 SELF-REPORT INVENTORIES127Box 6.3Understanding MMPI-A Results in a Developmental FrameworkArcher (2005) provided a review of MMPI/MMPI-A research that sought to demonstratehow the MMPI-A can contribute to knowledgeof adolescent development. In doing so,Archer also highlights some issues to considerwhen interpreting the results of an MMPI/MMPI-A and making clinical decisions andrecommendations. He emphasizes nine keypoints. The following is a brief summary ofthose points:1. “Generalizing adult findings to adolescentsis frequently inappropriate” (p. 258). Quitesimply, when adult norms are used withadolescents, even with available statisticalcorrections, adolescents tend to score in anelevated fashion2. “MMPI [MMPI-2] items are more effectivein discriminating normal from abnormalfunctioning for adults than [MMPI-Aitems are] for adolescents” (p. 260).3. “Maturational influences have profoundeffect on adolescents’ (and adults’) MMPIresponses” (p. 261). To illustrate this point,Archer presents data showing a decrease inadolescents’ MMPI-A raw scores throughoutthe teen years. In contrast, cross-sectionaldata reviewed by Archer show anincrease in scores on the MMPI-2 Hypochondriasisscale (Hs) with age.4. In general, “it is considerably more difficult todiscriminate normal from abnormal functioningamong adolescents than adults” (p. 263).Archer notes that many adolescents in clinicalsettings produce profiles with few, if any, elevations.Ironically, as Archer points out, thispattern may exist because many adolescentsin the non-clinical norm sample may haveendorsed a relatively high number of items.5. “The expression of psychopathology hasmany similarities across age groups (p.263). That is, despite notable developmentalinfluences on item endorsement, theunderlying implications of MMPI profilesfor adolescents or adults do not appear todiffer greatly.6. “Acting out is the ubiquitous defensemechanism among adolescents” (p. 265).To support this notion, Archer reports thelower frequency of L and K scale elevationsfor adolescents than for adults as wellas the relative commonality of adolescentclinical profiles that involve the PsychopathicDeviate (Pd) scale.7. “Adolescents in the juvenile justice andmental health systems are often similar”(p. 265) based on MMPI-A responses.8. “Given the fluid nature of symptomatologyduring adolescence, long-term predictionsbased on MMPI-A findings are ill-advised”(p. 267). Such a statement should be madeabout the results of any currently availableassessment tool for child or adolescent personality,emotional functioning, or behavioralfunctioning.9. A turbulent view of adolescence profferedby many early theories of adolescentdevelopment “receive substantialsupport from the MMPI/MMPI-A”(p. 267). This conclusion is drawn based onthe relatively high symptom endorsement byadolescents on the MMPI-A and the lack ofevidence indicating that such results are predictiveof long-term psychological difficulty.In other words, adolescents may, as a group,experience significant problems that warrantattention, but in many cases, these difficultiesare fortunately transient in nature.adolescent response patterns on the MMPI-A (and thus, evidence regarding potentialpsychopathology) within a developmentalframework.

128 CHAPTER 6 SELF-REPORT INVENTORIESInterpretationThe MMPI-A manual and subsequentliterature (e.g., Archer, Krishnamurthy,& Stredny, 2007) supplies several aids tointerpretation. Considerable psychometricinformation is available, including reliabilityand factor-analytic validity information.The potential import of reliability informationfor interpretation was outlined earlier.Similarly, factor-analytic data can help cliniciansunderstand MMPI-A results.If an adolescent obtains high score elevationson all of the clinical scales save Si andMf, for example, then the client is producinga high factor 1 score. This client, consistentwith the factor-analytic research, isshowing a high level of general anxiety ordistress. This result may not be of particularimport for differential diagnosis, but itis a sensible and predictable finding in lightof factor-analytic results.Furthermore, the MMPI-A, as with itsadult counterpart, includes Harris-Lingoesscales which are subscales that may helpthe clinician determine what types of itemsled to a clinical scale elevation. For example,The Depression (D) scale includes fiveHarris-Lingoes subscales (i.e., SubjectiveDepression, Psychomotor Retardation,Physical Malfunctioning, Mental Dullnessand Brooding).The previous description of scale itemcontent can also be most useful for understandingscale elevations. The MMPI-Amanual (Butcher et al., 1992) also providesa list of items and their scale membership(Table E-1), which can be useful for scaleinterpretation.Some questions to ask oneself wheninterpreting the MMPI-A could include:llHow does an adolescent get a high scoreon this scale? What are the behaviors,perceptions, and feelings assessed by thisscale (i.e., the item content)?How reliable are the scales that I wish tointerpret as being of some clinical value inthis case?lllllIs there a content scale analogue that ismore reliable and potentially more valid(e.g., D versus A-dep) than the clinicalscale?What does the pattern of scale elevationssuggest in terms of factor-analyticresearch?How is this scale reflective of nontestbasedclinical symptomatology?Does external validity evidence exist(e.g., differential validity studies) to supportor refute my interpretation of thisscale?Has the validity study been independentlyreplicated?Interpretation of the MMPI is also supportedby practical and thorough bookson the topic (e.g., Archer, 1997) as well asnumerous work-shops and other continuingprofessional development opportunities.The MMPI-A literature, while notnearly as extensive as that of the MMPI-2,is expansive. A specific example of this rangeof resources is a guide devoted singularlyto the use of the MMPI system in forensicwork (Pope, Butcher, & Seelen, 2000).A sample case (Box 6.4) on the followingpage illustrates briefly how MMPI-A scaleelevations can be used in conjunction withother assessment information to aid diagnosticdecisions and in determining the primarytarget(s) of intervention.Strengths and WeaknessesThe MMPI-A’s long history is simultaneouslyits greatest asset and liability.On the one hand, the volumes of MMPIresearch guide practice. In direct contrast,the original clinical scales are incongruentwith modern research on child and adolescentpsychopathology and test developmentmethods. The continually expandingresearch base and new scale developmentensures that the MMPI will enjoy continuedpopularity and utility. It is, after all, aunique self-report measure.

CHAPTER 6 SELF-REPORT INVENTORIES129Box 6.4A Sample Case Using the MMPI-AMichael is a 16-year-old high school juniorHe was referred for an evaluation becauseof recently increased anxiety symptoms thathave interfered with his social interactionsand functioning at school. More specifically,Michael has experienced some increasedconcerns with cleanliness or organization athome, and he has has begun to avoid socialsituations for fear of experiencing noticeableanxiety in front of others. Michael also has ahistory of depression, according to his mother,which has included a trial of antidepressantmedication about three years ago. Medicationwas discontinued after a couple of monthsbecause his symptoms had diminished.Michael currently lives with his parentsand his 11 year-old brother. He reportedlygets along well with his family, and he has historicallyperformed well academically until thecurrent school year when his grades have gonefrom “As” and “Bs” to “Bs,” “Cs,” and “Ds.”Michael was cooperative and attentivethroughout the test session. He appearedmotivated to do well. No speech, visual, auditory,or motor abnormalities were noted.Michael demonstrated tendencies towardperfectionism and anxiety during intelligencetesting. During the arithmetic subtesthe began to tremble and remarked, “I feelnervous.” He was visibly nervous (i.e., handsshaking, voice cracking) when faced withrelatively difficult items. He also seemed perfectionisticin his response style as evidencedby his unwillingness to give up on difficultitems. Michael also displayed depressed affectduring a clinical interview, although he didsmile on occasion. Michael’s intellectual testfindings were all within the High Averagerange. Likewise, all of his achievement subtestscores were in the High Average range.These findings are consistent with his educationalhistory.On the MMPI-A, Michael indicatedmoderate levels of depressive symptoms anda tendency to isolate himself socially. He wasself-described as shy, which is consistent withhis mother’s report. Content scale elevationson the Social Discomfort and School Problemsscales are consistent with referral concerns.More specifically, he reported beinguncomfortable in social situations, has difficultyinteracting with others, and avoids socialevents. In regards to problems at school,Michael indicated that he does not careabout doing well at school and that school isboring. Structured interviews with Michaelindicate that these issues are relatively recent.However, his mother expressed concern thatMichael has had a tendency to lose interestin activities such as school and social outingsover the last few years.The Depression (D) clinical scale wassomewhat elevated. Michael indicated that hehas feelings of inadequacy and worthlessness.Also, on the MMPI-A, Michael acknowledgedhaving lost interest in activities andlosing sleep due to worry. The reports ofMichael and his mother on structured interviewsindicate that he meets diagnostic criteriafor Dysthymic Disorder. Since childhood,Michael reportedly has had periods of cryingeasily, difficulty making decisions, feelings ofinadequacy, a lack of enjoyment from praiseor rewards, and feeling that he is not as goodas other people. He recently has experiencedincreasingly depressed mood, difficulty sleeping,and avoidance of activities.His social anxiety and negative attitudesabout school have apparently been recent developmentsthat do not warrant a diagnosis at thistime but should receive clinical attention.The findings of this evaluation support theneed for intervention at this time. Michaelcontinues to demonstrate depressive symptoms.In addition, he is beginning to exhibitsome anxiety in social situations which areinterfering with his desire to interact withothers and his enjoyment of school. A possible(Continues)

130 CHAPTER 6 SELF-REPORT INVENTORIESBox 6.4 (Continued)focus of psychotherapeutic interventions maybe on the self-deprecating nature of his cognitionsand anxiety during social interactions.ScaleL 56F 54K 49Hs 48D 67Hy 58Pd 51Mf 33Pa 52Pt 55Sc 58Ma 48Si 65Content Scale ElevationsSocial Discomfort (A-sod) 72School Problems (A-sch) 71T-scoreThe MMPI-A is likely to continue toremain as one of the most widely used adolescentmeasures of personality in the world.Claiborn (1995) concluded:Strengths of the MMPI-A include:1. Its familiarity to a large group of devotedusers2. The existence of a number of valuablevalidity scales3. Interpretive flexibility because of thenumerous scales4. A thorough evaluation of the adolescent’sself-appraisal due to the variety of itemspresented5. The availability of numerous books, chapters,and empirical articles devoted toMMPI interpretation, many of which offerhighly sensible interpretive guidancePotential weaknesses of the MMPI-Ainclude:“Its flaws are relatively minor, correctable,and enormously outweighed by the strengthsof the inventory. Clearly, the MMPI-A wasdeveloped with a great deal of care, expertise,and sensitivity to the problems of adolescentsand the needs of practitioners who work withthem” (p. 628).1. Retention of scales that have not beenwell-supported by validity evidence2. Retention of scales that lack internal consistency(Black, 1994)3. Failure to incorporate factor-analytic evidenceinto the test development process(e.g., consideration of composite scoresor clarifying how scales measure differingtraits or problems despite the fact that theyload on the same factor)4. Duplicating items on different scales whichproduces high intercorrelations, thusbringing into question the distinctivenessof measurement of individual constructs(Kline, 1995)5. Lack of a complete description of thenormative sample and little evidencethat the sample matches well a particularpopulation (e.g., US Census bureaustatistics)6. Length of the MMPI-A relative to otherself-report inventories, making the practicalityof administering the MMPI-A anissue.In spite of the extraordinary amount of MMPIresearch available, much remains to be done.The majority of the available research isbased on adult samples. However, we referthe reader to the work of Archer and colleagues(e.g., Archer, 2005; Archer, Handel, &Lynch, 2001; Archer & Krishamurthy, 1994)who have greatly contributed to the body ofknowledge regarding MMPI-A validity andinterpretation.

CHAPTER 6 SELF-REPORT INVENTORIES131Conners, 3rd Edition, Self-Report (Conners-3 SR;Conners, 2008a)The Conners-3 SR (Conners, 2008a) is themost recent addition to a rating scale systemwith a long history of research and clinicaluse. However, despite this long history, aself-report instrument had largely beenconsidered experimental until the mostrecent revision of the Conners rating scalesystem. The Conners-3 self-report ratingscale consists of 59 items which are writtenat approximately a third grade reading level.A 39-item Short Form also exists.It takes approximately 20 min to completethe Long Form of the Conners-3 SR.The inclusion of a standard self-report formmakes the Conners system competitive withother well-known rating scale systems highlightedin this chapter. It should be notedthat the Conners-3 SR includes extensiveassessment of externalizing problems, particularlyADHD, with screening of anxietyand depression. Depending on the referralissue, this design may be either ideal or lessthan-idealfor the clinician.We focus on the Conners-3 rather thanits companion rating scale system, theConners Comprehensive Behavior RatingScales (CBRS; Conners, 2008b), becausethe Conners-3 is relatively unique in itsextensive evaluation of externalizing problems.The CBRS self-report form (for ages8–18) provides an assessment of severalareas of behavioral, emotional, and academicfunctioning and is desirable if the clinicianneeds information in more domainsthan externalizing problems.Scale ContentThe Conners-3 includes four contentscales: Hyperactivity/Impulsivity, Inattention,Aggression, Family Relations, andLearning Problems. There are five DSM-IV-TR Symptom scales (i.e., ADHD,Combined Type; ADHD Inattentive;ADHD Hyperactive-Impulsive, ODD,and CD). As noted above, the Conners-3SR includes screening items for depressionand anxiety, as well as impairment itemsfor home, school, and social relationships.Like the BASC, the Conners-3 includescritical items that may signal the need forfurther follow-up. The critical items onthe SR are specific to severe conduct problems.Consistent with its predecessors,the Conners-3 includes a brief ADHDIndex. This scale is based on items thatbest differentiate ADHD from nonclinicalsamples. New to the Conners-3 are threevalidity scales: Positive Impression (or“fake good”), Negative Impression (“fakebad”), and the Inconsistency Index. Lastly, theConners-3 SR includes an open-endeditem assessing “strengths/skills.” Overall,the Conners-3 scales and the existence ofthe validity scales are in line with the currentstate-of-the-art in rating scale systems,as well as providing perhaps more detailedassessment of externalizing symptoms thanis typical for other rating scales.Administration and ScoringThe Conners-3 SR is designed for usewith youth ages 8–18. Responses are madeon a four-point scale, where 0 = not at alltrue (never, seldom), and 4 = very muchtrue (very much true, very frequent). Thetime frame for responses is the last monthprior to the assessment. Both hand scoringand computer scoring are availableas are secure Internet administration andscoring.The profile form that is included withthe response form is used to convert rawscores to T-scores. The T-scores for theConners-3 are linear T-scores, meaningthat the scales maintain their distributionswhen converted to T-scores. Maleand female norm-referenced scores are

132 CHAPTER 6 SELF-REPORT INVENTORIESobtained separately. In addition, each agehas separate norms. The T-scores shownon the available profile forms are truncatedsuch that T-scores below 40 are notspecified. Detailed step-by-step scoringprocedures are available in the Conners-3manual (Conners, 2008a).NormingThe normative sample of 1,000 caseswas collected mostly in the UnitedStates, with “a limited amount of data”(Conners, 2008a; p. 139) collected inCanada. Recruitment was targeted toapproximate the ethnic/racial distributionof the population according to USCensus statistics. Data presented in theConners-3 manual indicate that accuraterepresentation across ethnic groups wasattained. Only 8% of cases came fromthe western part of the United States,indicating some underrepresentation ofthis region. The majority (i.e., 70%) ofcases in the normative sample of the SRhad parents with at least some secondaryeducation.Equal numbers of girls and boys wereincluded in the normative sample at eachage. As noted above, norm-referencedT-scores are provided separately for boysand girls, which does limit interpretationto sex-only comparisons.ReliabilityInternal consistency coefficients for SR aregood for both the content scales and DSMsymptom scales. Specifically, coefficientswere all above .80 in all age groups acrossgenders, with the exception of the ConductDisorder scale for girls with coefficientsranging from .74 to .79 and the Aggressionscale for 8 and 9-year-old girls (i.e., .75).Two to four week test-retest reliabilitieswere also good, with coefficients all .71and higher (Conners, 2008a).ValidityFactor analysis was used extensively indevelopment of the Conners-3. Exploratoryand confirmatory factor analysessupport a four-factor model (i.e., FamilyRelations, Learning Problems, Aggression,and Hyperactivity/Impulsivity; Conners,2008a). Thus, these content scales are consideredempirically-derived, whereas theInattention scale is considered theoretical/rational. The SR content and DSM scalesshow moderate to high intercorrelationsand moderate correlations with analogousscales from the Conners-3 parent andteacher rating scales.Criterion-related validity was demonstratedthrough correlations with other ratingscales. Specifically, correlations betweenscales on the Conners-3 SR were moderatelyto highly correlated with analogousscales on the BASC-2-SRP and AchenbachYSR. The one exception was a non-significantnegative relation between the Relationswith Parents (higher scores indicatebetter relations) scale on the BASC-2-SRPand the Family Relations (higher scoresindicate worse relations) scale of the Conners-3SR for younger children (ages 8–11).This relation was significant and negativefor older youth (ages 12–18).Discriminant validity for the SR is alsosupported based on comparisons of clinicaland general populations. In particular, theLearning Problems, Inattention, Hyperactivity/Impulsivity,and Aggression scales alldifferentiated not only the clinical samplefrom the general sample, but they also tendedto differentiate among individuals within theclinical sample. For example, scores on theAggression scale were higher for individualsdiagnosed with disruptive behavior disordersthan individuals with other diagnoses.This pattern was also the case for the FamilyRelations scale. The DSM scales also faredwell in these analyses.It should be noted that the above informationpertains to the Long Form of theConners-3 SR in particular. Reliability and

CHAPTER 6 SELF-REPORT INVENTORIES133validity evidence of the Short and Indexforms are similarly good (see Conners,2008a).InterpretationConners (2008a) provides a clear recommendedapproach to interpretation ofthe SR and other scales in the Conners-3family. This approach is well-alignedwith the approach recommended in thistext. First, the validity scales should beexamined as to the potential usefulnessof the responses. The next interpretationshould occur at the scale level followedby a consideration of the overall “profile”or pattern of elevations. Then, item-levelresponses should be examined, includingthe anxiety and depression screeningitems, the items on elevated content orDSM scales, critical items, and strength/skills items. Lastly, reports from the SRshould not stand-alone in an assessmentbut instead should be integrated withother information.Strengths and WeaknessesSome of the strengths of the Conners-3 SRare:1. Good initial validity evidence.2. User-friendly manual that assists withscoring and interpretation.3. Availability of empirically-derived,rational, and DSM-oriented scales.4. Thorough coverage of externalizingsymptomatology5. Availability of a Short Form for efficiency6. Availability of validity scales.Some characteristics that may be consideredweaknesses are:1. Limited assessment of internalizingproblems.2. Limited assessment of adaptive functioning.3. Limited independent validity research todate.Personality Inventory for Youth(PIY; Lachar & Gruber, 1994)The Personality Inventory for Youth (PIY;Lachar & Gruber, 1994) has its roots inthe well-known Personality Inventory forChildren, which is designed as a parentrating scale (see Chap. 7). The PIY consistsof 270 True-False items designed toassess emotional and behavioral adjustment,school adjustment, family characteristicsand interactions, and academicability in children aged 9 through 18 years(Lachar & Gruber, 1994). The PIY offersa substantial array of scales designed toassess these issues.Scale ContentThe PIY features four broad-band factorscales; Externalizing/Internalizing, CognitiveImpairment, Social Withdrawal,and Social Skills Deficit. The Externalizing/Internalizingfactor includes severalscales, whereas the other three factors eachinclude the scale of the same name. Thescales and subscales of the PIY are:Cognitive ImpairmentPoor Achievement & Memory InadequateAbilitiesLearning ProblemsImpulsivity and DistractibilityBrashnessDistractibility & OveractivityImpulsivityDelinquencyAntisocial Behavior DyscontrolNoncomplianceFamily DysfunctionParent-Child ConflictParent MaladjustmentMarital Discord

134 CHAPTER 6 SELF-REPORT INVENTORIESReality DistortionFeelings of AlienationHallucinations & DelusionsSomatic ConcernPsychosomatic SyndromeMuscular Tension & Anxiety Preoccupationwith DiseasePsychological DiscomfortFear & Worry DepressionSleep DisturbanceSocial WithdrawalSocial Introversion IsolationSocial Skill DeficitsLimited Peer Status Conflict withPeersThe PIC item pool served as the basis fordeveloping items for the PIY (Lachar, 1999).In fact, only a few new items were added.The vast majority (i.e., about two-thirds)of PIY items are adapted from the PIC. Weraise the question regarding item contentbecause it is surprising to see a self-reportmeasure that does not produce familiarscales or composites (not subscales) forconstructs such as depression and anxiety.The authors explain (p. 34 of the TechnicalGuide) that they combined these scalesin order to measure “emotional distress.”The scale is labeled “Psychological Discomfort.”The lack of clear operationaldefinitions for Psychological Discomfort,Cognitive Impairment, Dyscontrol, andother scales and subscales hinders interpretation,and the clinician should pay closeattention to the item content within scalesand subscales in making interpretations.Interpretation could be hinderedbecause an idiosyncratically defined scalecannot be interpreted with the assistanceof the vast research associated with a wellresearchedconstruct such as depression.If a well-researched construct is definedand assessed (e.g., hyperactivity), then theuser can obtain information about diagnosis,prognosis, course of treatment, andso forth from the behavioral sciences atlarge-, rather than solely relying on theresearch base for a specific measure suchas the PIY. The content of the PIY is, inmany ways, broad and unique, but experiencewith administering and interpretingthe PIY would be necessary for successfullyarticulating the meaning of a client’sscale elevations and making appropriatedecisions.Validity ScalesFour validity scales are also featured. ThePIY Validity (VAL) scale is intended to assessthe presence of inattentive, oppositional,or provocative responses (e.g., responding“True” to “My teachers are trying to poisonme.”). The Inconsistency (INC) scale indicatesif a protocol was answered haphazardly.In other words, the INC scale gaugesthe respondent’s consistency on similarlyworded items. Dissimulation (FB) measuresthe tendency of the informant to fake bad ormalinger. Finally, the Defensiveness (DEF)scale provides an index that may reflect afake good or a social desirability responseset (Lachar & Gruber, 1994). A study ofthe validity of these scales found that theydetected inconsistent and overly positive ornegative responding as intended (Wrobelet al., 1999).Administration and ScoringThe PIY is a 270-item inventory in a True-False format. It takes approximately 30–60min to administer. The PIY manual (Lachar& Gruber, 1994) provides several helpfulguidelines for administration. The manualsuggests, for example, that the examinerexplain the directions to the examinee,even though the directions are includedin the Administration Booklet, in order toensure compliance and enhance rapport.The administration requires two components,the booklet, and WPS Autoscore TMAnswer Forms.A single “template” is included in theanswer form. The examiner adds rows

CHAPTER 6 SELF-REPORT INVENTORIES135and columns in order to obtain raw scores.These scores are then converted to genderbasedT-scores. The PIY does not offercombined norms by gender as an interpretiveoption. Both PC and mail-in computerscoringservices are also available.Abbreviated FormThe PIY also includes an 80-item abbreviatedform which consists of the first 80items of the full PIY. In addition, a uniquefeature within the abbreviated PIY is a32-item classroom screening tool (CLASS)that assesses for difficulties in classroomadjustment (Lachar, 1999).NormingThe PIY normative sample consists of2,327 regular education students testedin 1991–1992 in five states. The samplewas stratified to meet US Census Bureaustatistics for ethnicity, parental educationallevel (SES), and community size.In addition, the marital status of theparent(s) was also considered. Therewas a slight under-sampling of AfricanAmerican children. An additional normingsample of 1,178 cases was collected toproduce clinical norms. These cases werecollected from 50 facilities serving a varietyof inpatients and outpatients.ReliabilityInternal consistency coefficients are availablefor the PIY for both the normativeand clinical samples (see Table 6.8 for dataon the clinical samples). The reliabilities ofthe scales for the clinical sample are generallygood, producing a median of .85. Themedian reliability for the subscales is lower,at .73. In fact, for the clinical sample, eightof the internal consistency coefficients fallbelow .70. The internal consistency coefficientfor the learning problems scale, forexample, was the lowest, at .44.Table 6.8 PlY Scales and Internal ConsistencyReliabilities for the ClinicalSamplesScales and SubscalesInternalConsistencyCoefficientCognitive Impairment .74Poor Achievement and.65MemoryInadequate Abilities .67Learning Problems .44Impulsivity and.77DistractibilityBrashness .54Distractibility and.61OveractivityImpulsivity .54Delinquency .92Antisocial Behavior .83Dyscontrol .84Noncompliance .83Family Dysfunction .87Parent–Child Conflict .82Parent Maladjustment .74Marital Discord .70Reality Distortion .83Feelings of Alienation .77Hallucinations and.71DelusionsSomatic Concerns .85Pychosomatic Syndrome .73Muscular Tension and.74AnxietyPreoccupation with.60DiseasePsychological Discomfort .86Fear and Worry .78Depression .73Sleep Disturbance .70Social Withdrawal .80Social Introversion .78Isolation .59Social Skill Deficit .86Limited Peer Status .79Conflict with Peers .80Note: Adapted from Table 46 in Lachar and Gruber(1994).

136 CHAPTER 6 SELF-REPORT INVENTORIESThe 7–10 day test-retest coefficientsfor the clinical sample are also good atthe scale level, with a median of .82. Themedian coefficient at the subscale level wasagain lower, at .73. At the subscale level, atotal of seven subscales yielded test-retestcoefficients of .70 or less. The lowest coefficientwas .58 for the Impulsivity subscale,and the highest was .88 for the AntisocialBehavior and Dyscontrol subscales (Lachar& Gruber, 1994).The manual also reports useful estimatesfor the standard error of measurementfor each scale and subscale in T-scoreunits. The typical SEM at the scale levelis about 4 or 5 T-score points for the testretestestimates.ValidityThe major argument presented for contentvalidity of the PIY is that the itemswere derived from the PIC-R item pool.Lachar and Gruber argue that any validityevidence already gathered for the PIC-Rlends indirect support for the validity ofthe PIY. However, it is worth noting thatbecause items on the PIY are necessarilyself- referent and items on the PIC are necessarilynot, data regarding the PIC cannotbe extrapolated toward understandingresponse tendencies on the PIY. In regardsto adolescent self-reports of psychologicalconstructs, it is unclear why the PIY doesnot include any items directly relevant tohigh-incidence syndromes of childhoodsuch as depression and anxiety.Several criterion-related validity studiesare included in the manual. A study of 152adolescents produced very modest relationsbetween the PIY and the original MMPIclinical scales. Among the highest correlationswere Reality Distortion and MMPISchizophrenia (r = .66), and PsychologicalDiscomfort and Psychasthenia (r = .65).The majority of correlations were in the.20 to .50 range, and many of the correlationswere not statistically significant.Another study correlated the PIY withthe State-Trait Anxiety Inventory and theReynolds Adolescent Depression Scale for79 cases.These data allow for the evaluation ofthe criterion-related validity of the PIYPsychological Discomfort scale and itscomponent subscales. The correlationsranged from modest to strong. The State-Trait correlated .51 and the Reynolds .70with the PIY Psychological Discomfortscale, suggesting that this scale may measuremore depressive than anxiety symptomatology,although indicators of both constructsappear to be part of the Psychological Discomfortscale. It is difficult to summarizethe wealth of data included in these studiesin the limited space available here. The clinicianwho is seriously interested in the PIYwould be served well by reading the criterion-relatedvalidity studies included in themanual very carefully prior to interpretingthe scales and subscales.Several samples were also used to assessthe factor invariance of the PIY by genderand ethnicity (Lachar & Gruber, 1994).The results were generally supportive ofthe hypothesis of factor invariance acrossgroups, although cross-validation withindependent samples should be soughtbefore drawing definitive conclusions.An additional study supported the relevanceof PIY items for detecting peerrelatedproblems. Wrobel, Lachar, andWrobel (2005) found that PIY scales constructedof items indicative of peer problemsand peer withdrawal were significantlyrelated to analogous peer-reported items.Another study in Bilingual adolescentssupported the correspondence between theEnglish version of the PIY and a Spanishversion of the instrument (Negy, Lachar, &Gruber, 2001).InterpretationA five-step interpretive system for the PIYis offered by the authors (Lachar & Gruber,

CHAPTER 6 SELF-REPORT INVENTORIES1371994). The first step involves assessingthe validity of the obtained results usingthe four validity scales and a review of thecompleted form.The second interpretive step suggestsidentifying primary profile elevations.A primary scale elevation is defined asa clinical scale with a T-score of 60 andsubscale member of this same scale witha T-score of at least 65 for the majorityof scales, although this cut score varies.These primary scale elevations may identifythe diagnostic issues most likely to requireadditional attention and/or study.The third interpretive step pertains tosecondary scale elevations. These significantprofile elevations are defined as scalesthat exceed the clinical cut score that arenot accompanied by corresponding subscalescores that exceed the clinical cut score. Inaddition, subscales that exceed the cut score(without a corresponding high scale score)are also interpreted in this step. These secondaryscale elevations may reveal mildproblems, issues linked to the primary areaof concern (e.g., social problems secondaryto one’s externalizing behaviors), and/orfrequently occurring personality characteristics(Lachar & Gruber, 1994).Step four includes tallying and interpretingitems labeled as critical. This extensivelist is intended to identify items that suggestclinical issues that should be examinedin greater detail (Lachar & Gruber, 1994).The last interpretive step is probably themost crucial-the integration of PIY resultswith other findings. The complexity of thisinterpretive step cannot be overstated. Oursuggested approach for integrating informationis described in detail in Chap. 15.The PIY manuals provide detailedinterpretive guidance that is grounded inscientific evidence. Cut scores and manydecision rules are based on careful investigationsof clinical samples. The chapter ofcase studies included in the administrationand scoring manual should also be highlyvalued by users.Strengths and WeaknessesThe PIY is clearly the result of a thoroughand thoughtful test development process asis duly noted in reviews of the instrument.Destefano (1998), for example, concluded:“Given the shortcomings of projective testingand parent report for this age group, thePIY is a welcome addition to clinical assessmentbatteries for children of ages 9 through18” (p. 757).Marchant and Ridenour (1998) also havehigh praise for the PlY: “Seldom does aself-report instrument enter the field withthe background of the PIY” (p. 758). Thesereviews also offer some caveats that are ofinterest to potential users.We wish to add two caveats. First, wesuggest that PIY users study the itemcontent of the clinical scales, rather thanrelying on scale labels, in order to developan understanding of the constructs beingmeasured and assess their correspondenceto similar constructs in the child psychopathologyliterature. Second, we cautionusers that virtually all of the extant PIYresearch to date was generated by membersof the PIY authorship and researchteam. We hope that more independentinvestigations will be forthcoming tocross-validate, extend, and clarify the currentfindings.Some specific strengths of the PIY includethat:1. The PIY is supported by a thorough setof manuals that give considerable statisticaldata, guidelines for interpretation,and case studies.2. The PIY appears to assess a broad spectrumof child and adolescent behavior.3. The PIY is part of a comprehensiveassessment system that includes bothparent and teacher reports.4. The PIY interpretive system is thoughtfuland helpful.

138 CHAPTER 6 SELF-REPORT INVENTORIES5. Computer scoring options are available forthe PIY, which enhances its practicality.Some potential weaknesses of the PIY mayinclude:1. Evidence of content validity is not compelling.Some of the item placementson scales do not seem consistent withrational/theoretical approaches to testdevelopment, which are gaining favorin personality assessment.2. The use of norm-referenced scoresat age 9 seems questionable given thesmall norming sample at this age (i.e., N= 70 for the regular education sample)(Lachar & Gruber, 1994).3. The subscales should likely be usedwith caution for clinical purposes, ornot used by the inexperienced PIY user,until further reliability studies are conducted(Destefano, 1998).4. The PIY has a limited research baseoutside of studies published by thedevelopers.Single Construct PersonalityInventoriesIn addition to an ever-growing body of omnibusself-report rating scales, a number ofsingle construct or single domain scales exist.These scales are typically oriented towardolder children and adolescents and towardan additional assessment of internalizingproblems. Some examples of these scales arereviewed in later chapters when we discussthe assessment of specific constructs.Given the need to approach assessmentcomprehensively, particularly in regards toassessing for comorbidity, we recommendfirst administering interviews and ratingscales that evaluate a variety of domains. Asinformation indicates problems in a particulardomain or that more information ona construct would aid in decision-makingand recommendations, well-validated singleconstruct inventories may be useful.ConclusionsThe current state of self-report personalityassessment continues to improve. Inparticular, a variety of assessment systemshave developed methods that allow forrelatively easy interpretation across self-,parent-, and teacher-reports. Personalityassessment through self-report ratings continuesto lag behind assessment of intelligenceand achievement testing in manyways. Subscale reliabilities are often stilltoo low to support diagnostic decisions.In some cases, they are too low to supporthypothesis generation. Only a few of theinstruments discussed herein have beenempirically checked for item bias, and fewhave used modern statistical methods suchas structural equation modeling and latenttrait methodology as is commonly done forintelligence and achievement tests.Improvements have been made innorming and the inclusion of reliability andvalidity data in test manuals. Some of thenoteworthy contributions have been theextended age range and expansion of adaptivecompetencies on the BASC-2-SRP, theidentification of correlates of profiles for thePIY, the history of numerous cross-culturalstudies with the YSR, the extensive assessmentof various externalizing problems onthe Conners-3, and the burgeoning validityevidence for the MMPI-A content scales.Psychologists clearly have more and betteroptions than in years past, and youth selfreportof emotional, social, and behavioralfunctioning is valuable for obvious reasons.Still, much validity evidence remains to begathered, particularly with so many measuresbeing relatively new to the market.Relatively few validity studies are availablefor some of these scales. Much more evidencewill have to be gathered for all measuresin order to approximate the higher

CHAPTER 6 SELF-REPORT INVENTORIES139standards of practice required today and todevelop a truly evidence-based approach toassessment.Chapter Summary1. Until recently, parent and teacherreports have routinely been preferredover the use of self-report inventoriesin child personality assessment.2. Decisions to use self-report inventoriesshould be based, in part, on thechild’s developmental level, the constructsbeing assessed, and the purposeof the assessment. In general, olderchildren are more useful informants.Self-reports of covert conduct problemsand internalizing symptoms maybe particularly informative.3. The SRP is one of many componentsof the BASC-2. The SRP attempts togauge children’s perceptions and feelingsabout school, parents, peers, andtheir own behavior problems.4. Five composites are available for the SRP:the Emotional Symptoms Index (ESI),Inattention/Hyperactivity, InternalizingProblems, School Maladjustment, andPersonal Adjustment.5. The SRP includes three validityscales.6. The SRP was normed on a nationalsample of 1,500 children and 1,900adolescents.7. The reliability of the SRP scales is goodas indicated by a variety of methods.8. Until clinical experience and furtherresearch studies are available, initialefforts at SRP interpretation shouldfocus on the scale level rather thanthe composite level because these itempools have some rational, theoretical,and research basis.9. Relatively few independent studies havebeen conducted with the most recentlyupdated YSR. However, numerous studieshave demonstrated that the previousversion of the YSR produces similarscores across cultures and consistent sexdifferences across cultures.10. The YSR does not include validityscales. However, it includes an assessmentof clinical and adaptive domainsthat are relevant for most purposes ofchild assessment.11. The MMPI-A has ten clinical scales,eight validity scales, and several empirically-basedcontent scales.12. The MMPI-A is unusually lengthy iscomparison to other self-report inventoriesdesigned for children and adolescents.This length entails specialadministration guidelines.13. There are distinct differences in theinternal consistency estimates for theMMPI-A, with many adequate scalesand some that are below minimumstandards.14. Factor analysis of the MMPI-Areveals four factors: the first twoare labeled general anxiety andovercontrol or repression, and thethird and fourth factors are composedsolely of the Si and Mf scales,respectively.15. The self-report form of the Conners-3is a recent addition to a long-standingrating scale system.16. The Conners-3 SR has good initialvalidity evidence. It offers an extensiveassessment of ADHD and disruptivebehaviors but a limited assessment ofinternalizing problems.17. The PIY manual provides considerableinterpretive guidance.18. The PIY used the PIC item pool (seeChap. 7) as its source of items.19. A number of single domain self-reportrating scales, particularly for internalizingproblems, are in existence. Examplesof these instruments are discussedin later chapters.

C h a p t e r 7Parent and Teacher Rating ScalesChapter Questionsl How reliable are parent and teacherratings of child behavior problems?l What domains of behavior are assessedby parent and teacher rating scales?l How are parent and teacher ratingscales used in the typical psychologicalevaluation?l Why are teachers important sources ofinformation about a child’s emotionaland behavioral adjustment?l To what extent do parents and teachersagree in their ratings of childrenand adolescents?l What factors influence this agreement?l What factors should play a role in theuse and selection of parent and teacherrating scales?Evaluating Children viaParent RatingsIt has long been recognized that children areoften less-than-accurate reporters of theirown behavior. Furthermore, children maynot have sufficient reading or oral expressionskills for self-report purposes (Lachar,1990). Problems with underreporting andresponse sets have always been well-recognizedby clinicians and, to some extent, havebeen documented by research (see previouschapter). These concerns about child selfreportshave undoubtedly contributed tothe popularity of parent rating scales. Furthermore,the parental perspective is ofteninvaluable when conceptualizing a case; thatis, because children are often referred for anevaluation because of a parent’s concerns,information on the parent’s perspective of achild’s problems is critical.141P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_7, © Springer Science+Business Media, LLC 2010

142 chapter 7 Parent and Teacher Rating ScalesParent ratings of child behavior possessadditional advantages, including brevityand cost efficiency (Hart & Lahey, 1999).The time-efficient nature of parent ratingsmakes it easy to collect additional informationabout child behavior. Given theimportance of parental influence on childbehavior, parental perceptions of behaviorshould routinely be collected in clinicalassessments.Today, the commonly used parent ratingscales routinely provide a broad coverageof problems. For example, whilethe unstructured interview may allowthe clinician to carefully evaluate a specificarea of functioning, other importantbehavioral problems or areas of concernmay be missed (Witt, Heffer, & Pfeiffer,1990). Parent or other caretaker ratingsalso foster objectivity and clarity inthe assessment process. Because of thebehavioral specificity of the typical itemcontent of these measures, parents arerequired to operationally define theirconcerns and provide specific and objectiveratings of hyperactivity, depression,nervousness, and the like (Witt et al.,1990).All rating scales, including parentratings, can be influenced by bias andrater response sets (Witt et al., 1990).Even biased reporting, however, can beof value. If, for example, parent ratingsprovide very different results when comparedto the ratings of others, the cliniciancan develop some important insightsinto the child’s family functioning. If achild’s father rates his son as having significantlymore behavioral problems thanthe mother, the clinician can explore thedynamics behind the ratings. A straightforwardexplanation may be that thefather is doing the majority of the childcare. This information could be importantto acquire if the presumption hadbeen that the child’s mother was providingmost of the caretaking.Factors Influencing ParentRatingsAs discussed in more detail in Chap. 15,research has indicated that parental, specificallymaternal, distress may influencethe ratings of child functioning in a negativeway. Although the issue of whether ornot maternal distress is directly influentialon the ratings of a child’s symptoms is farfrom settled, it stands to reason that stressfulhome environments would be positivelycorrelated with parent reports of childsymptoms.The construct being evaluated andthe child’s developmental level are twoadditional factors that may influence parents’reports. Teachers have traditionallybeen considered superior to parents asreporters of a child’s ADHD-like symptoms(Loeber, Green, & Lahey, 1990;Loeber et al., 1991; Tripp, Schaughency,& Clarke, 2006). However, parents arestill considered necessary and useful inproviding information about inattentionand hyperactivity (Tripp et al., 2006) andin documenting the effects of treatmentfor ADHD (Biederman, Gao, Rogers, &Spencer, 2007). In addition, parents maybe in a unique position to understand theantecedents of a child’s disruptive behaviors,as they can observe their child moreclosely than a teacher who works withseveral children simultaneously. Parentshave also been discussed as particularlyimportant observers and informantsof child anxiety and depression (Klein,Dougherty, & Olino, 2005; Silverman &Ollendick, 2005).As reviewed by De Los Reyes and Kazdin(2005), research has reached mixedconclusions about the degree to whichthe child’s age influences agreement inratings across informants. Parents, inparticular, may be useful informants ofa child’s functioning throughout childhoodand adolescence, although there

chapter 7 Parent and Teacher Rating Scales143may be discrepancies between theirreports and the reports of others. Theinformation used in conjunction withparent reports may vary with age. Morespecifically, parents are obviously vitallyimportant sources of information foryoung children in such areas as conductproblems, whereas the childrenthemselves would not be reliable (andthus, not valid) informants. Teachers,however, could offer useful perspectivesof the young child’s social, academic, andbehavioral functioning. For adolescents,parents may still provide valid and usefulinformation, but their knowledgeof the child’s conduct and behaviorproblems may be more limited, as thebehaviors may sometimes occur outsideof the parent’s awareness. The adolescent– provided that he or she is willingto provide such information – wouldbe the most knowledgeable informantof these behaviors, and the teacher’scontribution would also presumablydiminish.Finally, parent ratings are more likelyto attribute the child’s problems to dispositionalfactors in the child, whereas youthself-reports are more likely to indicate thefamily environment as a factor in need ofintervention (see De Los Reyes & Kazdin,2005). Thus, informants (includingparents, teachers, and children) may basetheir ratings of a child’s functioning onthe attributions that they make regardingthe genesis and maintenance of the child’sproblems.That parent ratings may be influencedby factors that are not necessarily directlytied to the child’s actual functioning doesnot render parent ratings questionable.Instead, it calls to mind the many potentialvariables to consider in understanding thechild’s presenting problems – an understandingthat is critical for case conceptualizationand subsequent recommendationsfor intervention.Evaluating Children viaTeacher RatingsAlthough teachers have traditionally beenconsidered an important source of informationabout children’s academic performance,they have not often been used inthe assessment of children’s behavioraland emotional functioning. However,knowing how a child behaves in the classroomis important for several reasons.First, school is a setting in which thechild spends several hours a day. Therefore,a child’s adjustment to the schoolsetting can have a dramatic impact onhis or her overall psychological functioning.Second, the multiple demands of theschool environment (e.g., to stay seated,to follow the demands of adults, to interactwith classmates) present many challengesto the child— challenges that maynot be present in other settings. Third,the demands of the school setting changeas a child progresses through school (e.g.,demands for organization, the importanceof social acceptance). Therefore,understanding school-related problemsthat are unique to a given period can provideclues to specific problems in adaptationthat a child or adolescent mightexperience.On the basis of these considerations,there is increasing interest in assessing achild’s behavioral and emotional functioningin the school setting. Given the manyadvantages of behavior rating scales, suchas time-efficiency and objectivity, it is notsurprising that the primary assessmentinstruments for children’s school behaviorhave been teacher-completed behaviorrating scales. In addition to suggestionsfor appropriate use of rating scales in generaldiscussed previously, there are severalconsiderations for interpreting informationfrom teachers that warrant specialattention.

144 chapter 7 Parent and Teacher Rating ScalesFactors Influencing TeacherRatingsAs described above for parents, the usefulnessof teacher information may varydepending on what type of behavior isbeing assessed. Teachers are often consideredthe best source of information abouta child’s attention problems and overactivitybecause they have the opportunityto observe the child in a situation thatdemands sustained attention and inactivity.In contrast, teachers’ ratings tend to be lessuseful in assessing many types of antisocialbehavior that are unlikely to occur in theschool environment (e.g., setting things onfire, being cruel to animals) or for internalizedtypes of problems that may not bereadily observable in the classroom setting(Loeber et al., 1991).The usefulness of teacher informationmay also vary according to the age of thechild. Children in early elementary schoolfrequently have one teacher who observes achild across several class periods, if not theentire school day. In contrast, high schoolteachers frequently have students for oneclass period during the day. Therefore, theusefulness of information may decrease asa child advances in school and contact withany single teacher decreases (Edelbrocket al., 1985).A final issue in interpreting teacherrating scales is understanding the frameof reference or standard used by teachers.As discussed previously (e.g., Piacentini,1993), a number of characteristics ofa rater can influence his or her judgmentof the intensity, quality, and/or frequencyof a child’s behavior. In the case of teacherratings, a characteristic of teachers that caninfluence their ratings is their experiencewith many children of the same age. Experienceallows the teacher to make someinternal normative comparison of a child’sbehavior with the behavior of other childrenthe teacher has taught. This internalnorm is a double-edged sword. It oftengives the teacher a unique perspective ofknowing both the individual child and thebehaviors that are age-appropriate. However,some teachers, such as teachers whowork in special education classrooms, mayhave a skewed base of comparison thatcould influence their ratings. That is, theirratings of a child’s behavior may be influencedby a comparison of the child withother disturbed children.Despite these cautions and limitations,we feel strongly that teacher ratings arean essential element of a comprehensiveclinical assessment of children’s behavioraland emotional functioning. Carlson andLahey (1983), in an early review of teacherratings, reported that most of the teacherrating scales available at that time sufferedfrom significant psychometric problemsin development and inadequate norming.As a result, the available scales wereseverely limited in their usefulness forclinical evaluations. Fortunately, since that1983 review, there have been numerousadvances in the teacher rating scales andthe emergence of new scales, with manyof the inadequacies of earlier scales eliminatedor greatly reduced.Overview of Omnibus Parentand Teacher RatingScalesParent and teacher rating scales are notinterchangeable and, with the seeminglyexponential growth of such instruments,psychologists have to make many decisionsabout the utility of various measures.This chapter attempts to aid theclinician in decision making by providingan overview of the variety of scales available,with particular attention devotedto defining the strengths and weaknessesof each measure. Writing such a chapterrequires selectivity. Hence, if a scale is not

chapter 7 Parent and Teacher Rating Scales145mentioned in this chapter, it should notbe construed as a judgment of the qualityof the scale. As with our chapter on selfreportrating scales, we have attempted toreview those instruments that are widelyused and/or part of a long-standing systemof rating scales used for child andadolescent assessment. This broad overviewof the various scales is not designedto replace information provided in thetechnical manuals that accompany theseinstruments, to be reviewed by any userof the scales. Optimally, however, theprinciples applied to evaluating parentand teacher rating scales in this chaptercan be used by psychologists to evaluateother scales as well.The parent and teacher rating scalesreviewed in this chapter are highlightedin Tables 7.1 and 7.2, respectively. Commonlyused omnibus measures that assessmany different domains, as opposed tosingle construct measures, are the focus.Although these scales are discussed in isolationto balance clarity and specificity, itshould be recalled that they are often partof larger multimethod assessment methodsthat are discussed in various chaptersof this book. The integration of componentsand information from differentinformants and methods is discussed inthe context of interpretation in Chap. 15and in subsequent chapters that addressspecific syndromes.Behavior Assessment Systemfor Children, 2nd Edition(BASC-2)Parent Rating Scale (PRS)The BASC-2 Parent Rating Scale (BASC-2-PRS; Reynolds & Kamphaus, 2004) ispart of the larger BASC system. The PRSwas published concurrently with the SRP(discussed in Chap. 6) and TRS (see below)as well as other components of the BASCassessment system (Reynolds & Kamphaus,2004). The PRS has three formscomposed of similar items and scales thatspan the preschool (2–5 years), child (6–11years), and adolescent (12–21 years) ageranges. The PRS takes a broad sampling ofa child’s behavior in home and communitysettings.ContentAs with its predecessor, the BASC-2-PRSwas developed using both rational/theoreticaland empirical means in combination toconstruct the individual scales. The benefitof this approach is that the resulting scaleshave relatively homogenous content. Theuniqueness of the scales was also enhancedby not including items on more than onescale. Table 7.3 provides item examples foreach scale. There are four composites: ExternalizingProblems, Internalizing Problems,Adaptive Skills, and a Behavioral SymptomsIndex that includes some internalizing andexternalizing scales (i.e., Atypicality, AttentionProblems, Hyperactivity, Aggression,Depression, and Withdrawal).Two types of scales are included at eachage level: clinical and adaptive. Clinicalscales of the PRS are designed to measurebehavior problems much like other measuresdiscussed below in that behavioralexcesses (e.g., hitting others) are the focusof assessment. The PRS also includes criticalitems that are thought to warrant followupor clinical attention in their own right.These items (e.g., “Has a hearing problem.”)are not necessarily indicative of themost severe pathology; instead, they maybe worthy of further questioning or recommendationsby the clinician. The adaptivescales measure behaviors (e.g., complimentsothers) or skills that are associated with

146 chapter 7 Parent and Teacher Rating ScalesTable 7.1 Overview of Parent Rating ScalesInventory Ages ContentBehavior AssessmentSystem forChildren, 2ndedition (BASC-2;Reynolds &Kamphaus, 2004)Achenbach ChildBehavior Checklist(CBCL; Achenbach& Rescorla,2000; Achenbach& Rescorla, 2001)2–5; 6–11;12–21134 –160 items;BehavioralSymptoms Index,Adaptive Skills,InternalizingProblems; 3 validityscales; 9 clinicalscales, 5 adaptivescales; 7 optionalcontent scales1½– 5; 6–18 100–113 items;Internalizing,Externalizing,Competence; 8problem scales;6 DSM-orientedscalesChild SymptomInventory-4(Gadow &Sprafkin, 1998)ESI/CSI/ASI3–5; 5–12;12–1897 items correspondingto 17disordersReadingLevelTime toAdministerNormSample Reliability Validity MI4th grade 10–20 min E Internal consistency(good); 1–7week test–retest(good); Interrater(moderate to high)Moderate to highwith similar tools;Support forfactorsENot given 15–20 min G Internal consistency(good forschool age version);1 week test–retest(good); interrater(moderate forpreschool version;good for school ageversion)Not given 10–15 min F Internal consistency(moderate togood); Test-retest;content (moderateto good)Good validity(clinical vs.nonclinicalgroups); limitedinfo. on relationswith other tools;limited info. oncurrent versionContent derivedfrom DSMcriteria;Evidence ofconstruct validity(good)EG(Continues)

chapter 7 Parent and Teacher Rating Scales147Table 7.1 (Continued)Inventory Ages ContentReadingLevelTime toAdministerNormSample Reliability Validity MIConners 3rd Edition,Parent RatingScale (Conners-3;Conners, 2008a)Personality Inventoryfor Children,2nd edition(PIC-2; Lachar &Gruber, 2001)6–18 110 items;43-item shortform; 3 validityscales; 6 empiricalscales; 1 rationalscale; 5 DSM-IV-TR scales5–19 275 items, 96item short form;9 clinical scales(several subscales);3 validityscales4th–5thgrade20 min E Internal consistency,2–4 weektest–retest, andinterrater reliabilityall good4th grade 40 min E Internal consistency;test–retest;interrater; Goodreliability, somewhatlower forsubscalesGood evidence ofcriterion-relatedand differentialvalidityValidity (based onteacher and selfreportsof samesystem); lack ofresearch on relationswith othercriteriaE aGNote: Norm Sample evaluated as E = Excellent, G = Good, F = Fair; MI = Degree of enhancement of multi-informant assessment; E = Excellent, G = Good;ESI = Early Symptom Inventory; CSI = Child Symptom Inventory; ASI = Adolescent Symptom Inventory.aWe view the Conners-3 as providing an excellent means of assessment of externalizing problems; however, if an extensive assessment of internalizing or learningdifficulties is desired, the Conners Comprehensive Behavior Rating Scales (CBRS; Conners, 2008b) is recommended.

148 chapter 7 Parent and Teacher Rating ScalesTable 7.2 Overview of Teacher Rating ScalesInventory Ages ContentBehavior AssessmentSystem forChildren, 2ndedition (BASC-2;Reynolds &Kamphaus, 2004)AchenbachTeacher RatingForm (TRF;Achenbach &Rescorla, 2000;Achenbach &Rescorla, 2001)Child SymptomInventory-4(Gadow &Sprafkin, 1998)ESI/CSI/ASIConners 3rdEdition, TeacherRating scale(Conners-3-T;Conners, 2008a)2–5; 6–11;12–21100–139 items; BehavioralSymptoms Index,Adaptive Skills, ExternalizingProblems,Internalizing Problems,School Problems;3 validity scales; 10clinical scales, 5 adaptivescales, 7 optionalcontent scales1½–5; 6–18 100–113 items; Internalizing,Externalizing,Adaptive Functioning;8 problem scales; 6DSM-oriented scales3–5; 5–12;12–1877 Items correspondingto 13 disorders8–18 115 Items; 39-itemshort form; 3 validityscales, 6 empiricalscales, 1 rational scale,5 DSM-IV-TRsymptom scalesReadingLevelTime toAdministerNormSample Reliability Validity MINot given 10–20 min E Internal consistency(good); 1–9week test–retest(good); interraterreliability (goodwith exception ofSomatiz)Moderate to highrelations with similartools; Supportfor factorsENot given 15–20 min G Internal consistency(good); 7month test–retest(moderate);interrater (low tomode-rate)Good validity (clinicalvs. nonclinicalgroups); limitedinfo. on relationswith other toolsENot given 10 min F Test–retest;limited evidenceavailable4th–5thgrade20 min G Internal consistencyand test–retest reliability(good). Interraterreliability (moderateto good)Content derivedfrom DSM criteria;Classification ratesmatch well withepidemiologyGood evidence ofcriterion-relatedand differentialvalidityEE a(Continues)

chapter 7 Parent and Teacher Rating Scales149Table 7.2 (continued)Inventory Ages ContentReadingLevelTime toAdministerNormSample Reliability Validity MIStudent BehaviorSurvey (SBS;Lachar, Wingenfeld,Kline, &Gruber, 2000)5–18 102 items; AcademicResources, HealthConcerns, EmotionalDistress, UnusualBehavior, Social, andAggression compositesNot given 10–15 min E Good reliability(internal consistency;test–retest)Good convergentvalidity for mostscales; DBD scalesdo not overlap fullywith DSM criteria;Limited validityevidence for somescalesFNote: Norm Sample evaluated as E = Excellent, G = Good; MI = Degree of enhancement of multi-informant assessment; E = Excellent, G = Good, F = Fair; ESI =Early Symptom Inventory; CSI = Child Symptom Inventory; ASI = Adolescent Symptom Inventory; DBD = Disruptive Behavior Disorder.aWe view the Conners-3 as providing an excellent means of assessment of externalizing problems; however, if an extensive assessment of internalizing or learningdifficulties is desired, the Conners Comprehensive Behavior Rating Scales (CBRS; Conners, 2008b) is recommended.

150 chapter 7 Parent and Teacher Rating ScalesTable 7.3 BASC-2-PRS Scale Definitions and Key Symptoms as Indicated by Items with theHighest Factor Loadings Per ScaleActivities of Daily LivingAdaptabilityAggressionAnxietyAttention ProblemsAtypicalityConduct ProblemsDepressionFunctional CommunicationHyperactivityLeadershipSocial SkillsSomatizationWithdrawalSkills associated with performing everyday tasks; “Acts in a safe manner”,“Sets realistic goals”, “Attends to issues of personal safety”Ability to adapt to changes in the environment; “Adjusts easily to newsurroundings”, “Adjusts well to changes in family plans”, “Recoversquickly after a setback”Tendency to act in hostile or threatening manner; “Is cruel to others”,“Loses temper too easily”, “Annoys others on purpose”Tendency to be nervous, fearful, or worried; “Worries about makingmistakes”, “Worries about what other children think”, “Is nervous”Tendency to be easily distracted or have difficulty concentrating;“Has a short attention span” “Listens carefully” (reverse scored); “Iseasily distracted”Tendency to behave in odd manner; “Acts strangely.” “Says thingsthat make no sense”, “Seems out of touch with reality”Tendency to engage in antisocial and rule-breaking behavior; “Breaksthe rules”; “Deceives others”; “Gets into trouble”Feelings of unhappiness, sadness, or stress; “Is negative about things”,“Says ‘I don’t have any friends’”, “Seems lonely”Ability to communicate ideas and express oneself clearly; “Communicatesclearly”, “Responds appropriately when asked a question”,“Accurately takes down messages”Tendency to be overly active and act without thinking; “Acts out ofcontrol”, “Interrupts others when they are speaking”, “Disrupts otherchildren’s activities”Possessing skills needed to accomplish goals, ability to work with others;“Gives good suggestions for solving problems”, “Is creative”, “Isa ‘self-starter’”Having the skills necessary to interact successfully with peers andadults; “Encourages others to do their best”, “Offers to help otherchildren”, Congratulates others when good things happen to them”Tendency to be sensitive to, and complain about, minor physical ailments;“Complains about health”; “Gets sick”; “Complains of beingsick when nothing is wrong”Tendency to avoid others; “Makes friends easily” (reverse scored),“Avoids other children”, “Quickly joins group activities” (reversescored)Note: Adapted from Tables 7.7 and 10.3 of the BASC-2 manual (Reynolds & Kamphaus, 2004).The clinicalnorms may be especially important when assessing a child in a residential setting to be able to compare him orher to others with relatively severe difficulties. That is, it may be understood that a child is functioning poorlycompared to most other children his/her age (i.e., elevations on general norms), but it may be informative toconsider how the child functions (e.g., “How severe are his conduct problems?”) in comparison to other childrenwith emotional and behavioral difficulties for treatment planning purposes.

chapter 7 Parent and Teacher Rating Scales151good adaptation to home and community(see Table 7.3).Each of the parent forms of the BASC-2includes seven optional content scales:Anger Control, Bullying, Developmental,Social Disorders, Emotional Self-control,Emotional Self-control, ExecutiveFunctioning, Negative Emotionality, andResiliency. As with the BASC-2-SRP(see Chap. 6), the content scales for theBASC-2-PRS were constructed via theoreticaland empirical methods. Thesescales were developed based on currenttheoretical perspectives about importantdomains of youth functioning (Reynolds &Kamphaus, 2004). There exists very littleresearch on these scales, yet their labels anditem content are intriguing and warrantfurther investigation of their reliability, validity,and clinical utility. Initial analyses indicatethat the PRS content scales possess adequate(i.e., 0.70 and higher) internal consistencies(Reynolds & Kamphaus, 2004).Administration and ScoringThe PRS uses a four-choice response format(i.e., never, sometimes, often, almost always)with no space allowed for parent elaboration.According to the authors, the scale takesabout 10–20 min for parents to complete.A variety of derived scores and interpretivedevices are offered. Linear T-scoresare available for all scales and composites,meaning that the original distributions forthese indices in the norming sample weremaintained. Other scores available includepercentile ranks, confidence bands, andstatistical methods for identifying high andlow points in a profile. Both hand-scoringand computer entry scoring are availablefor the PRS.NormingPRS provides three norm-referenced comparisonsdepending on the questions ofinterest to the clinician. Some examples ofquestions and their implications for normIs Daniel inattentive incomparison to childrenof the same age?Is Daniel inattentive incomparison to a large samplesample of children who arecurrently diagnosed andreceiving treatment?Is Daniel inattentive incomparison to boys of thesame age?group selection include the following.Question Norm GroupGeneral nationalsampleClinical nationalsampleMale nationalThese various norm-referenced comparisonsare more than are typically offeredfor such scales. The general national normingsample is advised as the starting pointfor most purposes (Reynolds & Kamphaus,2004). Of course, as just noted, dependingon the question, the clinician may opt forgender-specific or clinical norms. Genderspecificnorms may be useful in trying toconvey a child’s current level of functioningto others, such as parents. In otherwords, one might present the child’s scoresrelative to the general population andthen emphasize how the child comparesto other boys/girls on areas of concern inorder to provide a more complete picture.However, too much information may causeconfusion for some parents. Gender-basednorms may also help answer some specificresearch questions (i.e., correlates of inattentionand hyperactivity among girls andamong boys).The general norm sample for the PRSincluded 1,200 preschoolers, 1,800 children,and 1,800 adolescents. Cases werecollected at test sites in 40 states. Across agegroups, the PRS sample closely matchesUS Census statistics (Current PopulationSurvey, 2001) in terms of sex, race/ethnic-

152 chapter 7 Parent and Teacher Rating Scalesity, and socioeconomic status (SES). Thenorming sample also represents a good fit tocensus data on geographic region (seeReynolds & Kamphaus, 2004 for moredetails). The clinical norming sample forthe PRS included responses from 1,975parents, with most cases being classified ashaving a learning disability or ADHD.ReliabilityThe median reliability coefficients providedin the manual suggest good evidence forthe reliability of the individual scales andcomposites. All scales and composites havemedian reliability estimates of 0.80 andabove, with the exception of the Activitiesof Daily Living and Atypicality Scales. TheBASC-2 manual also provides informationon 1–7-week test-retest reliability and interraterreliability between parents. Test-retestreliability coefficients were 0.70 and higher,with the exception of Depression for thepreschool form which was .66. Interraterreliability was generally good, with coefficientsin the same range, with the exceptionof Aggression on the preschool and childforms (i.e., 0.59 and 0.58, respectively) andAnxiety on the preschool form (.56).ValidityThe PRS appears to have a broad contentcoverage. The PRS assesses a variety ofexternalizing behavior problems (McMahon& Frick, 2002) and has an expandedassessment of adaptive skills. In addition,the PRS enjoys considerable factoranalytic support for a three factormodel consisting of externalizing problems,internalizing problems, and adaptiveskills. The strongest measures of theexternalizing factor are the Aggression,Conduct Problems and Hyperactivity.The Internalizing factor is marked byloadings by the Atypicality, Depression,Anxiety, Somatization, and Withdrawalscales. Adaptive skills scales that loadhighly on this factor include Activities ofDaily Living, Functional Communication,Leadership, and Social Skills.Some of the secondary loadings for thescales may also have implications for interpretation.Specifically, the factor-analyticdata suggest that the following profiles arereasonable:l Poor Adaptive Skills with AttentionProblemsl Good Adaptive Skills with Anxietyl Internalizing Problems accompanied byPoor AdaptabilityCriterion-related validity analyses producedconsistent associations between the PRS andother parent rating scales. This pattern particularlyholds for the composites and for theexternalizing problem scales. Generally, theinternalizing problem scales (e.g., Anxiety,Depression, Somatization) show moderatecorrelations with analogous scales from othermeasures (see Reynolds & Kamphaus, 2004).InterpretationThe same logical interpretive steps that wereoutlined for the BASC-SRP (discussed inChap. 6) also apply to the BASC-PRS. Briefly,the clinician should:1. Assess validity using validity indexes andinformal means (e.g., inspect for a highnumber of items with no response).2. Inspect critical items and follow-up asappropriate.3. Interpret scores on scales and composites,with particular attention to elevations(T-scores of 65–70 and higher) onclinical scales and low scores (T-scores of35 and below) on adaptive scales.4. Attend to items that appeared to haveled to scale elevations (or low adaptivescores).5. Integrate score with information fromother informants.

chapter 7 Parent and Teacher Rating Scales1536. Integrate data with information fromother assessment tools (e.g., interview,behavioral observations, intelligencetesting).7. Set objectives for treatment/interventionAs was the case with the SRP, we again recommenda focus on interpretation at the scale level, as thereliabilities of the PRS scales are generally good,and elevations on scales are more specificallyinformative than would be the case for elevatedcomposite scores.The original PRS enjoyed a great deal ofresearch support and research use. There isquite limited information available to dateon the BASC-2-PRS. Nevertheless, thecombined rational and empirical approachto scale development has intuitive appealfor use in clinical situations. Clinicians arestill urged to keep abreast of the researchliterature discussing the strengths and limitationsof any assessment tool.Strengths and WeaknessesThe BASC-2-PRS has a number of apparentstrengths and weaknesses as follows:The strengths of the PRS are:1. Good psychometric properties based onthe information reported in the BASC-2manual.2. A variety of scales that may be usefulfor differential diagnosis (e.g., AttentionProblems vs. Hyperactivity, and Anxietyvs. Depression).3. The availability of validity scales andcritical items.4. An expanded group of norm-referencedadaptive scalesAmong the weaknesses of the PRS are:1. A response format that does not allowparents to provide additional detail abouttheir responses2. Cross-informant and cross-scale comparisonsnot as readily made as on othermeasures, as different forms (e.g., parentvs. self-report) include differentitem content and scales3. Limited research on the latest edition ofthe PRS.Teacher Rating Scale (TRS)The BASC-2 Teacher Rating Scale (BASC-2-TRS; Reynolds & Kamphaus, 2004)allows the clinician to gather informationon a child’s observable behavior from thechild’s teacher and place that information inthe context of other information obtained inthe overall BASC system (e.g., self-reportscale, parent rating scales, classroom observationsystem). As with the PRS, there arethree forms of the BASC-2-TRS: preschool(ages 2–5), child (6–11), and adolescent(12–21). The three forms contain behavioraldescriptors that are rated by the teacher on afour-point scale of frequency, ranging from“Never” to “Almost Always.” The threeforms have 100 items for the preschool versionand 139 for both the child and adolescentversions.ContentAs with the other BASC-2 rating scales, theitems of the BASC-2-TRS were chosen tomeasure multiple aspects of a child’s personalityand behavior. The TRS includesboth positive (adaptive) and pathological(clinical) dimensions. For the most part,the BASC-2-TRS has maintained the contentareas of the original BASC. The onlyscale additions to the current version of theTRS were the Functional Communicationscale for all age groups and the Adaptabilityscale for the adolescent version.The BASC-2-TRS consists of five composites(i.e., Behavioral Symptoms Index,School Problems, Externalizing Problems,Internalizing Problems, Adaptive Skills)

154 chapter 7 Parent and Teacher Rating Scalesacross all age ranges, with 11 scales inthe preschool version and 15 scales in thechild and adolescent versions. The scalesgrouped into the composites – except forthe Behavioral Symptoms Index whichincludes the Hyperactivity, Atypicality,Depression, Aggression, Attention Problems,and Withdrawal scales – are providedin Table 7.4. The TRS also has the sameoptional content scales as those providedfor the PRS (see above). Because these area new feature of the BASC-2, very limitedinformation is available on their psychometricproperties or clinical utility.The content coverage of the BASC-2-TRS scales has several unique features relativeto other teacher rating scales. First,it provides comprehensive coverage ofseveral areas of adaptive behavior. Second,the current version of the TRS continuesthe strategy of including separate scales formotor hyperactivity and attention problems,which aids in the differentiation ofsubtypes of Attention-Deficit HyperactivityDisorder (Vaughn, Riccio, Hynd, &Hall, 1997). Third, there are separate scalesfor anxiety, depression, and withdrawal,which aid in the assessment of emotionaldifficulties. Fourth, the BASC-2-TRSTable 7.4 Composites and Scales of BASC-2-TRSCompositeExternalizing ProblemsInternalizing ProblemsSchool ProblemsAdaptive SkillsScalesAggression HyperactivityConductProblemsAnxiety DepressionSomatizationAttention ProblemsLearning ProblemsAdaptability FunctionalCommunicationLeadershipSocial SkillsStudy Skillsincludes items that screen for learningproblems that often accompany emotionaland behavioral problems in children.Administration and ScoringThe BASC-TRS takes approximately10–20 min to complete. The cover of therecord provides instructions to the teacherfor completing the form and space forrecording background information aboutthe child and teacher (e.g., age, gender,type of class, length of time in class). Bothhand scoring and computer scoring areavailable. Norm tables in the BASC manualare provided so that any of four sets ofnorms can be used: general, male, female,and clinical (see above for discussion ofthe uses of these different types of norms).Both T-scores and percentile ranks arelisted for each set of norms, with linearT-scores again being utilized for the TRS.As with the other BASC-2 rating scales,the BASC-2-TRS scoring sheet highlightscritical items (e.g., “I want to kill myself’)that are clinically important or that warrantfurther follow-up.NormingThe norming group included 1,050 preschoolers,1,800 children (ages 6–11), and1,800 adolescents (ages 12–21) with equalsex distributions in all age groups. Respondentswere recruited from sites throughoutthe USA. As described previously, the samplingprocedures for obtaining the normativesample were designed to closely mirrorUS Census statistics in terms of race/ethnicity,SES, and geographic region, andthis goal was accomplished (see Reynolds& Kamphaus, 2004). Details regarding the1,779-member clinical sample for the TRSare also provided in the BASC-2 Manual.ReliabilityThe manual for the BASC-2-TRS (Reynolds& Kamphaus, 2004) provides evidence

chapter 7 Parent and Teacher Rating Scales155on three types of reliability: internal consistency,test-retest reliability, and interraterreliability. With very few exceptions,the scales of the BASC-2-TRS proved tobe quite reliable in the normative sample.More specifically, internal consistency coefficientstended to average well above 0.80across all age groups, and all were 0.75 orhigher. Similarly, test-retest reliability overone to nine weeks was high, with the exceptionof the Anxiety scale, with coefficientsranging from 0.64 for the adolescent versionto 0.77 for the adolescent version. Still,these coefficients are adequate. Finally, theconsistency of ratings between two teacherswas tested in samples of preschool-agechildren (n = 74), school-age children (n =38), and adolescents (n = 58), with moderatereliability estimates emerging acrossage group samples (median coefficients of0.69, 0.60, and 0.52, respectively). Correlationcoefficients tend to be somewhathigher for externalizing than for internalizingproblems consistent with pastresearch (Achenbach, McConaughy, &Howell, 1987). It is also worth noting thatthe coefficients tended to be lower foradolescents, which may be associated withthe limited contact an individual teachermay have with students of that age group.Interrater agreement for Somatization(r = 0.25), Withdrawal (r = 0.24), and Atypicality(r = 0.31) was particularly low forteacher ratings of adolescents.ValidityThe TRS is closely, but not exactly, alignedwith the item content of the PRS. However,the TRS has additional scales (i.e., StudySkills, Learning Problems) that seem particularlyvalid for use with a teacher ratingscale. The BASC-2 manual provides factoranalytic support for the construct validityof the scales and composites of the TRS.In addition, initial research on the TRSshows generally high correlations withanalogous scales from other teacher ratingscales. However, the correspondenceto analogous scales is somewhat lower forinternalizing types of problems than forthe indices of externalizing problems (seeReynolds & Kamphaus, 2004). One notablefinding was the lack of a correlation (i.e.,r = .03) between the TRS Somatizationscale and the Somatic Complaints scale ofthe Achenbach Teacher Report Form. Asignificant limitation of the latest versionof the TRS is the very limited research onits validity and utility outside of what wasconducted by the developers.InterpretationThe BASC-2-TRS includes validity scalesthat provide a useful and efficient firstpoint of interpretation. More specifically,it contains a “fake bad” index (F), whichhelps to assess the possibility that a teacherrated a child in an overly negative pattern.Therefore, interpretation of this scale, inparticular, should be the first step in theinterpretative process, keeping in mind that ahigh score on the F index may actually indicatesignificantly problematic functioning.Therefore, this validity index shouldbe interpreted in the context of otherassessment data. The Consistency Indexand the Response Pattern Index availablefor the TRS (as are available for the PRSand SRP) provide another initial point ofinterpretation. Critical items should bereviewed promptly, because these itemstend to be clinically important indicatorsthat deserve careful follow-up assessment.The reliability estimates at the scalelevel of the TRS are good; therefore, weagain recommend focusing interpretationmainly at the scale, rather than composite,level since more specific information isavailable through the TRS scales. Interpretationsat the item level must be made quitecautiously because of the low reliability ofindividual items. It is often informative tosee which items led to a child’s or adolescent’selevation on a given clinical scales.For example, it may be informative for achild with an elevation on the Adaptabil-

156 chapter 7 Parent and Teacher Rating Scalesity scale to determine if this elevation waslargely due to problems specifically withinthe interpersonal domains or due to moregeneral problems in adapting to changes inroutine. Finally, interpretation at the scalelevel for the parent form is a viable earlystep in interpretation (see above); therefore,interpretation at the scale level of theTRS facilitates integration of informationacross parent and teacher ratings. In addition,considering individual items withinelevated scales on both rating forms mayhelp determine the source of consistenciesand inconsistencies across parent andteacher ratings, further informing caseconceptualization and recommendations.Strengths and WeaknessesLike its predecessor and its companion parentrating scale, the BASC-2-TRS has anumber of strengths and weaknesses. Notablestrengths include:1. It is part of a multimethod, multi-informantsystem that aids in a comprehensiveclinical evaluation with item content thatcovers important problematic and adaptivedomains of classroom behavioral andemotional functioning.2. The assessment of adaptive functioningis enhanced on this version of the TRS.3. The preschool age range of the BASC-2-TRS is expanded from the age rangeavailable from the original TRS.4. The BASC-2-TRS has a large nationwidenormative sample on which normreferencedscores are based, allowing forone to confidently make many normreferencedinterpretations of scores.Weaknesses of the BASC-2-TRS include:1. The limited research base for the currentedition.2. The relatively lower correlationsbetween internalizing scales on the TRFand analogous scales from other teacherrating scales.3. The different item content across informants,especially with the SRP, makesintegration of BASC-2 informationsomewhat more challenging.A sample case using the BASC-2-PRS andBASC-2 TRS is provided in Box 7.1.Achenbach System of EmpiricallyBased Assessment(Achenbach & Rescorla,2000, 2001)Parent Report: Child BehaviorChecklist (CBCL)The Child Behavior Checklist (CBCL;Achenbach, 2001) and its predecessors havea long history of prominence in child assessment.The CBCL scale is the product of anextensive multiple-decade research effort,and it has a distinguished history of researchusage. The current version of the CBCL ismuch like its predecessors with some itemchanges, response format changes, and theintroduction of DSM-Oriented scales (seebelow; see Achenbach & Rescorla, 2001).The CBCL is part of an extensive systemof scales including teacher rating (TRF),self-report (YSR), and classroom observationmeasures. The newest version of the CBCL(Achenbach & Rescorla, 2001) has two separateforms: one for ages 1½–5 and the otherfor children of ages 6–18.The development of the CBCL and itsrevisions reflects the author’s belief thatparent reports are an important part of anymulti-informant system of child evaluation.In Achenbach’s (1991) own words:Parents (and parent surrogates) are typicallyamong the most important sources of data aboutchildren’s competencies and problems. Theyare usually the most knowledgeable about their

chapter 7 Parent and Teacher Rating Scales157Box 7.1Sample Case Using the BASC-2 PRS and TRSJohnny was referred for a psychological evaluationby his mother because of his academicdifficulties and distractibility. He is a 12-yearoldsixth-grader. Johnny has reportedly hadtrouble completing schoolwork since the firstgrade. Johnny was retained in the second gradebecause of poor work completion and academicprogress. He was reportedly placed inspecial education in the third grade in a learningproblem class. He was reportedly placedin a regular class again in the fourth grade. Apsychological evaluation conducted at the endof the fourth grade resulted in the conclusionthat he was a “slow learner,” according to hismother. This year, he is again having difficultyconcentrating, and he rarely completes hisassignments. Academic progress is still unacceptableto his mother and teachers.Johnny’s developmental milestones wereslightly delayed. He still reportedly has problemsdrawing and using scissors. Last year, hewas diagnosed with juvenile diabetes, accordingto his mother. His family psychiatric historyis significant for depression (mother), andJohnny’s father reportedly had difficulty academicallywhen he was in school.Johnny was exceedingly cooperative duringthe evaluation. He addressed the examinerpolitely and would occasionally answer questionsby saying “yes sir.” He had considerable difficultycomprehending instructions on an intelligencetest. He was reluctant to admit to not knowingan answer, and he worked extremely slowly. Thetest session had to be conducted over two daysbecause of his slow response style.He responded impulsively to items on occasion.He also wiggled in his seat and frequentlylooked around the room and asked questionsof the examiner. His full scale IQ score in thisevaluation was 85. His achievement test scoreswere similarly in the Low Average range.His BASC-2-PRS (completed by hismother) and BASC-2-TRS (completed by hiscurrent teacher) results are highly consistentwith background information.HyperactivityT = 63 (parent report)T = 60 (teacher report)The reports of relatively mild levels of hyperactivityare consistent with Johnny’s historywhich indicates no history of disruptive orimpulsive behaviors.Attention ProblemsT = 76 (parent)T = 77 (teacher)The parent and teacher reports of significantattention problems (e.g., has trouble concentrating,is easily distracted, daydreams) isconsistent with reports of Johnny’s difficultiesdating back to the first grade.SomatizationT = 67 (parent)T = 63 (teacher)Mild to moderate concerns in the area of Somatizationappear to be related to Johnny’s historyof diabetes and its attendant difficulties.Learning ProblemsT = 72 (teacher)Johnny’s teacher reported significant learningproblems for Johnny, indicating concerns inall academic areas. It was recommended thathe be further evaluated through a Responseto Intervention (RTI) procedure at his school,and services available to him should be plannedaccordingly.AdaptabilityT = 39 (parent)T = 32 (teacher)Functional CommunicationT = 30 (parent)T = 32 (teacher)Johnny’s mother and teacher reported concernswith his ability to adjust to changes inplans and to adequately communicate his needfor help. These difficulties may interfere withhis academic performance.Johnny’s results were strikingly similarfor both teachers and parents. The diagnosisof ADHD, Predominantly InattentiveType was made based on Johnny’s history asreported by his mother and teacher duringinterviews and based on these scores on thePRS and TRS.

158 chapter 7 Parent and Teacher Rating Scaleschild’s behavior across time and situations. Furthermore,parent involvement is required in theevaluation of most children, and parents’ viewsof their children’s behavior are often crucial indetermining what will be done about the behavior.Parents’ reports should therefore be obtainedin the assessment of children’s competencies andproblems whenever possible (p. 3).Table 7.5 Sample Content of CBCL(6–18-Year-Old Version) Syndrome ScalesAnxious/Depressed: Cries a lot, is fearful,feels too guilty, talks of suicideWithdrawn/Depressed: Would rather bealone, shy/timid, sadSomatic Complaints: Feels dizzy, constipated,has headaches, nauseaSocial Problems: Dependent, lonely, getsteased, prefers to be with younger kidsThought Problems: Cannot get mind off ofcertain thoughts, sees things, stores things,strange behaviorAttention Problems: Cannot concentrate,daydreams, impulsive, cannott sit stillRule-breaking Behavior: Drinks alcohol,lacks guilt, breaks rules, sets fires, prefers to bewith older kidsAggressive Behavior: Argues a lot, destroysothers’ things, gets in fights, mood changes,attacks peopleFrom Achenbach & Rescorla (2001).ContentThe CBCL includes 100 items for thepreschool version and 113 items for theschool age version. Responses are made ona three-point scale (i.e., Not True; Sometimes/SomewhatTrue; Very True/OftenTrue).The CBCL syndrome scales are primarilyempirically derived, with substantialuse of factor-analytic methods. TheCBCL scales were also derived separatelyby gender and age group (see Achenbach &Rescorla, 2000, 2001). Throughout the testdevelopment process, the CBCL developersalso emphasized the derivation of scalesthat were common across raters (e.g., parentsand teachers). The CBCL parent andteacher scales have closely matched itemsand scales that make it easier for cliniciansto make cross-informant comparisons.Sample item content from the CBCL scalesis shown in Table 7.5.The item content for the preschool(ages 1½–5) version of the CBCL is notablydifferent from the version for 6–18-yearolds, with somewhat different syndromescales. The syndrome scales for the 1½–5-year-old version are Emotionally Reactive,Anxious/Depressed, Somatic Complaints,Withdrawn, Sleep Problems, AttentionProblems, and Aggressive Behavior(Achenbach & Rescorla, 2000).On both versions, there is a Total Problemsscore (the most global score availableon the CBCL) as well as composites forInternalizing Problems and ExternalizingProblems (see Achenbach & Rescorla,2000, 2001). The CBCL also includes competencescales (except for the preschoolversion) that are designed to discriminatesignificantly between children referred formental health services and non-referredchildren (Achenbach & Rescorla, 2001).DSM-Oriented scales were formedbased on experts’ ratings (see Achenbach,Dumenci, & Rescorla, 2001) of how wellthe items fit DSM criteria for relevant disordersor groups of disorders (e.g., MajorDepression and Dysthymia for the AffectiveProblems scale). For the school-ageversion of the CBCL, the DSM-Orientedscales are Affective Problems, Anxiety Problems,Somatic Problems, Attention/HyperactivityProblems, Oppositional DefiantProblems, Conduct Problems. The fiveDSM-oriented scales on the preschool versionof the CBCL are Affective Problems,Anxiety Problems, Pervasive DevelopmentalProblems, Attention Deficit/HyperactivityProblems, and Oppositional DefiantProblems. The DSM-Oriented scales are anew feature to the Achenbach system and

chapter 7 Parent and Teacher Rating Scales159were designed to more closely align scoresthat were available from these instrumentsto current diagnostic nomenclature.Administration and ScoringThe CBCL is easily administered in 15–20min. The CBCL is somewhat unique inthat adaptive behavior is assessed with acombined fill-in-the-blank and Likert scaleresponse format. In addition, some of theproblem behavior items require the parentto elaborate on or describe the problemendorsed. This format is advantageous inthat it allows the parent to respond in anopen-ended format. Clinicians can gainaccess to qualitative information of valueusing this format. Open-endedness, however,also has a disadvantage: It may extendadministration time and requires moredecision making on the part of the parent.Hand scoring and computer scoringare available for the CBCL. The CBCLoffers normalized T-scores as the featuredinterpretive scores. Percentile ranks arealso provided. T-scores are available for allscales and three composites: Externalizing,Internalizing, and Total. T-scores are nowalso offered for the Competence scales.The advantages and disadvantages ofusing normalized versus linear T’s aredebatable (see Kline, 1995). On the onehand, the advantage of comparable percentileranks across scales was recognizedby the MMPI-A author team (see Chap.6). Normalized scores, however, clearlychange the shape of the many skewed rawscore distributions forcing the T-score distributionto take a shape that it does notactually take in the general population(see Chap. 2). In addition, the reporting ofT-scores on the CBCL is truncated for theSyndrome and DSM-Oriented scales suchthat low scores are reported simply as T ≤50. For the Competence scales, the distributionis truncated above a T-score of 65and below a T-score of 35.NormingThe norming of the school age CBCL isbased on a national sample of 1,753 childrenaged 6 through 18 years. This samplewas collected in 40 states and the Districtof Columbia (Achenbach & Rescorla,2001). Relevant stratification variablessuch as age, gender, ethnicity, region, andSES were recorded in an attempt to closelymatch US Census statistics on these variables.The respondents were mothers in72% of the cases and fathers in 23% of thecases (5% of the cases used “others”). Sixtypercent of the respondents were classifiedas White, with Hispanics appearing to besomewhat underrepresented (9%). Fiftyonepercent of cases were from a middleSES background, 33% were from an upperSES background, and 16% were from alower SES background. Forty percent ofrespondents were from the southern partof the USA (see Achenbach & Rescorla,2001). From this sample, separate normswere developed for ages 6–11 and 12–18,with each of these groups further delineatedby gender.The norming sample for the preschoolCBCL version for ages 1½–5 was alsorecruited in an attempt to match US Censusstatistics on the same variables. Thissample consisted of 700 respondents (76%mothers, 22% fathers, 2% “others”). Fiftysixpercent of respondents were White,21% African American, 13% Latino, and10% Mixed or Others. In the preschoolnorming sample, 33% of respondentswere from an upper SES background, 49%from middle SES, and 17% from a lowerSES background. Again, 40% of theserespondents were from the southern USA(Achenbach & Rescorla, 2000).Children were excluded from the sampleif they had “received mental health orspecial education classes during the previous12 months” (Achenbach & Rescorla,2001, p. 76). Separate clinical norms arenot offered for the CBCL.

160 chapter 7 Parent and Teacher Rating ScalesReliabilityThe CBCL has good evidence of reliabilitywith internal consistency coefficients rangingfrom 0.78 to 0.97 on the Syndrome scales,0.72 to 0.91 on the DSM-oriented scales, andsomewhat lower internal consistency on theCompetence scales (i.e., 0.63 to 0.79; Achenbach& Rescorla, 2001). On the preschoolversion of the CBCL, the internal consistencycoefficients for the Syndrome scalesand composites ranged from 0.66 to 0.95. Forthe DSM-Oriented scales, internal consistenciesranged from 0.63 to 0.86 (Achenbach &Rescorla, 2000). The data from a test-reteststudy for a sample of 73 (mean interval = 8days) children yielded coefficients rangingfrom 0.80 for the Anxiety DSM-Orientedscale to 0.93 for the DSM-Oriented ConductProblems scale on the 6–18-year-oldversion (Achenbach & Rescorla, 2001). Forthe preschool CBCL (n = 68), 8-day testretestreliabilities were good, ranging from0.74 for the Attention-Deficit/HyperactivityProblems DSM-Oriented scale to 0.92for the Sleep Problems scale (Achenbach &Rescorla, 2000).A two-year stability study of 67 childrenyielded coefficients ranging from 0.45 forSomatic Problems to 0.81 for AggressiveBehavior for the school age CBCL.Twelve-month test-retest coefficients forthe preschool version (n = 80) ranged from0.52 for two of the DSM-Oriented scales to0.62 for the Anxious/Depressed SyndromeScale. These coefficients are indicative ofstrong reliability in light of the lengthyinterval and the expected natural instabilityin some of these areas over time. Lastly,mother–father interrater agreement on theCBCL was generally good on the schoolage version with all coefficients except theActivities scale being 0.63 or higher. Theinterrater agreement on the preschoolversion was lower, with coefficients rangingfrom 0.48 to 0.67 (see Achenbach &Rescorla, 2000).ValidityMuch of the validity evidence reported bythe authors of the CBCL focuses on theability of the scale to differentiate clinicalfrom nonclincal samples. Results of theseanalyses indicated good differential validityacross scales for both boys and girls. Asnoted in Chap. 2, however, evidence of differentialvalidity must now also show differentiationbetween clinical samples. Todate, such evidence is lacking for the latestversions of the CBCL.The factor structure of the CBCLcontinues to raise some conceptual issuesregarding the content validity of the scales.For example, it is unusual for depressionand anxiety items to be included on the samescale. In addition, the Attention Problemsscale includes items that appear more indicativeof hyperactivity and impulsivity thaninattention (see Table 7.5). High scores onthese scales still require a great deal of clinicaljudgment as to what characteristics ledto the high scores and should be the focusof further attention.Validity studies as well as basic andapplied research investigations using theprevious versions of the CBCL are legion.Although the research base of the currentCBCL is not as well-established, some evidenceon its validity is promising. For example,the preschool version of the CBCL hasbeen found to be useful in screening forAutism Spectrum Disorders based on theWithdrawal and Pervasive DevelopmentalProblems scales (Sikora et al., 2008), andthe CBCL has been touted for its abilityto screen for a variety of problem areas andits strong convergent and divergent validity(Scholte, Van Berckelaer-Onnes, & Van derPloeg, 2008). However, the correspondenceof the DSM-Oriented Anxiety scale on theCBCL to DSM criteria for anxiety disordershas been called into question (Ferdinand,2008). Clearly, more research is needed onthe latest rating scales in the Achenbachsystem, but just as clearly, the CBCL has

chapter 7 Parent and Teacher Rating Scales161enjoyed and continues to enjoy a great dealof empirical support.InterpretationInterpretation of the CBCL is bolstered bymany articles by Achenbach and colleaguesdevoted to its clinical use dating to McConaughyand Achenbach’s (1989) informativework on this subject. The CBCL useris fortunate to have many interpretiveresources available.More specifically, McConaughy andAchenbach (1989) provide an assessmentmethodology for the identification ofsevere emotional disturbance in the schools.Their multi-axial empirically based assessmentmodel proposes five axes for suchassessment situations: (1) parent reports(Achenbach, CBCL), (2) teacher reports(Achenbach Teacher Report Form), (3)cognitive assessment, (4) physical assessment,and (5) direct assessment of the child(i.e., Achenbach Direct Observation Formand Youth Self-Report). McConaughy andAchenbach assist the psychologist workingin schools further by linking each CBCLscale to the accepted criteria for severeemotional disturbance. High scores on theAnxious/Depressed scale may, for example,indicate the presence of a general pervasivemood of unhappiness, which, in turn, mayqualify a child as severely emotionally disturbedand document eligibility for specialeducation and related services. Of course,these examples fit best with previous versionsof the CBCL and are only as useful astheir degree of correspondence with eligibilitycategories used by school systems.For interpreting the CBCL specifically,because there are no established validityscales and because there are some scalesthat include heterogeneous content, werecommend more attention to item-levelinterpretation than we have for other measures.That is, the clinician should pay closeattention to scale elevations and draw mostconclusions at the scale level; however, itwould behoove practitioners to determineany concerning aspects of the parent’s itemresponse style that would render the protocolinvalid. Additionally, interpretationshould not stop at the scale level. Rather,one should inspect the items that led to thescale elevations to determine the best wayto describe the child’s difficulties. Fortunately,the scoring methods available forthe CBCL make linking item responses toscale elevations a straightforward process.Strengths and WeaknessesThe CBCL has many strengths that continueto make it a popular choice for clinicians.Noteworthy strengths include:1. An ever-growing research base on thecurrent CBCL, as well as a wealth ofvalidity research on its predecessors.2. Its popularity among professionalswhich facilitates communication aboutits results3. Several writings that provide interpretiveguidance above and beyond thatprovided by the manual4. Improved approach to assessing competenceand available of new DSM-Oriented scales that are aligned to DSMcriteria.5. Some response flexibility in that parentsare asked to elaborate on their answersto some items.Weaknesses of the CBCL include:1. Lack of validity scales which are nowcommon among behavioral ratingscales.2. Lack of close correspondence betweenthe empirically derived scales and commondiagnostic criteria (e.g., DSM;Hart & Lahey, 1998)3. Heterogeneous content within somescales.

162 chapter 7 Parent and Teacher Rating ScalesThe CBCL continues to be a preferredchoice of many child clinicians because ofits history of successful use and popularitywith researchers. The continuing developmentof the CBCL database bodes wellfor its future. The most recent versions ofthe CBCL would benefit from researchaimed at assessing the construct validity ofits scales, particularly the DSM-Orientedscales which were not part of the previousCBCL. Such research efforts are necessaryto define further the degree of confidencethat a clinician can place on specific scalesfor making differential diagnostic decisions.Teacher Report: Teacher ReportForm (TRF)The Achenbach Teacher Report Form(TRF; Achenbach & Rescorla, 2001) isdesigned to be completed by teachers ofchildren between the ages of 6 and 18 forthe school-age version and between 1½ and5 for the preschool version which is labeledthe “Caregiver-Teacher Report Form”(Achenbach & Rescorla, 2000). The itemcontent of the TRF is very closely matchesthe CBCL item content.ContentThe school age version of the TRF includesseveral background questions (e.g., “Howlong have you known this pupil?” “Howwell do you know him/her?”), a teacher’srating of a child’s academic performance,and a four-item screening of a child’s adaptivebehavior with scoring on a 1–7 scale(e.g., “How hard is he/she working?” “Howappropriately is he/she behaving?” “Howmuch is he/she learning?” “How happy ishe/she?”). The preschool TRF includes thesame background questions but does notinclude the other items. These backgroundquestions have associated norms and fallunder the “Adaptive Functioning” domainof the TRF. The major portion of the TRFconsists of 100 items for the preschoolversion and 113 items for the school ageversion. These items describe problematicbehaviors and emotions that the teacherrates as being Not True, Somewhat True/Sometimes True, or Very True/Often Trueof the child. The problem behavior itemscover a broad array of both internalizing(e.g., anxiety, depression, somatic complaints)and externalizing (antisocial behavior,aggression, oppositionality) behaviors.As with the CBCL, the TRF now includesDSM-Oriented scales (6 for the school ageversion; 5 for the preschool version). Theonly scale difference is that the TRF doesnot include a Sleep Problems scale.Administration and ScoringThe TRF takes approximately 15–20 minto complete. The instructions to the teacherare printed on the front of the answer sheet.Scoring of the TRS can be done by handusing the TRF Profile Sheets with separateprofile sheets available for boys and girls.However, a computer-scoring system isavailable that greatly facilitates scoring byautomatically calculating raw scale scoresand converting them to norm-referencedscores appropriate for the child’s age andgender.Both the Profile Sheets and the computer-scoringprogram provide raw scoresand norm-referenced scores for severalscales. As with the CBCL, a Total Problemscore, which is an overall indicator ofa child’s classroom adjustment, and twobroadband scores consisting of Internalizingand Externalizing behaviors areincluded. These broad dimensions arefurther divided into the eight Syndromescales.The TRF allows for raw scores on allscales to be converted to T-scores andpercentile ranks based on the standardizationsample. The T-scores are normalizedstandard scores. That is, the raw scoredistributions are transformed to a normalizeddistribution. This procedure allowsT-scores on all scales to have similar dis-

chapter 7 Parent and Teacher Rating Scales163tributions and corresponding percentilesbased on the assumptions of a normaldistribution. However, as noted previouslyfor the CBCL, this transformationassumes that the dimensions assessed bythe scale should be normally distributed inthe general population, an assumption thatis questionable because most children tendto cluster in the normal end of the distribution.Norm-referenced scores are basedon gender and age-specific norms. In addition,as with the CBCL, T-scores on theTRF are truncated, such that the lowestscore provided is ≤ 50.NormingThe norming sample for the school-ageversion of the TRF consisted of 2,319 children,72% of whom were White. Fourteenpercent were identified as African American,and 7% were identified as Latino;thus, both ethnic minority groups appearto be underrepresented in this sample. TheTRF normative sample appears to be geographicallyrepresentative. Thirty-eightpercent of children in this sample werefrom an upper SES background, 46% froma middle SES background, and 16% froma low SES background.For the preschool version, the normingsample consisted of 1,192 children. Thissample was geographically diverse, butonly 10% of the sample came from a lowerSES background. The sample representedWhites and African Americans well (i.e.,48% and 36%, respectively), with only 8%of the sample identifying as Latino.For both versions of the TRF, the normativesample excluded children who hadreceived mental health or special educationservices within the preceding 12 months.Therefore, as with the other rating scales inthe Achenbach system, the sample shouldbe considered a normal comparison group,rather than one that is normative and representativeof the general population.ReliabilityAchenbach and Rescorla (2000, 2001) providethree types of reliability information onthe TRF. Internal consistency estimates wereprovided for the Syndrome and DSM-Orientedscales. Coefficients indicated goodinternal consistency, ranging from 0.72 to0.95. For the preschool version, coefficientswere quite variable, ranging from 0.52 forSomatic Complaints to 0.96 for the AggressiveBehaviors scale. Test-retest reliabilityover an average of a 16-day interval is presentedon a sample of 44 children in the agerange of 6–18. The test-retest coefficientswere generally high (i.e., 0.80s and higher),with the exception of the Withdrawn/Depressed scale (r = 0.60) and the AffectiveProblems scale (r = 0.62). Four-month testretestreliability was variable, ranging from0.31 (Affective Problems) to 0.72 (Hyperactivity-Impulsivity).The 8-day test-retestreliability for the preschool version of theTRF was somewhat variable, ranging froma coefficient of 0.57 for Anxiety Problems to0.91 for Somatic Complaints (the scale withthe lowest internal consistency). Threemonthtest-retest reliability for the preschoolTRF was variable with coefficientsranging from 0.22 for Somatic Complaintsto 0.71 for Emotionally Reactive.Correlations between ratings from twodifferent teachers are provided for 88 children(Achenbach & Rescorla, 2001). Thecorrelations were modest across scales, withthe mean coefficient for the full sample being0.49 for the Competence scales, 0.60 for theSyndrome scales, and .58 for the DSM-Oriented scales. Similar analyses for 102preschoolers revealed an overall mean coefficientof 0.62, with a range of 0.21 (SomaticComplaints) to 0.78 (Aggressive Behavior).ValidityThere is relatively limited validity informationavailable for the current versionsof the TRF. The validity data reported inthe manuals (Achenbach & Rescorla, 2000,

164 chapter 7 Parent and Teacher Rating Scales2001) mainly focus on the ability of thescales to differentiate non-referred fromclinical samples within gender. The TRFscales generally show such differential validity.However, the Somatic Complaints scaleof both versions does not appear to consistentlydifferentiate the two groups (Achenbach& Rescorla, 2000, 2001). As notedabove, validity studies for the BASC-2-TRSdemonstrated good correspondence withthe TRF, especially for externalizing problems.An exception was the lack of correlationbetween scales on each form assessingsomatic complaints. Research has supportedthe predictive validity of the TRF at agethree in predicting problems, especiallyexternalizing problems, at age five (Kerret al., 2007). The TRF, like the CBCL,was based heavily on its well-researchedpredecessor. There is a wealth of researchon the previous versions of the TRF whichsupports its validity in (1) differentiatingclinic-referred children from non-referredchildren, (2) correlating with classroomobservations of children’s behavior, and (3)correlating with independent clinical diagnoses(see Achenbach, 1991; Casat, Norton,& Boyle-Whitesel, 1999; Piacentini, 1993).InterpretationInformation on the reliability and validityof the adaptive functioning component ofthe TRF is lacking; therefore, interpretationsof these scales should be done verycautiously, if at all. Subsequently, althoughit may be useful to next consider the TRFTotal Problems score and composites,more specific information can be gleanedfrom interpretations of the eight syndromescales and the six DSM-Orientedscales. The reliability of these scales (withthe exception of the Somatic Complaintsscale) is good. However, because the initialdevelopment of the TRS item poolwas done in an attempt to be atheoretical,the item content of the TRS scales tendsto be more heterogeneous than other ratingscales that used a more explicit guidingtheory for scale development. For example,the Attention Problems scale consists ofitems traditionally associated with inattention(e.g., difficulty concentrating) but alsoincludes items associated with immaturity,overactivity, poor school achievement, andclumsiness. Therefore, it is imperativethat the items that led to a clinical scaleelevation be reviewed to understand themeaning of the elevation. For example, achild may show an elevation on the AttentionProblems scale because of problemswith immaturity, clumsiness, or academicproblems, or a child may have an elevationdue to problems of inattention and/or hyperactivity. However, because of theunreliability of individual items, this itemlevelanalysis should be conducted onlywhen there is an elevation on the clinicalscale. The Attention-Deficit/HyperactivityProblems DSM-Oriented scale can bequite helpful in this type of scenario as itis more closely aligned with characteristicsindicative of ADHD.Interpretation of the Thought Problemsscale on both the TRF and CBCLdeserves several cautionary notes. Thisscale has an especially heterogeneous content,consisting of items describing obsessions(e.g., “Cannot get mind off certainthoughts”), compulsions (e.g., “Repeatsacts over and over”), fears (e.g., “Fears certainanimals, situations, or places”), andpsychotic behaviors (e.g., “Hears sounds orvoices that are not there”), many of whichare fairly ambiguous (e.g., strange behavior,strange ideas). For this scale, it shouldbe apparent that item level interpretationand integration with other informationcollected during the assessment are crucialbefore drawing conclusions.One final note is in order for interpretingTRS scales. Because the norm-referencedscores of the TRS are based on anormal sample and not a normative sample,it is recommended that a more conservativecut-off score be used than would be

chapter 7 Parent and Teacher Rating Scales165Box 7.2Sample Case Using the CBCL and TRFDoug is an 11-year-old, fifth-grade boywho was referred for an assessment becauseof parental and teacher concerns about hisschool performance. He is suspected of havingsignificant attention problems. Doug also hassignificant trouble in peer and other relationships.He often fights and argues with peers,resulting in his often playing by himself.Doug has a history of significant medicaldiffi culties. He is reportedly the product ofan at-risk pregnancy. Although he reportedlyreached most developmental milestoneswithin normal age limits, he has a history ofmotor delays. In second grade, he was diagnosedwith muscular dystrophy. He does nottolerate many foods well, and consequently,his appetite is poor.In the first grade, Doug was diagnosedwith a learning disability in reading. He is ina resource special education program. He isdescribed by his teacher as having significantsocial difficulties. He is reportedly often disrespectfultoward teachers and peers. His gradesdeteriorated significantly toward the end of thelast academic year. His teachers consider himto be a capable underachiever with behaviorproblems such as inattention, excessive talking,fighting, arguing, and poor work completion.Doug’s performance on intelligence andachievement testing indicate that his cognitivefunctioning is consistent with what would beexpected for his age, whereas his writing andreading skills were slightly below what wouldbe expected for his age.Doug’s ratings from the CBCL and TRFcompleted by his mother and teacher reflectthe multitude of his difficulties as follows:Comparisons across these two raters, behavioralobservations, and background informationindicate that Doug is experiencing difficultiesin a number of behavioral, emotional, andsocial domains. In particular, reports by Doug’smother and teacher of his tendency to daydream,his difficulty sustaining attention, hisimpulsive behavior, and his restlessness, as wellas the fact that such behaviors were reportedfor Doug at an early age, suggest that he meetscriteria for a diagnosis of Attention-DeficitScaleMother/Teacher(T)Withdrawn/Depressed 70/61Somatic Complaints 91/64Anxious/Depressed 79/55Social Problems 80/75Thought Problems 79/67Attention Problems 78/69Rule-Breaking Behav. 73/76Aggressive Behav. 85/65Hyperactivity Disorder (ADHD)- CombinedType. Ratings on the CBCL and TRF Attention-Deficit/HyperactivityProblems scale alsoindicated significant problems with behaviorsrelated to ADHD. In addition, Doug’sreported argumentativeness, disrespect towardothers, and tendency to break rules at homeand at school indicate that he meets criteriafor Oppositional Defiant Disorder as well.Some concerns regarding depression werealso reported, particularly by Doug’s mother.However, these issues do not warrant anotherdiagnosis, but they do warrant continued monitoringand some interventions.Indications of thought problems were notcorroborated by other findings. The problemsreported by Doug’s mother on this scale(e.g., trouble sleeping, cannot’t get his mind offcertain things) seemed particularly tied to hisreported problems with depression.CBCL and TRF Social Problems scoreswere corroborated by reports of his difficultygetting along with others. Doug’s social interactionskills are in need of intervention.Although the CBCL and TRF clearlyindicate some areas of concern and in needof intervention, this case also highlights theneed, more pressing in a case like this, tocomplement the CBCL and TRF with otherassessment strategies. In this case, backgroundinformation was particularly important forcorroborating rating scale information andindentifying treatment objectives.

166 chapter 7 Parent and Teacher Rating Scalesthe case for other rating scales. Any elevations,regardless of the degree of elevation,should still be considered in conjunctionwith other assessment results (e.g., parentreport, self-report, history, observations,etc.). The sample case that follows gives abrief example of an interpretive approachto the CBCL and TRF (Box 7.2).Strengths and WeaknessesThe TRF remains one of the most widelyused of the teacher-completed behaviorrating scales. In addition to its popularityand familiarity with a large number ofprofessionals, the strengths of the TRFinclude:1. The large research literature on theTRF and its predecessors which demonstratesgood correspondence betweenthe TRF and other indicators of childfunctioning, particularly on externalizingbehaviors.2. The inclusion of DSM-Oriented scalesaids the clinician in interpreting teacherreports in terms of diagnostic categories.3. A larger normative sample than wasavailable for the previous versions of theTRF.Some weaknesses of the TRF include:1. An underrepresentation of Hispanics inthe normative sample.2. The exclusion of children with mentalhealth or special education services inthe normative sample, indicating thatsuch children are not represented.3. The questionable reliability and validityof the Somatic Complaints scale.4. A relatively limited assessment of adaptivefunctioning.Child Symptom Inventory-4(CSI-4)Parent and Teacher ReportChecklistsThe Child Symptom Inventory-4 (CSI-4;Gadow & Sprafkin, 1998) is a standardizedrating scale designed to assess the symptomsof over a dozen childhood disorders. Thiscontent is unique from other rating scalesin that it is the only omnibus rating scalewhose entire content is explicitly tied to thediagnostic criteria specified in the DSM-IV(American Psychiatric Association, 1994).Therefore, its content reflects the researchthat went into developing these diagnosticcriteria, which is excellent for some disordersbut more suspect for others especiallyfor children (Widiger et al., 1998).The CSI-4 has both parent and teacherreport versions that contain analogousscale content, which enhances its usefulnessfor comparing and combining ratingsacross informants. The CSI-4 was designedfor use with children of ages 5–12, butthere is an analogous Adolescent SymptomInventory-4 for youth ages 12–18 (ASI-4;Gadow & Sprafkin, 1998) that has bothparent and teacher versions and an adolescentself-report checklist, the Youth SymptomInventory-4 (YSI-4). As part of thesame system, the Early Childhood Inventory-4(ECI-4; Gadow & Sprafkin, 1997)assesses DSM-IV symptoms in preschoolchildren of ages 3–5.The content of these forms is mostlyidentical; however, they also each includesome domains that may be particularlydevelopmentally relevant. For example, theASI-4 includes assessments of AntisocialPersonality Disorder, Anorexia, and Bulimia.The ESI-4 omits items screening forpsychosis but includes items for SelectiveMutism, Reactive Attachment Disorder,sleep problems, and elimination problems.

chapter 7 Parent and Teacher Rating Scales167A fairly unique aspect of this system is theinclusion of a symptom checklist specificallyfor ADHD (ADHD-SC4). This inventoryincludes 50 items that assess the core symptomsof inattention and hyperactivity as wellas other areas of interest related to ADHD.More specifically, the ADHD-SC4 includesa Peer Conflict scale to assess the social difficultiesthat often accompany ADHD and aStimulant Side Effects Checklist as a meansto monitor side effects of medication a childmay be taking for the management of his/her ADHD symptoms.ContentThe CSI-4, because of its explicit link tothe DSMIV system for classifying mentaldisorders (Table 7.7), covers many symptomdomains that are not assessed by otherrating scales (e.g., tic disorders), especiallysymptoms of more severe types of childhoodpsychopathology (e.g., ObsessivecompulsiveDisorder, PosttraumaticStressDisorder, schizophrenia, autism, Asperger’sDisorder). As a result, the CSI-4 may beespecially useful in the assessment of moreseverely disturbed children. The items onthe CSI-4 were designed to approximatesymptoms from the DSM-IV with rephrasingdone to eliminate jargon, to emphasizeobservable behavior, rather than makinginferences about internal processes, and toeliminate descriptions of frequency (e.g.,“often” acts without thinking). The CSI-4is fairly long (i.e., 97 items for the parentform, 77 items for the teacher form), but thescales are grouped according to each individualdiagnosis and, as a result, the wholescale need not be given. Instead, symptomsof certain disorders can be selected basedon the specific purpose of the evaluation(e.g., Frick, Bodin, & Barry, 2000).Administration and ScoringThe 97 items on the CSI-4 are rated ona 0 (“Never”) to 3 (“Very Often”) scale.Table 7.6 Prevalence of DSM-IV Disorders in a Normal Sample using the CSI-4 ScreeningCriteriaParent ChecklistTeacher ChecklistDSM-IV Category Boys Girls Boys GirlsAttention-Deficit Hyper. (n = 134) (n = 129) (n = 662) (n = 661)Inattentive 6.4 2.4 11.2 4.2Hyperactive-Impulsive 4.1 3.2 3.5 0.5Combined 4.1 0.8 4.7 1.2Oppositional Defiant 9.2 7.0 6.3 1.8Conduct 6.8 2.3 3.5 1.1Generalized Anxiety 3.7 2.3 0.8 0.8Separation Anxiety 3.0 3.1Schizophrenia 0.0 0.0 0.0 0.0Major Depression 0.0 0.0 0.0 0.0Dysthymia 2.2 0.0 0.8 0.6Autism 0.7 0.8 0.4 0.3Asperger’s 0.0 0.0 0.0 0.0

168 chapter 7 Parent and Teacher Rating ScalesLike other rating scales, quantitative scalescores corresponding to each diagnosticcategory (e.g., Conduct Disorder) can bedetermined by simply summing the ratingsacross items, and this score is calledthe “symptom severity” index. However, a“symptom count” score can be used to moreclosely approximate the DSM-IV methodof considering symptoms as either presentor absent. Using this method, any itemrated as being present “Often” or “VeryOften” is considered to indicate the presenceof the symptom, and any item rated as“Never” or “Sometimes” is considered toindicate the absence of the symptom.NormingThe normative sample of the CSI-4 included552 parent ratings (272 boys, 280 girls) and1,323 teacher ratings (662 boys and 661 girls)in three states (Gadow & Sprafkin, 2002).The children were elementary school-age.Children receiving special education serviceswere not included, making this sample a normalrather than normative sample.In addition to being somewhat geographicallylimited, there was great overrepresentationof Caucasian children, particularlyfor the teacher rating sample, with thatsample being 95% Caucasian, 2.8% AfricanAmerican, and 0.7% Hispanic. Because ofthese limitations in the CSI-4 normativesamples, norm-referenced interpretationsshould only be made very cautiously. However,because the CSI-4 was not designedprimarily to be used as a norm-referencedinstrument but instead was designed as ascreener for DSM diagnoses, the more criticalpsychometric consideration is its reliabilityand validity for this purpose.ReliabilityOne study reporting on the reliability ofthe parent CSI-4 found moderate to goodinternal consistency for both symptomseverityscores and symptom-count scores(Sprafkin et al., 2002). More specifically,internal consistency coefficients rangedfrom a low of 0.45 for the symptom-severityindex for schizophrenia to 0.92 for symptomseverity of ADHD-PredominantlyInattentive Type. Four-month test-retestreliability coefficients ranged from 0.35 forMajor Depression to 0.88 for ADHD PredominantlyHyperactive-Impulsive Type,with all but two coefficients being 0.65 orhigher (Sprafkin et al.).Relatively limited information on thereliability of the teacher version of theCSI-4 is available. For example, the CSI-4manual describes test-retest reliability forthe ADHD and ODD categories during amedication trial for children with behavioralproblems. One-week test-retest coefficientsfor these two diagnoses averaged 0.62for ADHD and 0.90 for ODD (Gadow &Sprafkin, 1998).ValidityThere are several pieces of evidence for thevalidity of the CSI-4 as a screener for DSM-IV diagnoses in school-aged children. First,the preva lence of the diagnoses, based onthe symptom-count scoring method ofthe CSI-4 in the norm sample, seemed toapproximate those found in communitysamples of children using structured diagnosticinterviews (Frick & Silverthorn, 2001).These estimates, computed separately forboys and girls, are provided in Table 8.4.Second, when these prevalence estimateswere compared to a clinic-referred sampleof school-aged children, the prevalence ofDSM diagnoses was higher in the clinicreferredsample for almost all diagnoses.The exceptions were ADHD Hyperactive/Impulsive Type for both boys (7.5% clinicvs. 4.1% norms) and girls (4.7 vs. 3.2%),Asperger’s Disorder for both boys (2.7 vs.0%) and girls (1.3 vs. 0%), and Schizophreniafor boys (1.1 vs. 0%). The primaryconcern is the finding for the one ADHDsubtype, because the failure to find signifi-

chapter 7 Parent and Teacher Rating Scales169cant differences for the latter two disordersseems largely due to the very low base rateof these disorders in both samples.Third, and probably most importantly,Gadow and Sprafkin (1998) reportedon a clinic-referred sample of 101 referrals(between the ages of 6 and 12 years)to an outpatient child psychiatry service,in which they tested the correspondencebetween CSI-4 diagnostic cut-offs andclinical diagnoses made by mental healthprofessionals. In general, the sensitivityand specificity rates for the disordersassessed by the CSI-4 generally indicatedquite good correspondence with clinicaldiagnoses. This correspondence was especiallygood when parent and teacher ratingswere combined, such that a disorderwas considered present if either the parentor teacher ratings led to a CSI-4 screeningdiagnosis. For this multi-informant composite,the Sensitivity rates ranged from0.87 to 1.00, and the specificity rates rangefrom 0.40 to 0.92. For example, a diagnosisof Generalized Anxiety Disorder (GAD)showed a sensitivity rate of 0.93 indicatingthat, of those in the sample who hada clinical diagnosis of GAD, 93% crossedthe screening cut-off for a diagnosis on theCSI-4. The specificity rate of 0.71 indicatesthat, of those without the diagnosisof GAD in the sample, 71% did not crossthe screening cut-off on the CSI-4.Sprafkin and colleagues (2002) foundgood convergent validity for the CSIdomains (parent form) based on theirrelations with the CBCL Syndromescales. Of note, virtually all CSI domainswere moderately correlated with theAnxious/Depressed scale of the CBCL,which may speak more to the generaldistress nature of that CBCL scale thanthe lack of discriminative validity of theCSI domains. They also concluded thatthe CSI-4 is a good screener of a varietyof child disorders based on the highcorrect classification rates found in theirsample.In a separate study, the teacher formof the CSI-4 showed similarly good diagnosticaccuracy with diagnoses made fromstructured interviews and moderate relationswith parent ratings (Gadow et al.,2004).It should be noted that the research onthe parent and teacher forms of the CSI-4far outpaces the research available on theircompanion measures, the ESI-4 and ASI-4. However, this issue is of less concerngiven the highly similar framework underwhich these measures were developed andthe true intent of these measures (i.e., toscreen for symptoms included in a widelyused diagnostic nosology).InterpretationAlthough the CSI-4 content is designedto correspond to the symptoms of DSM-IV disorders, the authors of the scale arevery clear in stating that the scale shouldnever be used in isolation to make diagnoses(Gadow & Sprafkin, 1998). Instead,the CSI-4 is a screener that could indicatethe need for a more complete diagnosticassessment. Rather than being a significantlimitation, it highlights some very importantuses of the CSI-4. As mentioned inChapter 3, there is a great deal of overlapand co-occurrence among the variousforms of childhood disorders. The CSI-4provides an efficient way of screening fora large number of potential comorbiditiesthat can allow for a more focused andintensive assessment in the specific areas ofconcern indicated by this screening. Also,such a screening, because it is time- andcost-efficient, may be quite beneficial fordefining smaller samples at high risk forpsychopathology from larger non-referredsamples (see Frick et al., 2000).Given the fairly substantial limitationsin the normative samples for the CSI-4 andits companion measures, norm-referencedinterpretations are not recommended.Instead, the symptom-count method of

170 chapter 7 Parent and Teacher Rating Scalesscoring is recommended to provide thebest approximation of DSM-IV disorders.Although the normative data suggest thatthe symptom-severity method of scoring issomewhat more reliable, it is not as consistentwith the structure of the DSM criteriathat relies on the presence or absence ofsymptoms to make diagnoses. Also, withoutgood normative data, it is difficult to judgewhen symptom severity should be considered“significant,” unless one is simply trying tomake relative comparisons between groupsof children. In addition, the symptomcountmethod provides a very easy methodfor combining information from multipleinformants, which as the available dataclearly suggest also provides the best correspondenceto clinical diagnoses. Specifically,a symptom can be considered presentif endorsed by any informant (e.g., eitherteacher or parent), and the rate of symptomatologybased on this multi-informantprocedure can be compared to DSM-IVcriteria (see Piacentini, Cohen, & Cohen,1992).Strengths and WeaknessesStrengths of the CSI-4 system include:1. Its uniqueness in attempting to assesscontent that directly corresponds toDSM-IV classifications of childhoodpsychopathology.2. Efficiency in gaining diagnosis-relevantinformation.3. Good correspondence with clinicaldiagnoses, especially when using bothparent and teacher informants.Weaknesses of note include:1. The lack of a large normative base; thus,norm-referenced interpretations shouldnot be made from this rating scale system.2. A relative lack of research, particularlyon the ESI-4 and ASI-4, as well as theself-report component of this system.The CSI-4 and its related measures offera potentially useful component to childassessment, particularly when preliminarydiagnoses are needed for reimbursement/insurancepurposes. However, asthe authors note, the CSI-4 (or any otherassessment technique) should not be usedas the sole criterion for making a clinicaldiagnosis. Instead, such decisions must bebased on a combination of many sourcesof information.Conners, 3rd Edition(Conners-3)Parent Rating ScaleThe Conners-3 (Conners, 2008a) ParentRating Scale (Conners-3-P) is the mostrecent revision to a widely used behaviorrating scale system. The Conners ParentRating Scale is designed similarly tothe BASC and Achenbach systems in thatit includes a number of clinically relevantdomains for which normative scores arederived. The parent rating scale is designedfor ages 6 through 18. The Long Formcontains 110 items and the Short Formcontains 45 items. There is also a 10-itemGlobal Index form. The Conners-3-Ptakes 10–20 min to complete, dependingon which form is used. The following discussionwill focus on the Long Form.As noted in Table 7.1, we recommendthe Conners Comprehensive BehaviorRating Scales (Conners, 2008b) for anassessment that covers externalizing,internalizing, and academic issues. However,as the information below indicates,the Conners-3 is unique in its detailed

chapter 7 Parent and Teacher Rating Scales171evaluation of ADHD and other externalizingissues.Scale ContentThe Conners-3-P includes five empiricallyderivedscales: Hyperactivity/Impulsivity,Executive Functioning, Learning Problems,Aggression, and Peer Relations. AnInattention scale developed theoreticallyis also available, as are five DSM-IV-TRSymptom scales for each of the DisruptiveBehavior Disorders (i.e., 3 ADHD subtypes,ODD, and CD). The Conners-3-Pincludes screening items for depressionand anxiety, as well as impairment itemsfor home, school, and social relationships.Like the BASC, the Conners-3 includescritical items that may signal the need forfurther follow-up. These critical items areparticularly geared toward severe conductproblem behaviors (e.g., uses a weapon, iscruel to animals). Consistent with its predecessors,the Conners-3 includes a briefADHD Index. This scale is based on itemsthat best differentiate ADHD from nonclinicalsamples. As described in Chap. 6for the Conners-3 SR, the Conners-3-Phas three validity scales: Positive Impression(or “fake good”), Negative Impression(“fake bad”), and the Inconsistency Index.These scales are new to the Conners system.Two open-ended questions regardingother concerns and particular strengths/skills are also included. Detailed informationon the generation and selection ofitems is provided in the Conners-3 manual(Conners, 2008a).Administration and ScoringThe Conners-3-P uses a four-choiceresponse format where 0 = not at all true(never, seldom), and 4 = very much true(very often, very frequently). A Spanishtranslation is available. Both hand scoringand computer scoring are available.Raw scores are transformed to linearT-scores, meaning that each scale maintainsits natural distribution in the conversionto norm-referenced scores. Separatenorms are used for boys and girls, as is thecase for the other versions of the Conners-3.Norms are also computed by age.NormingThe normative sample of 1,200 cases wascollected mainly in the USA, with a smallnumber of cases coming from Canada.Recruitment of the normative sample wasaimed at reflecting US Census data regardingethnicity/race. Data reported by Conners(2008a) indicate that the normativesample closely reflects the Census statistics.This representativeness is a notableimprovement over previous versions of theConners rating scale, in that the previoussamples were predominantly Caucasian.As noted for the Conners-3 SR, the WesternUSA appears to have been somewhatunderrepresented. The majority of the parents(63.5%) in the normative sample hadat least some post-secondary education. Aclinical sample of 718 participants was alsocollected for validation purposes, with over35% of that sample being diagnosed withADHD or one of its subtypes.ReliabilityInternal consistency coefficients for thecontent and DSM scales of the Conners-3-P are all 0.80 and higher, and many are0.90 and higher for the overall sample. ThePeer Relations scale for girls had somewhatlower coefficients (i.e., 0.72 for 6–9-year olds; 0.78 for 10–13-year olds) Twoto four week test-retest coefficients weregood (i.e., all higher than .70). Interraterreliability for the parent form was alsogood, with adjusted rs all 0.74 and higherfor the content and DSM scales (see Conners,2008a).

172 chapter 7 Parent and Teacher Rating ScalesValidityBecause the content scales were empirically-derived,it is not surprising thatconfirmatory factor analyses supportedthe five-factor model for those scales. Allintercorrelations among content and DSMscales were moderate to high in magnitude(i.e., ranging from 0.36 to 0.98). Correlationswith analogous scales from theteacher and self-report forms of the Conners-3were all moderate (i.e., rs = 0.49to 0.67). Criterion-related validity wasdemonstrated through moderate to highcorrelations between Conners-3-P scalesand analogous scales from the BASC-2-PRS, CBCL, and BRIEF (see Conners,2008a). The associations betweenthe Conners-3 and CBCL were particularlyhigh. Differential validity evidencealso indicates that the Conners-3-P wassuccessful in distinguishing both a generalpopulation sample and within clinicalsamples. That is, scores on scales such asthose tied to ADHD tended to be elevatedfor clinical groups relative to non-clinicalgroups and higher for individuals withADHD relative to others within a clinicalpopulation. The correct classificationrate based on content and DSM scale elevationswere also relatively high (i.e., 57to 86%). More validity evidence for theShort Form and the Indexes are availablein the manual (Conners, 2008a).and (e) integration with other assessmentinformation.Strengths and WeaknessesSome of the strengths of the Conners-3-Pare:1. The improved representativeness ofnormative sample2. Availability of complementary teacherand self-report forms that provide acomprehensive assessment of externalizingproblems3. Good initial reliability and validity evidence4. Brevity of Short and Index FormsSome characteristics that may be consideredweaknesses are:1. Limited assessment of internalizingproblems and adaptive functioning (anissue that is addressed through use ofthe Conners CBRS)2. Uniform negative wording of items,which may result in a negative responseset3. A lack of available validity research conductedby persons other than the developersInterpretationAs discussed in Chap. 6, Conners (2008a)provides a clear step-by-step approach forinterpreting ratings on the various forms ofthe Conners-3. This approach involves (a)examining the validity scales; (b) evaluatingscale elevations; (c) examining the overallprofile of scores (i.e., determining the constructsthat seem to be represented acrosselevations); (d) item-level interpretation,including critical items and screener items;Teacher Rating ScaleThe Teacher Rating Scale in the Conners-3system is very similar to the parentrating scale. In fact, the teacher ratingsscale and parent rating scale includethe same scales. The Long Form of theteacher rating scale is slightly longer thanthat of the parent rating scale (i.e., 115items), whereas the Short Form is slightlyshorter (i.e., 41 items). Two 10-itemHyperactivity Index forms are also available.The following discussion will focus

chapter 7 Parent and Teacher Rating Scales173primarily on the Long Form of the Conners-3-T.As with the self-report and parentforms of the Conners-3, we discussthe Conners-3-T because of its relativelyunique focus on ADHD and behavioralproblems. The companion teacher ratingscale from the Conners CBRS (Conners,2008b) provides a more extensive assessmentof broader domains of functioning.Therefore, the selection of one set of ratingscales versus the other in the Connersfamily should be dictated by the purposeof the evaluation.ContentThe Conners-3-T has some item overlapwith the Conners-3 parent rating scale, butthere are also unique items in each form.The same four-point response scale usedfor the self-report and parent-report versionsof the Conners-3 is also used for theteacher-report scale. As noted above, thescales are the same as those for the parentrating scale, including validity scales,impairment items, and critical items.Administration and ScoringThe Conners-3-T can be completed in10–20 min, or less if the Short Form is used.The Conners-3-T has both hand-scoringand computer-scoring formats that allowfor easy calculation of norm-referencedscores. As with the self-report and parentreport forms, only sex-specific T-scorescan be calculated. The scores are LinearT-scores and are based on each age group,which allows it to capture potential variabilityin discrete developmental stages.NormingThe norming process for the Conners-3-T was essentially the same as that usedfor the Conners-3-P and Conners-3 SR.Specifically, the norming sample for theConners-3-T consists of 1,200 teachersfrom throughout the USA, with a fewrespondents from Canada. Recruitment wasaimed at a sample that would reflect U.S.Census data on ethnicity/race. The studentsrated by teachers in the norming sampledo appear to match the Census data onethnicity/race (Conners, 2008a). However,the sample appears to be somewhat skewedtoward middle to high SES – based on parenteducation – as 76.9% of the studentsrated by teachers in the norming samplehad parents with at least some post-secondaryeducation. Almost as many cases came fromCanada as came from the western USA inthe Conners-3-T norming sample.ReliabilityInternal consistency coefficients for theteacher report version Conners-3 werequite high. Specifically, the coefficientsfor each of the content and DSM scaleswere 0.90 or higher, with the exceptionof the Conduct Disorder scale (0.77).Two- to four-week test-retest reliabilitycoefficients were also good, ranging from0.72 to 0.83 (see Conners, 2008a). Lastly,and perhaps particularly importantly forteacher ratings, interrater reliability coefficientsfor pairs of teacher raters weremoderate to high. The Peer Relationsand Oppositional Defiant Disorder scaleshad the lowest adjusted coefficients (i.e.,0.52 and 0.55, respectively), whereas theHyperactivity/Impulsivity and ConductDisorder scales had the highest coefficients(i.e., 0.77). It should be noted thatthe lower coefficients from these analysesmay reflect less-than-ideal rater agreement,or they may reflect real differencesin a child’s behavior from one classroomcontext to another. Additional analyses,particularly in determining whetherteacher agreement might change as afunction of the child’s age, are needed.Validity

174 chapter 7 Parent and Teacher Rating ScalesFactor analyses revealed a four-factorsolution for the Conners-3-T: Hyperactivity/Impulsivity,Aggression, Peer Relations,and a combined Learning Problems/Executive Functioning scale. Conners(2008a) also found support for consideringthe Learning Problems/ExecutiveFunctioning as consisting of two subscalesconsisting of items intended to loadon a Learning Problems and an ExecutiveFunctioning scale. As noted above,the Conners-3-T scales were moderatelycorrelated with the same scales fromthe parent and self-report versions. Thescales on the Conners-3-T were all moderatelyinterrelated. Criterion-relatedvalidity for the Conners-3-T was supportedthrough moderate to high correlationsbetween Conners-3-T scales andanalogous scales on the BASC-2-TRS,Achenbach TRF, and BRIEF TeacherForm. Similar to the parent version ofthe Conners-3, the teacher version demonstratedgood differential validity inthat scales were elevated for individualsfrom a clinical sample relative to a generalsample, and scale scores tended todiffer within the clinical sample in intuitiveways. For example, ADHD scalescores tended to be higher for youthsdiagnosed with ADHD than for youthswith other difficulties who did not haveADHD diagnoses (see Conner, 2008a).InterpretationAt the very least, the Conners-3-T appearsto be useful as a screening for problems inclassroom adjustment, particularly in termsof learning or externalizing problems, andas part of a comprehensive assessmentbattery. The recommended approach forinterpreting the Conners-3-T mirrors thatdescribed for the Conners-3 SR and Conners-3-P.Strengths and WeaknessesThe strengths of the Conners-3-Tinclude:1. Content that allows for an extensiveassessment of ADHD symptoms andother behavioral problems.2. Good correspondence across scales onthe parent and teacher versions, whichfacilitates comparisons in a multi-informantassessment.3. The presence of several short screeningscales which may be more feasiblefor many teachers.Apparent weaknesses of the Conners-3-T include:1. Minimal assessment of depression andanxiety, as well as adaptive functioning.The Conners CBRS should be used ifextensive assessment of these domains isdesired.2. Lack of research on reliability and validityconducted by persons other than thedeve loper.3. The normative sample is not quite asdiverse as that for the parent and selfreportforms of the Conners-3, yet it isstill diverse in terms of race/ethnicity.Personality Inventory forChildren-2 (PIC-2); StudentBehavior Survey (SBS)Parent Report PIC-2The Personality Inventory for Children-2[(PIC-2); Lachar & Gruber, 2001] is basedclosely on its predecessor, the PIC-R(Wirt, Lachar, Klinedinst, & Seat, 1990).The original development of the PIC followedclosely on the heels of the MMPI,with much of the early development worktaking place in the 1950s. The PIC-2 is a275-item rating scale designed for use with

chapter 7 Parent and Teacher Rating Scales175parents of children between the ages of 5and 19 years (Lachar & Gruber, 2001).Scale ContentThe PIC-2 scales, although revised, havea long clinical history. The PIC-2 includesscales that were developed via a mixture ofempirical means with considerable use ofexternal validation techniques and scalesTable 7.7 PIC-2 Clinical Scales, Subscales, and Internal Consistency CoefficientsCognitive Impairment (r = 0.87)Inadequate Abilities (r = 0.77)Others think my child is talentedMy child seems to understand everything thatis said.Poor Achievement (r = 0.77)It is hard for my child to make goodgrades.Reading has been a problem for my child.Developmental Delay (r = 0.79)At one time my child had speech difficulties.My child could ride a tricycle by the ageof 5.Impulsivity and Distractibility (r = 0.92)Disruptive Behavior (r = 0.91)My child jumps from one activity toanother.My child cannot keep attention on anything.Fearlessness (r = 0.69)My child will do anything on a dare.Nothing seems to scare my child.Delinquency (r = 0.95)Antisocial Behavior (r = 0.88)My child has been in trouble with thepolice.My child has run away from home.Dyscontrol (r = 0.91)When my child gets mad, watch out!Many times my child has become violent.Noncompliance (r = 0.92)My child often breaks the rules.My child tends to see how much he/she canget away with.Family Dysfunction (r = 0.87)Conflict among members (r = 0.83)There is a lot of tension in our home.Our family argues a lot at dinner time.Parent Maladjustment (r = 0.77)One of the child’s parents drinks too muchalcohol.The child’s parents are divorced or livingapart.Reality Distortion (r = 0.89)Developmental Deviation (r = 0.84)My child often gets confused.My child needs protection from everydaydangers.Hallucinations and Delusions (r = 0.81)My child thinks others are plotting againsthim/her.My child is likely to scream if disturbed.Somatic Concern (r = 0.84)Psychosomatic Preoccupation (r = 0.80)My child is worried about disease.My child often has an upset stomach.Muscular Tension and Anxiety (r = 0.68)Recently my child has complained of chestpains.My child often has back pains.Psychological Discomfort (r = 0.90)Fear and Worry (r = 0.72)My child will worry a lot before startingsomething new.My child is often afraid of little things.Depression (r = 0.87)My child has little self-confidence.My child hardly ever smiles.(Continues)

176 chapter 7 Parent and Teacher Rating ScalesTable 7.7 (continued)Psychological Discomfort (r = 0.90)Sleep Disturbance/Preoccupation withDeath(r = 0.76)My child’s sleep is calm and restful.My child thinks about ways to kill himself/herself.Social Withdrawal (r = 0.81)Social Introversion (r = 0. 78)My child is usually afraid to meet newpeople.Shyness is my child’s biggest problem.Isolation (r = 0.68)My child does not like to be close withothers.My child often stays in his/her room forhours.Social Skill Deficits (r = 0.91)Limited Peer Status (r = 0.84)My child often brings friends home.(reversed)My child is very popular with other children.(reversed)Conflict with Peers (r = 0.88)My child seems to get along with everyone.(reversed)Other children make fun of my child’sideas.Note: From Lachar & Gruber (2001).developed through rational/ theoreticalapproaches.Many changes and improvements havebeen made in the PIC-2 scales. Contentoverlap was either reduced or eliminatedbetween scales, item-total correlation hadto be high, and validity scales were added(Lachar & Gruber, 2001). Scale content wasalso better articulated with that of the PIYin order to enhance score comparisons. ASpanish translation was developed as well.The PIC-2 also includes a 96-item shortform (the first 96 items of the StandardForm) called the “Behavioral Summary.”An overview of the PIC-2 clinical scalesis provided in Table 7.7. In addition to thesescales, the PIC-2 provides three validityscales (i.e., Inconsistency, Dissimulation,and Defensiveness) and critical items.Administration and ScoringIt takes a parent about 40 min to completethe 275 true–false statements of the PIC-2. All administrations require at least twocomponents, an administration bookletand hand-scoring or computer-scoringanswer sheets. The hand-scoring processinvolves the use of four forms with a CriticalItems Summary Sheet as an option.The use of either PC or mail-in computerscoring limits the number of componentsto only two (administration booklet andanswer sheet).NormingThe norming sample included 2,306 childrenin the kindergarten through 12thgrades. The normative sample appears torepresent 1998 US Census data – whichwere the data available at the time of thePIC-2 norming – well in terms of ethnicity,parents’ education level, and geographicregion of residence.Linear transformations of T-scores wereutilized. The range of derived scores islimited to T-scores based only on withinsexcomparisons. Therefore, as alluded toin the discussions of other tests, one is notable to determine how a child’s behaviorcompares to that of children in general.Percentile ranks are also not available.ReliabilityInternal consistency coefficients for thescales are for the most part acceptable andare shown in Table 7.7. The results of one-

chapter 7 Parent and Teacher Rating Scales177week test-retest studies are also generallysupportive (see Lachar & Gruber, 2001).Interrater reliability between mothers andfathers was generally very good, with coefficientsmostly 0.75 and higher for nonclinic-referredchildren. One exceptionwas the Somatic Complaints scale and itssubscales, with coefficients of 0.49 to 0.54(Lachar & Gruber, 2001).ValiditySeveral types of validity evidence arereported in the PIC-2 manual includingcriterion-related, differential diagnosis,and factorial validity. Factors correspondingto the Externalization, Internalization,Social Adjustment, and Total compositescores are described.The relations between PIC-2 scores andexternal indicators of adjustment are describedin detail in the manual (see Lachar & Gruber,2001). Some of the indicators include teacherSBS and child self-report PIY ratings. Unfortunately,such studies, by being limited tothe PIC “family” of measures, do not allowclinicians to determine the degree to whichPIC-2 results will differ from CBCL, BASC-2, MMPI-A, or other results. Evidence ofthis nature is important, as clinicians oftenuse multiple measures and frequently haveto describe their findings in comparison toprevious evaluation results. The extent ofPIC-2 criterion-related validity evidence tobe found in the manual is sometimes difficultto discern. Considerable reference is made toSBS and PIY validity studies.Children with diagnoses in the clinicalsamples were used to compare PIC-2results for several diagnostic groups usingMANOVAs. Many significant effects werefound. However, sensitivity, specificity, andother typical indices of diagnostic accuracyare not provided.As is the case with the PIY, independentevidence of validity is difficult to obtain atthis time. Several aspects of validity remainto be assessed in order to support clinicians’use of the scale. First priority for furthervalidation is to assess the criterion-relatedvalidity of the PIC-2 with widely used scales,such as the CBCL and BASC-2 PRS becausemany clinicians will be faced with having tointerpret PIC-2 results in tandem with thesemeasures.InterpretationChapter 3 of the PIC-2 manual providesconsiderable guidance to the user. In fact,the sheer amount of tabular informationpresented is potentially overwhelming. Thefrequency of item endorsements for varioussamples, for example, is presented foreach scale. The value of such informationis questionable because it is based on theassumption that an item response is a reliableand valid indicator of some construct,which is a dubious assumption. Nevertheless,the manual provides numerous useful casestudies and correlates of profiles. In addition,the meaning of various T-scores forthe individual scales is thoroughly describedin an additional set of tables. Clinicians willprobably find these descriptions of T-scoreoutcomes to be valuable for deriving scoremeaning.Otherwise, we reiterate our recommendedsequential approach to interpretation(i.e., checking validity scales, criticalitems, scale elevations, subscale elevations,relevant item endorsements, consideringprimary vs. secondary concerns, integrationwith other information).Strengths and WeaknessesPIC-2 strengths include:1. A thorough manual by Lachar and Gruber(2001) that summarizes importantstudies of scale development.2. A great variety of subscale scores thatmay be of value for specialized uses.3. The inclusion of valuable interpretiveguidance in the manual.

178 chapter 7 Parent and Teacher Rating Scales4. Norming sample that closely matchescensus data at the time of the scale’sdevelopment.Weaknesses of the PIC-2 may include:1. Test length.2. A lack of criterion-related validity studiesand shortage of validity studies independentof the test developers.3. Limited score options (i.e., absence ofgeneral norm-referenced comparisonsand percentiles).The PIC-2 represents a significantupgrade of the PIC-R. The most importantimprovements are a reduction of itemoverlap between scales and the collectionof new norms. Both independent validationresearch and clinical experience are necessaryto determine the ultimate utility of thescale.Teacher Report: The StudentBehavior SurveyThe Student Behavior Survey (SBS; Lachar,Wingenfeld, Kline, & Gruber, 2000) is theteacher version of the rating scale systemthat includes the parent-completed PIC-2youth self-report PIY. As a result, SBSrounds out a rating scale system with a longand distinguished history in the assessmentof children and adolescents by providing asource of information on a child’s classroomadjustment based on teacher report. TheSBS is not as long as its parent-report andself-report siblings, containing 102 itemsthat are rated on a four-point Likert scale.This rather moderate length allows mostteachers to complete the form easily in15–20 min. The scale has normative informationfor children of ages 5 through 18.ContentDespite being developed to complementthe PIC-2 and PIY scales, theSBS was not beholden to the item contentof the parent-report and self-reportscales. Instead, the content of the SBSwas developed based on teacherendorsements of statements thatseem to reflect important dimensionsof classroom adjustment. The contentof the SBS can be divided into threemajor categories. The first category isAcademic Resources, which containsfour subscales: Academic Performance(eight items), Academic Habits (thirteenitems), Social Skills (eight items), andParent Participation (six items). Thesesubscales are adaptive scales focusing onpotential strengths of the child in the academicenvironment, and therefore, itemson these subscales are worded in a positivedirection. The second category is AdjustmentProblems, which includes sevensubscales: Health Concerns (six items),Emotional Distress (fifteen items), UnusualBehavior (seven items), Social Problems(twelve items), Verbal Aggression (sevenitems), Physical Aggression (five items),and Behavior Problems (fifteen items).These two areas include the main clinicalscales of the SBS focusing on emotional,social, and behavioral areas of concern forthe child’s classroom adjustment.The third section is a Disruptive BehaviorDisorders category that includes threesubscales: Attention-Deficit Hyperactivity(16 items), Oppositional-Defiant (16items) and Conduct Problems (16 items).As the names of the subscales imply, thesescales were developed to provide a screeningfor the major disruptive behavior disordercategories specified in the DSM-IV.However, the individual items were notspecifically developed to tap DSM criteria.Instead, three clinicians chose itemsfrom the existing 102-item pool that werejudged to be most indicative of the DSM-IV criteria, a similar approach to thatemployed for the Achenbach measures

chapter 7 Parent and Teacher Rating Scales179(discussed earlier). This procedure led tosome criteria not being assessed (e.g., “Isspiteful and vindictive”) and other itemsincluded that are not part of the DSM criteria(e.g., “Demonstrates polite behaviorand manners” reverse-scored). This issueis especially relevant to the Conduct Problemsscale, which is fairly divergent fromthe content of the DSM-IV definition ofConduct Disorder, including such items as“uses drugs or alcohol” and “preoccupiedwith sex.”Administration and ScoringThe items on the SBS are groupedaccording to their subscales, such thatthe 8 items for the Academic Performancesubscale are items 1 through 8,the 13 items for the Academic Habitssubscale are items 9 through 21, and soon. In addition, the subscale titles documentthis explicit grouping to the teacherraters. This is a somewhat unique formatin that other rating scales have items forthe subscale intermixed throughout thescale. There could be both positive andnegative consequences of this format. Forexample, it makes scoring much easier andreduces the likelihood of clerical errors incomputing raw scores, because it is readilyapparent which items are included oneach subscale. Also, it makes inspection ofitems that led to subscale elevations a verysimple process. Alternatively, it opens thepossibility that teachers may be influencedby the name of the construct (e.g., socialskills) and rate children according to theiroverall perceptions of a child’s adjustmentfor that domain rather than basing theirratings on their perceptions of the individualbehaviors. For example, a teacherwho views a child as socially unskilledmay rate items under that heading asmore problematic than if he or she wasnot explicitly informed about the overalldomain being assessed.However, there is no empirical evidencethat this item format affects ratingsin any systematic way, and as mentionedpreviously, it greatly simplifies the scoringprocess. There are two “Auto-Score”forms for the SBS: one for children ofages 5–11 and one for adolescents of ages12–18. Raw scores are simply computedby summing the ratings within each ofthe 11 subscales included in the AcademicResources and Adjustment Problemdomains. Between the two sides of theratings is carbon paper that copies ratingson only those items that correspond to thethree disruptive behavior subscales. Rawscores are based on a sum of these items aswell. These 14 raw scores are then transferredto a cover Profile page with separatecolumns for boys and girls. Theseprofiles reflect a conversion to T-scoresand show the relative elevations amongsubscales based on this norm-referencedscore. Importantly, the conversions andprofiles can only be computed for separatemale and female norms, and not forboth sexes combined.NormsThe primary normative sample for the SBSincludes 2,612 children from regular educationclassrooms from 22 schools in 11states. The sample was fairly evenly dividedbetween boys and girls and had substantialrepresentation at each year of age from 5 to18. Also, the regional breakdown, parentaleducational level, and ethnic composition(e.g., 70% Caucasian, 15% African American,10% Hispanic American) was fairlyrepresentative of US Census Bureau statistics(see Lachar et al., 2000). The one relativelyminor exception was the somewhathigh rate of college graduates in this normsample (i.e., 35 vs. 26.9% cited for the USCensus).One of the unique features of the SBSis that, in addition to the regular educationnorm sample on which T-score conversa-

180 chapter 7 Parent and Teacher Rating Scalestions were based, the manual also reportedon a large referred sample (n = 1,315) thatobtained teacher ratings on children from41 different sites in 17 states in the USA.These children included those in specialeducation classes, those referred to bothinpatient and outpatient mental healthclinics, and those referred to juvenile justicefacilities. This large sample allows for acomparison of the psychometric propertiesof the SBS in both a large normal sampleof children and in a large disturbed sample.Overall, each of the SBS scales differentiatedthe referred and normal samples withCohen’s d ranging from 0.23 (Parent Participation)to 0.98 (Academic Performance;see Lachar et al., 2000).ReliabilityThe information provided in the manual(Lachar et al., 2000) on the reliability of theSBS is exemplary. Internal consistency estimatesfor the 14 subscales across both thenormal and referred samples ranged from0.85 to 0.95, indicating uniformly excellentinternal consistency. Test-retestcorrelationsare provided for four samples ofchildren ranging in age from 5 to 18 andwith retest intervals ranging from 2 to 30weeks. Again, all scales showed quite goodtemporal stability, with the test-retest ofthe Unusual Behavior scale over a 20-weekperiod in adolescents being the only indexto be somewhat low (i.e., r = 0.29). A thirdtype of reliability, inter-rater agreement,was tested in two samples of 30 children,one sample including fourth and fifth graderegular education students and a secondsample including children (ages 5–12)receiving special education services. Thecorrelations between two teacher ratingsacross these samples ranged from 0.44 to0.91, with most indexes being above 0.70.ValidityThe dimensionality of the SBS was tested ina way that was somewhat different from thatreported for other behavioral rating scales.That is, rather than conducting a factoranalysis on the individual items, the itemsubscalecorrelations were compared foreach item’s correlation with the dimensionit is purported to assess and its correlationswith other dimensions. While this methodled to rationally derived scales that werefairly homogeneous in content, the decisionas to whether an item is “more strongly”associated with the dimension it is purportedto measure is somewhat subjective inthe absence of factor analysis. For example,“Blames others for own problems” is correlated0.79 with the Behavior Problems subscaleon which it is included, but it is alsocorrelated 0.76 with the Verbal Aggressionsubscale, 0.61 with the Physical Aggressionsubscale, and 0.54 with the Social Problemssubscale. The most problematic in this regardare the three Disruptive Behavior Scales,on which many items load equally high onall three dimensions, although this is likelydue to the nature of the criteria they weredesigned to assess, which tend to be substantiallyoverlapping (Frick et al., 1994).The manual of the SBS (Lachar et al.,2000) provides (1) the correlations of theSBS subscales with clinician ratings ofadjustment problems in 129 primarilyclinic-referred children, (2) the correlationsamong SBS scores and parent- andself-report ratings using the PIC-2 andPIY, and (3) the correlations between theSBS and an early version of the ConnersRating Scale for teachers (see also Piseccoet al., 1999; Wingenfeld, Lachar, Gruber,& Kline, 1998).In general, these correlations supportthe convergent validity of the SBS scales,but like most rating scales, the divergentvalidity was less clear. That is, the SBS subscaleswere often correlated with the otherscales designed to measure similar constructs(i.e., convergent validity), but theywere also correlated with other dimensions

chapter 7 Parent and Teacher Rating Scales181of maladjustment as well. For example, theEmotional Distress subscale was significantlycorrelated with clinician ratings ofpsychological discomfort (r = 0.55), butthis subscale was also highly correlatedwith the ratings of disruptive behavior (r= 0.44). Again, this pattern is common forratings of children’s adjustment becausechildren with problems in one area oftenhave problems in many other areas ofadjustment as well, and raters may alsodemonstrate response sets in that a childrating negatively in one area is rated similarlyin other areas. One notable weaknessuncovered in these validity analyses wasfor the Unusual Behavior subscale, whichseemed to be more strongly associatedwith measures of disruptive behaviors andADHD than with more severe psychopathologyor reality distortion. For example,it was correlated 0.40 with clinician ratingsof disruptive behavior but 0.25 with clinicianratings of serious psychopathology.Similarly, the Unusual Behavior subscalewas correlated at 0.41 with parent ratingsof impulsivity and distractibility on thePIC-2, but at 0.27 with the Reality Distortionsubscale of the PIC-2.One additional set of validity analysesprovided in the manual were comparisonsbetween groups of children either diagnosedwith Disruptive Behavior Disordersby clinicians or children elevated on theHyperactivity Index in an earlier versionof the Conners Rating Scale compared tocontrol children. As would be expected, theSocial Problem subscale, the three behaviorproblem subscales, and the disruptivebehavior disorder subscales all differentiatedchildren with behavior problems fromcontrol children. Also as expected, theacademic resources subscales tended to belower in groups of children with behavioralproblems, with the exception of the ParentParticipation subscale.InterpretationWithin the tradition of the PIC-2, which, inturn, was based on the MMPI tradition, themanual of the SBS provides a very detailedstep-by-step interpretative guide (Lacharet al., 2000). First, the manual recommendsexamining items for response adequacy,including ensuring that there are only a fewmissing responses. The one exception notedin the manual is that many teachers abovethe early elementary school grades may havedifficulty completing the Parent Participationscale because they are less likely to conversewith parents on a regular basis (Lacharet al., 2000). Also, it is important to notethat, unlike the PIC-2 and PIY, there are novalidity indexes on the SBS designed to helpin detecting potential threats to the qualityof the teacher ratings. Second, and themain focus of the interpretative approach inthe manual, is a description of the characteristicsthat are often associated with childrenwho score in a given range on eachsubscale.These interpretive guidelines weredeveloped by correlating the T-scores onthe SBS subscales with descriptors providedby clinicians (n = 379), parents (n= 425), and students (n = 218). Descriptorsare provided for T-scores below 40for the academic resources subscales andfor (1) T-scores between 60 and 70, and(2) T-scores above 70 for the adjustmentproblems scales. The authors note that thedescriptors for the higher elevations (above70) should be considered more definitivethan those between 60 and 70. The authorsclearly note, however, that all interpretations,even those above 70, should be consideredonly as “interpretative hypotheses,”and additional information (e.g., fromparent report, child self-report, and clinicalobservations) should be used to betterdetermine if these hypothetical descriptorsare appropriate for a given case.Strengths and WeaknessesThe strengths of the SBS include:

182 chapter 7 Parent and Teacher Rating Scales1. Content that includes a number of adaptivedimensions of classroom adjustmentand a rather comprehensive assessmentof conduct problems, including separatesubscales for verbal and physical aggression,and a general Behavior Problemssubscale.2. Fairly homogeneous subscale content,which greatly enhances the interpretationof scale elevations, as does thevery easy-to-use, step-by-step interpretiveguidelines, which provide the mostcommon characteristics for childrenwith specific scale scores.3. A large and representative norm sample.4. The evidence for subscale reliabilityusing both community and clinicreferredsamples is exemplary.All of these characteristics make the SBS avery useful tool for obtaining teacher ratingsof classroom adjustment.Weaknesses of the SBS include:1. Limited research on the validity of theSBS scales and subscales2. A lack of cross-validation in other samplesof the interpretative descriptorsprovided for children who score in aspecific range on each subscale need tobe cross-validated.3. The heterogeneous content of theUnusual Behavior subscale includessome items related to inattention(e.g., “Daydreams”) and some vaguebehaviors (e.g., “Behavior is strangeand peculiar”). Early evidence suggestedthat it is more associated withBox 7.3Sample Case Using the PIC-2 and SBSRicky is a 7-year-old boy who was referred foran evaluation by his teacher because of concernsabout inattention, overactivity, and poorpeer interactions. Ricky was reportedly bornat home without the benefit of any medicalattentionMost of Ricky’s developmental milestoneswere delayed, especially language. Accordingto his mother, he did not speak his first wordsuntil the age of 2 years. He has reportedlydemonstrated some improvements in languagesince beginning speech therapy at age 4.According to his mother, as a toddler, Rickybegan demonstrating significant behavioralproblems including frequent temper tantrums,overactivity, and oppositional behavior. Suchbehavioral concerns continued when Rickybegan school. He has a history of getting introuble at school due tobecause of overactivityand defiance toward his teachers.Throughout the evaluation, Ricky demonstrateda short attention span and a highlevel of motor activity. He did not sit still andwas easily distracted by other objects in theroom. When he was unable to testing objectsand toys in the waiting room with him, Rickydisplayed tantrum behavior (e.g., kicking andcrying).Ricky’s cognitive assessment results indicatedoverall functioning in the Low Averagerange, with his verbal skills being in theBorderline range. Tests of his achievementin reading and math indicated slightly belowaverage achievement relative to his sameagedpeers.PIC-2 Scale elevations were as follows:Impulsivity/Distractibility T = 80Delinquency T = 82Family Dysfunction T = 75Social Skill Deficits T = 68(Continues)

chapter 7 Parent and Teacher Rating Scales183Box 7.3 (Continued)SBS Scale elevations were as follows:Academic Habits (adaptive) T = 33Social Skills (adaptive) T = 32Emotional Distress T = 71Social Problems T = 67Unusual Behavior T = 63Verbal Aggression T = 80Physical Aggression T = 72Behavior Problems T = 82Attention-Deficit/Hyperactivity T = 73Oppositional Defiant T = 73On rating scales, both Ricky’s motherand teacher reported significant concernsregarding inattention, overactivity, impulsivity,and externalizing behaviors. Thesebehavioral concerns have apparently beenpresent for some time. Furthermore, basedon background information as well as ratingsby Ricky’s mother and teacher, Ricky’sbehavioral problems are causing impairmentin his relationships with others and his abilityto perform required tasks in the classroom.Therefore, diagnoses of Attention-Deficit/Hyperactivity Disorder (ADHD)-CombinedType and Oppositional Defiant Disorder arewarranted. The elevation on the SBS EmotionalDistress scale is consistent with reportsthat Ricky gets upset easily and throws tantrumsin the classroom. The slight elevationthe Unusual Behavior scale of the SBS completedby Ricky’s teacher appears consistentwith attention problems in that she reportedthat he daydreams and seems disoriented.In addition, Ricky’s mother indicated concernsregarding a high level of conflict inthe home, particularly between Ricky andhis parents. This conflict appears to also berelated to Ricky’s behavioral problems, particularlyhis defiance.Recommendations for Ricky includedparental consultation with a mental healthspecialist to address his behavioral problems,consultation with a physician regardinga possible medication trial for hisADHD symptoms, and classroom accommodationsto help minimize the impact ofRicky’s behavioral concerns and inattentionon his academic and social functioning. Theresults of parent and teacher rating scales inthis case highlight the dissimilar structure ofthe PIC-2 and SBS. However, similar informationcan still be gleaned from these ratingscales that can aid in case conceptualizationand recommendations.disruptive behavior dimensions thanwith indexes of more severe psychopathologyand thought disturbances.4. The lack of direct correspondencebetween the three disruptive behaviordisorder subscales and DSM criteria. Thisis especially true for the Conduct Problemsscale, which appears quite divergentfrom the criteria for Conduct Disorder.In addition, there is no evidence for howwell the specific SBS subscales (e.g.,Attention-Deficit Hyperactivity) correspondto specific DSM-IV diagnoses(e.g., ADHD). As a result, the usefulnessof SBS as a screener for specific DSM disordershas not been established.In addition to these issues, it is worth notingthat while SBS was developed to bepart of the assessment system that includesthe PIC-2 and PIY (reviewed previously)the item content and scale structure of theSBS is substantially different from these

184 chapter 7 Parent and Teacher Rating Scalesother scales. The result is a tool that is veryrelevant for assessing children’s classroomfunctioning. However, it also makes it moredifficult to integrate information from thedifferent informants. A case example withPIC-2 and SBS data follows (Box 7.3).Sample Impairment-Oriented ScalesAs can be determined from the previous review,omnibus rating scales can provide invaluableinformation about a variety of domains ofchild functioning. This information, however,tends to describe functioning in termsof severity of problems and/or frequency ofproblems. Rating scales typically stop short ofproviding an indication as to what extent theproblems interfere with the child’s functioning.Information on impairment is often leftto the clinician to infer based on interview orother information. However, this informationis no less important for case conceptualizationand treatment planning. In additionto assessing for impairment via structured orunstructured interviews, one may employan inventory to gather such information ina time-efficient manner and then follow-upaccordingly. A brief discussion of some suchinventories follows.Home Situations Questionnaire(HSQ) and School SituationsQuestionnaire (SSQ)The content of the Home Situations Questionnaire(HSQ; Barkley & Edelbrock,1987) and the School Situations Questionnaire(SSQ; Barkley & Edelbrock, 1987) ismarkedly different from the other ratingscales reviewed in this chapter. Rather thanhaving items that describe different typesof child behaviors, these measures includesituations (e.g., while playing alone, whenvisitors are in the home, during individualdesk work, at recess, on the bus) in whicha child may have problems. That is, theHSQ and SSQ were not designed to assessspecific behaviors but to assess specificsituations in which problem behaviors canoccur. Therefore, these measures providean indication of the specific situations inwhich the child may demonstrate particulardifficulty or impairment.Both measures were designed to be completedin the same manner. The respondent(parent or teacher) rates whether ornot the child has any problem in a givensituation and then rates the severity of theproblem on a 1–9 scale. These measuresmay be used with a variety of clinical problems,as the respondent can be directed torespond as to whether or not the child “hasproblems” in the situations provided.The psychometric development of bothmeasures is limited. Normative informationis available from Altepeter and Breen (1989)as well as Barkley and Edelbrock (1987).However, norm-based comparisons maynot represent the best use of these tools.Factor analyses have revealed four factorsfor the HSQ (i.e., Non-Family Transactions,Custodial Transactions, Task-PerformanceTransactions, and Isolate Play) andthree factors for the SSQ (i.e., UnsupervisedSettings, Task Performance, and SpecialEvents; Altepeter & Breen, 1989).The HSQ has demonstrated good testretestreliability and internal consistency(Altepeter & Breen, 1989). The number ofproblems and mean severity rating of theHSQ have been found to be related to ratingsof impulsivity and hyperactivity (Altepeter& Breen, 1989). Test-retest reliabilityof the SSQ in a sample of 119 regular educationchildren was estimated at 0.68 for thenumber of problem situations and 0.78 forthe mean severity score (Barkley & Edelbrock,1987). Also, inter-rater agreement forthe SSQ was tested in a sample of 46 studentsages 8–17. The correlation between teacherswas 0.68 for the number of problem situationsand 0.72 for the mean severity score(Danforth & DuPaul, 1996). Barkley and

chapter 7 Parent and Teacher Rating Scales185Edelbrock (1987) reported numerous significantcorrelations between the SSQ and ratingscale measures of externalizing behaviorproblems and evidence that the SSQ differentiateschildren with ADHD from childrenwithout ADHD. However, for both theHSQ and SSQ criterion-related validity evidenceis more difficult to operationalize, asthese measures have a different focus thanratings of symptoms or problems. Still, situationsin which the child has difficulties, asindicated on the HSQ and SSQ, can assistthe clinician in appropriately designing andprioritizing intervention strategies.Child Global Assessment Scale(CGAS)Another example of an assessment of impairmenttakes a different approach. The ChildGlobal Assessment Scale (CGAS; Shafferet al., 1983) is an adaptation of an adult scaledesigned to assess overall level of impairmentat home, in school, or with friends. The scaleextends from a low of 1 (extremely impaired)to a high of 100 (no impairment). A parent,teacher, or interviewer is asked to rate thechild on this scale where deciles are accompaniedby a descriptor (e.g., 51–60, “somenoticeable problems”). Previous studies havedemonstrated some evidence of reliabilityand validity. A cut score is commonly usedin studies of child psychopathology (e.g.,CGAS 70 or below identifies a clinical case).The CGAS was used as one of the criteriafor validating the DSM-IV criteriafor the diagnosis of ADHD (Lahey et al.,1994). Lahey and colleagues used a CGASscore of 60 or less as an indication of significantimpairment associated with symptomsof ADHD. A noteworthy finding ofthis study was the differential results forthe parent and teacher CGAS scores. Theparent CGAS scores were significantlyrelated to symptoms of hyperactivity/impulsivity but not to inattention. TeacherCGAS scores were not significantly relatedto hyperactivity/impulsivity problems.These same teacher scores were, however,related to ratings of academic problems.The Lahey et al. investigation then usedthe relation between teacher and parentCGAS scores and inattention symptoms toshape the DSM-IV criteria for inattentionproblems associated with ADHD.The psychometric properties of theCGAS have been well-studied (see reviewby Schorre & Vandvik, 2004). Of course,the accuracy of CGAS ratings (as is thecase for all ratings) depends heavily on therater’s knowledge of the child’s functioningin a variety of spheres (Weissman, Warner,& Fendrich, 1990). Can parents, for example,validly rate school and peer functioningas is required by the CGAS? Schorreand Vandvik (2004) call for increased consistencyin how clinicians assess and thenrate impairment. Certainly consistency inconceptualizing constructs such as attentionproblems or depression aid in communicationand treatment planning. Suchcould also be the case for assessing impairmentcaused by these problems.An additional consideration is whetherthe best approaches to assessing impairmentare already embedded in rating scalessuch as those reviewed in this chapter. Forinstance, a study by Bird et al. (1990) founda strong association between CGAS scoresand the Total T-score of the CBCL. Themost impaired group had a mean Total Tof 70, the next most impaired group hada mean of 67, the next group produced amean of 59, and the no-diagnosis groupmean was 53 (Bird et al.). Clinical elevationson standard rating scales may, then, providean indicator of impairment. However, Mashand Hunsley (2005) concluded that “Assessmentsof children and adolescents need tofocus not only on specific disorders andproblems but also on specific impairmentsthat may occur in the absence of a diagnosabledisorder” (p. 368). Therefore, it isquite likely and important that measures ofimpairment will see increasing use in clinicalassessment practice (Bird, 1999).

186 chapter 7 Parent and Teacher Rating ScalesBox 7.4Assessing ChangeAssessment of change is important in clinicalpractice, particularly in light of increasingcalls for accountability for the delivery ofhealth care. Throughout this text, our focusis primarily on assessment as a mechanism toprovide an answer to an initial referral questionand treatment recommendations. Quiteobviously, however, assessment is an ongoingprocess throughout treatment as well. If for noother reason, assessment of change should beroutine because research has shown that suchevaluations lead to better fidelity to evidencebasedtreatments, and ultimately, to bettertreatment outcomes (Lambert et al., 2003).Some considerations of assessment tools asuseful for answering referral questions may notapply to assessments of change. Demonstrationof utility for intervention planning and assessmentis a tricky endeavor since psychometricevidence of validity is not clearly applicable toquestions of treatment progress. A measuredesigned for evaluating change, for example, maynot need norms. Consequently, the quality of thenorming sample is not relevant. Moreover, a ratingscale that assesses a child’s tendency towardproblems in various areas may not be sensitiveto changes over a short period of time. Furthermore,it may not be necessary to interpret totalscores if one is interested only in change in theindividual behaviors. Perhaps ironically, goodtest-retest reliability may be undesirable in theassessment of session-to-session change. Therefore,ratings scales of the nature discussed in thischapter may not be suitable for assessing change– at least not in the short-term.Clinicians may wish to assess changethrough instruments that they develop for usewith a specific client with a specific treatmentplan. Such strategies, if true to the treatmenttargets, would have strong validity and utility.However, to the extent that standardized measurescan be used, a larger evidence base onmeaningful ways to assess change will develop.The Eyberg Child Behavior Inventory (ECBI;Eyberg & Ross, 1978) is an instrument that hasa long history of use for evaluating treatmentresults. It should be noted, however, that itsusefulness for assessing particular constructsis not well supported.The ECBI is a 36-item parent rating scaledesigned to assess behavior problems for childrenof age 2 through 17 years (Eyberg &Ross, 1978). Each item is rated in two ways:(1) a Likert-type scale that is used for markingfrequency, and (2) a dichotomous scale thatthe parent uses to identify if the issue is in facta problem.The ECBI has been used as a measure ofconduct-problem behaviors (Burns & Patterson,1990), and it does possess some advantages,including the fact that it provides an indicationof both frequency and severity for individualbehavior problems, which is not common forparent rating scales. This characteristic maymake the scale particularly useful for planningand evaluating treatment. The ECBI alsoproduces total scores for both the frequency(Intensity) and severity (Problem) measures.The available research suggests that theECBI is an example of an instrument that maybe useful for identifying treatment objectivesfor children referred for disruptive behaviorproblems and for evaluating response to treatment.Certainly, it was ahead of its time in itsamenability to evaluations of change. Whenthe ECBI is used for other purposes, such asnorm-referenced assessment of constructs,research to date does not reveal significantevidence of construct validity (using the termas outlined by Anastasi, 1988). The ECBI isan excellent example of a scale that has somevalue only in the hands of the well-informedclinician who applies the scale only in circumstanceswhere it possesses empirical strengths.We recognize that the preponderance ofour attention is devoted to other forms ofassessment rather than assessment of change.Nevertheless, an increased research base onindividualized as well as large scale approachesto evaluating treatments is imperative. Wehave every reason to believe that discussionsof evidence-based assessment will also includeassessments of treatment progress (e.g., Mash& Hunsley, 2005).

chapter 7 Parent and Teacher Rating Scales187ConclusionsParent and teacher rating scales are nowcommon methods for assessing child problems.The quality of parent and teacherrating scales has improved considerably inrecent years. Routinely, scales have nationalnormative samples and provide expansiveinformation about their reliability andvalidity. In essence, rating scales providea time-efficient and reliable method forobtaining assessment information fromparents and teachers.We focused primarily on global scalesthat assess multiple domains of functioningbecause the nature of childhoodproblems is such that dysfunction in onedomain is often associated with problemsin other areas of functioning. Our reviewof rating scales was not intended to beexhaustive but was designed instead tofocus on some of the most commonly usedscales and to illustrate what we feel aresome crucial areas to consider in evaluatingscales for use in a clinical assessment.Also, our overview was not intended toreplace a careful reading of the technicalmanuals of these scales but to highlightsome of the important features of thescales that might influence their use inclinical assessments.Furthermore, the ECBI (Box 7.4) is anexample of a parent rating scale that couldbe used to evaluate change. The OutcomesQuestionnaire-45 (OQ-45) is a questionnairethat has been used as a means to providetherapists with feedback from adultclients as often as after every session (Okiishiet al., 2006). The suitability and feasibilityof such an approach with parents/child clients and in many clinical settingsis uncertain. Therefore, it is likely the casethat the clinician is routinely left to evaluatechange, whether formally or informally.This strategy has the advantage of beingexecuted by someone trained to define anddetect the problems of focus. It has thedisadvantage of being utilized by the veryperson or persons trying to implement anddemonstrate the effectiveness of their therapeuticstrategies. Research has increasinglyaddressed the implications of this approach(e.g., Lambert et al., 2003), but relativelylittle is known. Far less is known about theteacher assessment of changes in behavioraland emotional functioning during and followinginterventions. An exception wouldbe single-case designs tracking behavioralchanges resulting from classroom interventions;many times, these interventions areevaluated by school psychologists or othermental health professionals.Chapter Summary1. Concerns about child self-reports andpracticality have made parent andteacher rating scales commonplacein modern child assessment practice.These tools tend to be a very efficientmeans of gathering clinically relevantinformation.2. Research has indicated that the constructbeing evaluated and the child’sdevelopmental level influence ratingsprovided by parents and teachers andeven the usefulness of such ratings.3. The Behavior Assessment System forChildren (BASC-2) Parent RatingScales (PRS) and Teacher Rating Scales(TRS) have three forms of similar itemsthat span the preschool (2–5), child(6–11), and adolescent (12–21) ageranges. The PRS takes a broad samplingof a child’s behavior in home andcommunity settings, whereas the TRSdoes the same for the school setting.4. The PRS and TRS were developedusing both rational/theoretical andempirical means in combination toconstruct the individual scales.

188 chapter 7 Parent and Teacher Rating Scales5. The BASC-2 measures include a relativelycomprehensive assessment of adaptivefunctioning.6. The Achenbach CBCL and TRF andtheir predecessors have long been consideredone of the premier rating scalemeasures of child psychopathology.7. The CBCL and TRF continue to bea preferred choice of many child cliniciansbecause of its history of successfuluse and popularity with researchers.8. The CBCL and TRF now includeDSM-Oriented scales that are moreclosely aligned to DSM criteria thanthe Syndrome scales of both measures.9. The CSI-4 is unique in its contentbeing explicitly tied to the diagnosticcriteria in DSM-IV. Thus, it provides ascreening of severe forms of childhood

C h a p t e r 8Behavioral ObservationsChapter Questionsl Why have direct observations often beenconsidered the standard by which otherassessment techniques are judged?l What are some of the characteristics ofbehavioral observations that limit theirusefulness in many clinical situations?l What are the basic components of observationalsystems?l What are some examples of observationalsystems that might be used as partof a clinical assessment of children andadolescents?Direct observation of a child’s or adolescent’sovert behavior has held a reveredstatus in the clinical assessment of youth.Frequently, the validity of other methodsof assessment is judged by their correspondencewith direct observations of behavior.In fact, behavioral observation is oftenviewed as synonymous with the practice ofbehavioral assessment (Shapiro & Skinner,1990). There are two primary reasons forthis importance provided to direct observations.First, as the term direct implies,observations of behavior are not filteredthrough the perceptions of some informant.Instead, the behaviors of the childare observed directly. As we have discussedin the chapters on behavior rating scales,information provided by others in thechild’s environment or by the child himselfor herself can be influenced by a hostof variables and biases. This increases thecomplexity of interpreting these types ofassessment by requiring assessors to accountfor these influences in their interpretations.Therefore, direct observations of behavioreliminate a great deal of the complexityin the interpretive process. Second, directobservations of behaviors frequently allow189P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_8, © Springer Science+Business Media, LLC 2010

190 chapter 8 behavioral observationsfor the assessment of environmental contingenciesthat are operating to produce,maintain, or exacerbate a child’s behavior.For example, direct observations can assesshow others respond to a child’s behavior, orthey can detect environmental stimuli thatseem to elicit certain behaviors. By placingthe behavior in a contextual framework,behavioral observations often lead to veryeffective environmental interventions.To illustrate this potential of behavioralobservations, Carroll, Houghton, Taylor,West, and List-Kerz (2006) conducted a studyof 58 students (ages 8 to 11) in which twostudents, one with Attention Deficit HyperactivityDisorder (ADHD) and one withno disorder, were observed for 40 min. Aswould be expected, the children with ADHDshowed more off-task behavior. However, theobservational system documented “triggers”to this off-task behavior. Nearly, half of theoff-task behaviors of students with ADHDcould be attributed to environmental distractions,and over a quarter were precededby specific teacher behaviors.While these characteristics of behavioralobservations make their use an importantcomponent of many clinical assessments,we feel that the importance of directobservation is sometimes overstated. Likeany assessment technique, direct observationshave several limitations. One of theirmajor limitations is that direct observationsare often expensive and time consuming, ifone is to obtain high-quality information.Because of their cost, many assessors simplyeliminate this source of information fromtheir assessment battery. Alternatively,assessors may attempt less rigorous observationsthan are appropriate. For example,an assessor may observe a child interactingon a playground for a 20-min periodand record the child’s behavior in a narrativeform, without clearly specifying whatbehaviors will be observed or how they willbe recorded. These informal observationsare dangerous if the assessor is unaware ofthe severe limitations and potential biasesin the data that are collected and, instead,interprets the data as if they were objective(see Harris & Lahey, 1982a).Another result of the costliness of directobservations is that the development ofmany obser vational systems has ignoredbasic psychometric considerations (Hartmann,Roper, & Bradford, 1979). In theprevious chapters on rating scales, wefocused a great deal of attention on the psychometricproperties of scales such as thedifferent types of reliability that have beenestablished, the information on the validityof the scales, and the normative base withwhich to compare scores. Because of costfactors, few observational systems haveestablished their reliability or validity inmultiple samples. An even more widespreadproblem for observational systems isthe lack of a representative normative samplethat would allow for a comparison of achild’s scores with those from the generalpopulation. As we have discussed in earlierchapters, having norm-referenced scores iscrucial in the clinical assessments of childrenand adolescents, given the rapid developmentalchanges they are experiencing.Even if one were to use an observationalsystem in the most sophisticated manner,direct observations are still limited by (1)the reactivity of the observational setting,(2) difficulties in obtaining an adequatesample of behaviors, and (3) an inabilityto detect internal events such as cognitionsand emotions. Reactivity refers to awell-documented phenomenon that a personwill change his or her behavior whenit is being observed (Kazdin, 1981; Mash& Terdal, 1988). As a result, the sampleof behavior may not be as objective as onewould hope. There is a significant amountof research on factors that influence thedegree of reactivity that results from directobservations (Harris & Lahey, 1982b;Kazdin, 1981). For example, the age of thechild can affect the degree of reactivity,with preschool children showing less reactivityto observation than older children

chapter 8 behavioral observations191(Keller, 1986). Also, steps can be taken toreduce reactivity during observation suchas allowing the child time to get used to(habituate to) the observational settingand reducing the conspicuousness of theobservational system (Keller, 1986). But,even under optimum conditions, reactivityis still likely to affect the results of theassessment to some degree.Another liability of direct observations isthe difficulty of obtaining an adequate sampleof behaviors. There are several facetsto this issue. The first issue involves ensuringthat the sample of behavior is obtainedunder the most ecologically valid conditions;that is, under conditions that will generalizeto other times and situations. Although theissue of ecological validity is most importantfor observational systems that use contrived(analog) conditions (e.g., observing the childin a clinic playroom), it is also important inselecting the natural setting most appropriatefor conducting the observation. The secondissue is that, even if one selects the bestsetting, one must ensure that a large enoughtime frame is used, so that behaviors will berepresentative and generalizable to othertimes and settings. In the previous exampleof a child being observed in a playgroundsetting for a 20-min period, it cannot bedetermined how typical a child’s behaviorwas during this observational period. Heor she may have had an especially goodor especially problematic day on the playground.A third issue, which encompassesboth the selection of settings and adequacyof the observational period, is the difficultyin assessing many behaviors that are veryinfrequent (e.g., cruelty to animals, hallucinations,panic attacks) or by nature covert(e.g., stealing, lying). In most cases, onewould not ethically want to contrive a situationthat would prompt such behaviors, andthe behaviors are often too infrequent to beobserved naturally occurring in the child’senvironment.A final issue in the use of behaviorobservations is the fact that observationsare limited to the assessment of overtbehaviors. They do not provide a meansfor assessing the cognitive, affective, andmotivational components of a child’s functioning(Mash & Terdal, 1988). This doesnot negate the importance of having agood assessment of a child’s overt behaviorin making diagnostic and treatment decisions.However, it has become increasinglyclear that overt behavior is only onepiece of a complex puzzle. Research inseveral areas of child psychopathology hassupported the importance of intrap sychicvariables for both assessing (e.g., Frick,2006) and treating (e.g., David-Ferdon &Kaslow, 2008) children and adolescents.In summary, direct observations areaffected by some factors that often precludetheir use in many clinical settings andlimit the usefulness of the data obtained.We spent a great deal of time reviewingthe factors that affect behavioral observations,not because of a bias against thisform of assessment, but because we havefound that assessors sometimes ignorethese issues. We feel that a clinical assessorshould be aware of these issues in decidingwhether or not direct observations shouldbe included in an assessment battery andshould consider these issues when interpretingobservational data. However, theselimitations should not be considered anygreater than those associated with otherassessment techniques, and the limitationsmust be weighed against some veryimportant advantages of direct observation(e.g., elimination of reporter bias and readytranslation into environmental interventions).Direct observations can be an integralpart of many assessment batteries but,as is the case for all assessment techniques,they also have limitations in the informationthey provide in isolation.In the following section, we discussbasic issues in the development and use ofobservational systems. As was mentionedearlier in this chapter, many clinical assessorsuse informal observational techniques

192 chapter 8 behavioral observationsin their assessment battery without establishinga well-defined system. Unfortunately,the information obtained fromsuch systems is difficult to interpret.Unlike rating scales, there are few standardizedobservational systems that arereadily available for clinical use that havewell-established psychometric properties.Therefore, the next section focuseson basic considerations in designing anobservational system for one’s own clinicaluse. Following this discussion, someexamples of observational systems that arecommercially available or that have beenused in research are reviewed.Basics of ObservationalSystemsDefining Target BehaviorsThe basic components of observationalsystems can be broken down into the what,where, how, and by whom of the system.The first part of developing a system ofdirect observation involves defining whatbehaviors one wishes to observe. Definingthe behaviors of interest first involvesdeciding on the level of analysis one wishesto use (Barrios, 1993). Specifically, thelevel of analysis can be at the level of isolatedbehaviors, at the level of constellationsof behaviors (syndromes), or at thelevel of interactions within a social unit.As an example of the social unit level ofanalysis, many observational systems allowfor the recording of how a child behavesin response to parental behavior and howa parent responds to a child’s behavior(Gelfand & Hartmann, 1984). Also, theexample given previously demonstratesan observational system that focuses onbehaviors by fellow students and teachersthat can influence the on-task behavior of astudent with ADHD (Carroll et al., 2006).Because these systems allow one to documentevents (stimuli) that elicit a behaviorand responses to the behavior that mayhelp to maintain or increase it, this levelof analysis provides important informationon potential targets of intervention. Anexample of a simple antecedent-behaviorconsequence(A-B-C) type of observationis provided in Box 8.1. From this example,it is clear that recording antecedents andconsequences allows one to determine thesequence of events within which a behavioris embedded.After the level of analysis is chosen, onemust operationally define what behaviors,what constellation of behaviors, or whatantecedents/consequences will be observedwithin this window. These definitions aremade prior to beginning a direct observationand must be specified in objective andunderstandable terms in order to reducethe potential for bias and increase the reliabilityof the observation. Some examplesof target behaviors used in observationalsystems are described in Table 8.1.The target behaviors in Table 8.1 aresimply lists of behaviors from severaldomains that can be assessed by observationalsystems. In order to be reliable,coding systems must have very explicitdefinitions of each behavior. This is necessaryto reduce the possibility that theobserver will use subjective and idiosyncraticdefinitions of the behaviors,thereby making interpretations from theobservations difficult. Without such definitionsthe primary advantage of directobservations, objectivity, is severely compromised.One would think that behaviorssuch as those in Table 8.1 are easy todefine and that simple definitions wouldlead to different observers being able tocode the same behavior in the same way.Decades of research have found that thisis not true. To reliably code behaviors,one must develop very detailed definitions.Box 8.2 provides an example of avery detailed definition of behavior froma frequently used coding system.

chapter 8 behavioral observations193Box 8.1A Hypothetical Example of Simple A-B-C Observational System of an8-Year-Old Boy (B)Time/Setting Antecedent Behavior Consequence8:30/Math classcopyingfromboard8:35/Math classdoingseatwork8:55/Math classcompletingseatworkChild ignoreshimTeacher reprimandsBTeacher ignoresBTeacher puts B’sname on boardB takes pencil fromanother childB tears paper on child’sdeskB sulksB leaves seat to sharpenpencilB raises handB gets out of seat and pullson teacher’s shirt to getattentionB starts to cryChild ignores himChild tells teacher andteacher reprimands BTeacher allows B to eraseboardTeacher asks B to raise handto leave seatTeacher continues to workwith other studentTeacher scolds B for leavingseat and places name onboardChild teases BChild teases B B tries to hit other child B sent to officeB returns toclassB sullen and refuses toworkTeacher allows B to collectassignmentsTable 8.1 Examples of Target Behaviors from Several Behavioral DomainsADHD(Carroll et al.,2006)ConductProblems(Patterson,1982)SocialCompetence(Dodge, 1983)Depression(Kazdin, 1988)Autism (Lord,Rutter, DiLavore,& Risi, 1999)Off task Noncompliance Solitary play Talking Asking for helpFidgeting Destructiveness Cooperative play Playing alone Symbolic playInappropriate Aggressive play Smiling Negativism Taking turnstalkingGazing around Insults/threats Compliments Frowning Reciprocal playOut of seat Aggression Rule making Complaining Telling a sequentialstoryLoud talking Arguing Turn taking WhiningTeasing

194 chapter 8 behavioral observationsBox 8.2Criteria for “Whine” from theDyadic Parent–Child Coding SystemDefinitionA whine consists of words uttered by thechild in a slurring, nasal, high-pitched, falsettovoice.ExamplesWhen can we go home?Mommy, I hurt my finger.I have to go to the bathroom. This is toohard.I don’t want to play this anymore.Guidelines1. The voice quality of the word or phraseis the primary distinguishing element forcoding whine.2. Each whined sentence constitutes a separatewhine. Whined phrases separatedfrom one another by a pause of 2 s orlonger are coded as separate whines.Examples:Child: I have a headache. I want to go home.(2 whine)Child: I don’t like the red blocks… 2-spause… and I don’t like the Legos.(2 whine)Child: Please let me take it home… 2-spause… Please. (2 whine)3. The content of the word or phrase maybe anything except smart talk.Examples:I don’t like this anymore.I hate you.I feel sick.You make me sick.You hurt my feelings.You’re a jerk.(whine)(smart talk)(whine)(smart talk)(whine)(smart talk)4. Whining is a verbal behavior and canoccur simultaneously with a nonverbaldeviant child behavior (destructive orphysical negative child).Decision Rules1. When uncertain as to whether the child’svoice quality is actually a whine or normalvoice quality, do not code whine.2. When uncertain as to whether a child’sverbalization is a whine, smart talk, or acry, code whine.3. When uncertain as to whether the deviantbehavior is a whine or a yell, code yell.Source: Summarized from the manual forthe Dyadic Parent–Child Interaction CodingSystem (Eyberg & Robinson, 1983) with theauthors’ permission.SettingOnce the target behaviors are defined,the next decision is to determine where toobserve these target behaviors. Naturalisticobservations involve observing the child inhis or her natural setting (e.g., in the classroom,at home). The kind of behaviors ofinterest (e.g., social interactions duringfree play) often determines what naturalsetting is best to conduct the observation(e.g., on the playground). In its purestform, naturalistic observations involveplacing no constraints on a child’s behaviorother than those naturally occurring in theobservational setting. However, sometimesit is necessary to place some restrictions onthe observational setting to enhance thequality of the observations. For example,an observer who is in the home of a childto observe parent–child interactions mayneed to place some constraints on the childand parents to ensure that there are sufficientopportunities to observe interactionsduring the observational session. Forexample, one may wish to place restrictionsthat parents and children must stayin the same room and that there is no talkingon the telephone, working on a computer,playing video games, or watchingTV. Another example is an observational

chapter 8 behavioral observations195system designed to observe a child’s anxiousbehavior. The observer may wish theteacher to “create” a situation that seems tolead to anxiety in the child, such as beingcalled on in class or taking a test, in orderto observe the child’s response.In Box 8.3, we provide an exampleof a study by Ostrov and Keating(2004) in which observationsof aggressive behavior of preschool childrenwere observed in both free playand during several structured interactions.Both types of observations wereconducted in a naturalistic setting (i.e.,the child’s school). The use of differentobservational situations within the samestudy allowed the authors to determinethe types of settings in which aggression ismost likely to occur.Naturalistic observations are oftenpreferred because they generally providemore ecologically valid data. However,time and cost constraints may prevent onefrom conducting a naturalistic observation.For many clinical assessments, it is oftenimpossible for the assessor to make severalBox 8.3An Observational Study of Preschool AggressionOstrov and Keating (2004) reported a studyof aggression in preschool children usingnaturalistic observations in the child’s schoolsetting. This study provides a good exampleof two common types of naturalistic observationaltechniques, one in which no restrictionsare placed in the natural setting (i.e., free play)and one in which the situation is structured(i.e., coloring task). The study involved 46children (mean age of 64 months) in ruralpreschools.What: The observation coding systemfocused behaviors in four main categories: (1)physical aggression: hitting, pushing, pulling,punching, forcibly taking objects; (2) verbalaggression: teasing, calling mean names, verbalthreats of harm, insults; (3) relational aggression:excluding from play group; spreadingrumors, withdrawing friendship; telling lies;ignoring peer; (4) number of male and femaleplaymates: number of children of each sex theobserved child directly interacted.Where: Free play observations were conductedduring regularly scheduled free playperiods in large indoor playrooms, in classrooms,and outdoors on the playground. Forthe coloring task, pairs of children were givena series of three pictures to color. However,the potential for mild conflict was introducedby providing one colorful crayon and onewhite one.How: Each observational session was 10min and every instance of the specified behaviorswere coded. Each child was observed forfive sessions. Behavioral counts were summedacross observational periods to determine ascore in each of the four behavioral categories.By Whom: Observers were three femaleand one male undergraduate students whowere trained on the observational system.Prior to conducting observations, observerswere introduced to the teacher and studentsand they spent a few days in the classroom tolet the students adjust to their presence.Results: Boys exhibited more physicalaggression but girls displayed more relationalaggression. Aggression was less overall andthese gender differences were less pronouncedduring the coloring task. However, there wasfairly high stability in a child’s level aggressionacross contexts.Summarized from: Ostrov, J. M., &Keating, C.F. (2004). Gender differences inpreschool aggression during free play andstructured interactions: An observationalstudy. Social Development, 13, 255–277.

196 chapter 8 behavioral observationshome visits to observe a child or adolescentinteracting with his or her parents. Also,for some behaviors, there may not be a wayof obtaining unobtrusive observations ina child’s natural environment. As a result,the level of reactivity would be so high thatthe data would be meaningless. In additionto these more practical considerations,sometimes there is a need to exert morecontrol over the situation than is possible in anatural setting. For example, one may wishto observe a child’s activity level in a freeplay situation by determining how manytimes a child passes from one part of a roomto another. To code this reliably, one candivide the room into sections with tapeand then code the number of times a childcrosses over a tape divider (Milich, 1984).This type of control (e.g., dividing theplayroom into grids) may not be feasiblein a child’s natural environment.For these reasons, it is sometimes necessaryor desirable to conduct analog observationsin a laboratory or clinic. Analog refersto the creation of a contrived setting thatapproximates the natural environment.Dividing a clinic playroom into grids toobserve a child’s activity level is one exampleof an analog setting. However, thekey to these observations is how well theanalog situation approximates the naturalenvironment. Staying with our example,it would be imperative that the playroombe similar to a play area that a child wouldbe in outside of the clinic (e.g., with ageappropriatetoys available). There aremany other examples of analog settings forbehavioral observations, but each involvesthe basic component of simulating a child’snatural environment in a clinic setting.Sometimes it is not feasible to have theclinic setting approximate the natural setting.In these cases, children may be askedto imagine themselves in a situation, andtheir behavior is observed in this role-playsituation. An area in which role play observationshave been frequently used is theassessment of children’s social competence(e.g., Bornstein, Bellack, & Hersen, 1977;Dodge, McClaskey, & Feldman, 1985). Forexample, Dodge et al. (1985) had childrenpretend that they were in certain social situationsand then pretend that the assessorwas another child. An explicit coding systemwas developed to code the degree ofsocial competence of a child’s behavior ineach of the imagined situations. An exampleof one of the role-play situations usedin this study is included in Box 8.4.Data CollectionThe next stage in developing an observationalsystem is to determine how one willcode the target behaviors in the selectedsetting. There are several data collectionmethods that can be used, with the methodof choice depending on the characteristicsof the behaviors of interest. Although thereare many variations of these basic data collectionmethods, the techniques can belargely placed into three categories: EventRecording, Duration Recording, and TimeSampling.Event RecordingEvent recording is the simplest of the datacollection methods. It involves recordingthe number of times that a target behavioroccurred during preset intervals orduring an entire observational session.This method was illustrated in the studyof preschool aggression described in Box8.3 (Ostov & Keating, 2004). Due to itssimplicity, event recording is the mostfrequently used method of direct observation.However, to use event recording,target behaviors must have discrete beginningsand endings, such as hitting anotherchild, raising one’s hand, and asking toplay a game (Shapiro, 1987). In contrast,behaviors that are continuous and persistfor long periods of time are more difficultto code using event recording because it

chapter 8 behavioral observations197Box 8.4A Role-Play Situation from the Dodge Study of Social CompetenceSituation #2Let’s pretend that I’m playing blocks with someof my friends after lunch. We’re building a reallyneat house. You come in the schoolroom and seeus. Pretend that you really want to play blockswith us. what do you do and say?a. Was the child role playing?b. How competent was the response? Score:8-Complimentary or evaluative remark with are-quest to play: “Boy, that’s neat.Can I play?”6—A simple request to play: “I’d ask, MayI please play?”4—Rhetorical question or evaluativeremark: “What are you doing?” or “That’sneat.”2—Suggestion for different activity:“Want to play a game?”0—Aggressive responses: “I’d knock theblocks over”, “I would say nothing or sitdown at desk without speaking”: “I do’ntknow what to do”, or no answer.------------ ------------ ------------ ------------0 2 4 6 8VeryIncompetentSomewhatIncompetentNeither Competentnor IncompetentSomewhatCompetentVeryCompetentReproduced with permission of authors from the Scoring System for Child Role Plays: Role Playing Criteriaused in Dodge, K. A., McClaskey, C. L., & Feldman E. (1985). Situational approach to the assessment of socialcompetence in children. Journal of Consulting and Clinical Psychology, 53, 344–353.is difficult to distinguish the occurrenceof one incident from the next. Examplesof such continuous behaviors are off-taskbehavior, talking out loud, and engagingin solitary play. Event recording is especiallyuseful for recording behaviors thatoccur only briefly and for recording lowfrequencybehaviors that only occur onceor twice in an observational period, such asswearing or hitting another child (Keller,1986; Shapiro, 1987).behavior. In duration recording, theobserver records the length of time fromthe beginning to the end of an instance ofbehavior. Duration and event recordingcan be combined to provide an even richersource of information (Shapiro, 1987). Forexample, in observing the temper tantrumsof a young child, it may be helpful to recordnot only the frequency of tantrums withina given period, but to also record the durationof each tantrum episode.Duration RecordingFor some assessments, it may be moreimportant to know how long a behavioroccurs rather than the frequency of theTime SamplingIn both event and duration recording techniques,all instances of the behaviors arerecorded during the observational period.

198 chapter 8 behavioral observationsHowever, some behaviors occur too frequentlyto obtain accurate counts, or thereare no clear beginning or end points to thebehaviors, which prevents effective eventand duration recording. For these types ofbehaviors, the observation period can bedivided into predetermined intervals, andthe behaviors are simply coded as beingeither present or absent during each interval.Therefore, rather than yielding anexact count of the number of times that abehavior occurred during the observationalperiod, time sampling allows one to determinethe proportion of intervals in whichthe behavior occurred.Shapiro (1987) reviews three types oftime-sampling techniques. In whole-intervalrecording one codes a behavior as presentonly if it occurs throughout a time interval.For example, an observational period in achild’s classroom can be broken down into20-s intervals and the number of intervalsin which the child remained on task forthe entire interval is recorded. In partialintervalrecording, one records whether ornot a behavior occurred at any time duringthe interval. Shapiro gives the example of ateacher dividing the day into 15-min segmentsand noting whether or not certainbehaviors occurred during each segment.A final type of time sampling is momentaryrecording, in which one records whether abehavior was present or absent only duringthe moment when a time interval ends.For example, when observing the degreeof social withdrawal of a child, one maydivide an observational period into 60-sintervals and record whether or not a childwas engaged in interactions with otherchildren at the end of each interval.Selecting the ObserversAfter determining what behaviors will beobserved, where the observations will takeplace, and how the observations will beconducted, one still must determine whois best suited to conduct the observation.Having someone who is in the child’snatural setting (e.g., teacher or parent)conduct the observation is often useful innaturalistic observations because it helps tomaintain unobtrusiveness. However, it isoften difficult to teach people in the child’snatural environment how to use a codingsystem and to ensure that it is being usedappropriately.In order to exert more control over theobservational methodology, many observationalsystems require rigorously trainedand monitored observers. Barrios (1993)provides a summary of steps required intraining and monitoring observers (seeTable 8.2). As one can see from this summary,using specially trained observers isquite costly. Such stringent methodologyis feasible only when a large number ofchildren are being observed with the sameobservational method. Therefore, thismethodology may not be optimal in manyclinical settings.One type of observation that is frequentlyused in clinical assessments involvestraining a person to observe his or her ownbehavior. The same steps of selecting targetbehaviors, determining the setting forthe observation, and determining how thetarget behaviors will be recorded are followed.However, in self-monitoring the childis trained to record his or her own behavior.Although self-monitoring has beenused largely with adults, children have usedself-monitoring systems to monitor suchdiverse behaviors as classroom attending,class attendance, talking out in class, roomcleaning, aggression, and inappropriateverbalizations (Mash & Terdal, 1988).One method of self-monitoring thathas begun to receive some attention inresearch is the use of electronic diaries.That is, with the advent of small computersand personal digital assistants (PDAs),children and adolescents can be taught to

chapter 8 behavioral observations199Table 8.2 Steps in Training and Monitoring ObserversStepOrientationEducationEvaluationApplicationRecalibrationTerminationDescriptionInforming observers of the importance of objective assessment in theunderstanding and treating of childhood disorders. Informing observers oftheir duties and responsibilities, in particular their independent, unbiased, andfaithful recording of the behavior of interest.Instructing observers in the response definitions and recording scheme throughthe use of written materials, filmed illustrations, and live demonstrations.Assessment of observers’ knowledge of the response definitions, coding system,and recording scheme through the use of written and oral examinations.Representation of materials until observers are thoroughly acquainted with allaspects of tracking and recording of the behaviors of interest.Graduated implementation of the observation system across a range of situations,beginning with analog ones and ending with actual setting of interest. Transitionfrom one situation to the next contingent upon observers achieving a criterionlevel of agreement and accuracy.Assessment of the accuracy and agreement of observers’ recordings in the settingof interest. Identification and correction of any breakdowns in the fidelity ofobservers’ recordings.Questioning observers as to the merits of the observation system. Informingobservers of their contributions to the understanding and treating of the behaviorsof interest. Reminding observers of the need to maintain confidentiality.Reproduced with permission from Barrios (1993). Direct observation. In T. H. Ollendick and M. Hersen (Eds.),Handbook of child and adolescent assessment (pp. 140–164). Boston: Allyn & Bacon.carry a small PDA which cues the child torespond to certain prompts and questions.This allows the child to record responsesin real time (i.e., when it is actually occurring).For example, Whalen et al. (2006)had 52 children (mean age of 10.58) reporton their moods, behaviors, and socialcontexts every 30 min during nonschoolhours. The children carried a PDA thatbeeped every 30 min to signal it was timeto respond to certain questions. The children’sresponses were saved in the PDAfor later analyses. Twenty-seven of thechildren were diagnosed with ADHD andthe results showed that these children hadmore behavioral problems, negative mood,and conflict with parents.Research suggests that children canself-monitor their behavior accuratelyif they are trained appropriately, have aclear and simple observational system,have an outside monitor of the accuracyof their recording, and are reinforced forthe accuracy of their recording (Keller,1986). The advantages of self-monitoringare that it is cost effective and is lessintrusive than many other forms of behavioralobservation. However, researchclearly suggests that children change theirbehavior as they become more aware ofit through self-monitoring (Keller, 1986;Shapiro, 1987). Whereas this change inbehavior may be a beneficial aspect ofself-monitoring in a treatment program,this reactivity limits the usefulness of selfmonitoringas a means of obtaining objectiveinformation on a child’s behavior forassessment purposes.

200 chapter 8 behavioral observationsExamples of ObservationalSystemsIn contrast to behavior rating scales, thereare few observational systems that arewidely used, have standardized procedures,and are readily available for clinical use. Inthis section, we review some notable exceptions.The goal of this overview is not toprovide an exhaustive list of observationalsystems but to provide a carefully selectedlist of observational techniques that varyin terms of target behaviors, settings ofobservation, method of data collection,and degree of training needed to reliablyuse the observational system. We feel that,even if one does not choose to use one ofthe specific systems discussed here, thesesystems provide concrete examples of someof the issues discussed in this chapter andtherefore can serve as a guide for the developmentand use of other observational systems.Also, some additional observationalsystems that focus on a specific areas ofadjustment are reviewed in other chapters.Achenbach System ofEmpirically Based Assessment:Direct Observation Form andTest Observation FormThe ASEBA system of assessments containstwo observational systems. The firstis the Direct Observation Form (DOF;Achenbach, 2001) which is designed foruse with children and adolescents ages 5 to14. It provides a method of coding observationsin academic classrooms and othergroup activities. The Test ObservationForm (McConaughy & Achenbach, 2004)is designed for use with children and adolescentsages 2 to 18 and allows for the codingof behavioral observations during theindividual administration of standardizedability and achievement tests. Both of thesesystems are part of the ASEBA system andare designed to be interpreted in conjunctionwith the parent, teacher, and child selfreportversions of the ASEBA, all of whichhave been discussed in previous chapters.The DOF is designed to provide adirect observation of a child in a classroomor group setting during a 10-min period.There are three parts to the DOF. First,the observer is asked to write a narrativedescription of a child’s behavior throughoutthe 10-min observational period, notingthe occurrence, duration, and intensityof specific problems. Second, at the end ofeach minute the child’s behavior is codedas being on- or off-task for 5 s. Third, atthe end of the 10-min period, the observerrates the child on 96 behaviors that mayhave been observed during the observationalperiod using a 4-point scale (from 0= behavior was not observed to 3 = definiteoccurrence of behavior with severe intensityor for greater than 3 min duration).The 96 problem behaviors on the DOFhave a high degree of item overlap withthe behaviors rated on the parent andteacher rating scales of the ASEBA system.Therefore, the DOF is nicely suitedfor a multimodal assessment of a child’sor adolescent’s emotional and behavioralfunctioning. Like the other parts of theASEBA system, the DOF can be used tocalculate a Total Problem score, which isa sum of the ratings of all 96 problems,two broadband scales (Internalizing andExternalizing), and six narrowband scales(Withdrawn-Inattentive, Nervous-Obsessive,Depressed, Hyperactive, Attentiondemanding,Aggressive) (Achenbach,2001). The DOF does report norms from arelatively small sample of 287 nonreferredchildren (Achenbach, 2001).There is evidence that the DOF canbe used reliably by observers with minimaltraining. Inter-observer correlationshave been calculated on the Total Problemsscale in several samples. Correlationsbetween observers range from .96 in a

chapter 8 behavioral observations201residential treatment center (Achenbach,1986) to .92 in a sample of boys referredfor special services in school (Reed &Edelbrock, 1983) to.75 in a sample of outpatientreferrals to a child psychiatry clinic(McConaughy, Achenbach, & Gent, 1988).Inter-observer correlations for the On-taskscores were .71, .71, and .88 in the threesamples, respectively. Reed and Edelbrock(1983) reported inter-observer reliability intheir sample of 25 boys for a selected set ofindividual items from the DOF. In general,most items showed high inter-observercorrelations (most above .80), with theexceptions of Nervous, high-strung, ortense (.20); Picks nose, skin, or other partsof body (.52); and Compulsions, repeatsbehavior over and over (.53).Reed and Edelbrock (1983) reported thatthe Total Problems scale and On-task scoresfrom the DOF correlated in expected directionswith teacher ratings of total problemsand adaptive behaviors. In addition, theDOF Total Problems scale and On-taskscores have been shown to differ in normaland disturbed children (McConaughyet al., 1988; Reed & Edelbrock, 1983). Interms of discriminating within disturbedchildren, the evidence for the DOF is lessclear. McConaughy et al. (1988) reportedthat the Total Problems, On-task, andExternalizing scores differentiated childrenclassified with internalizing or externalizingproblems. However, the internalizing scaleof the CBCL-DOF did not demonstratediscriminant validity in this study.The TOF is designed to rate children’sbehavior during an individual standardizedtesting session. It has 125 items thatare rated on a four point scale (0 = “nooccurrence” – 3 = “definite occurrencewith severe intensity or 3 or more minutesduration”). Items are rated by the examinerimmediately after the testing session. Likethe DOF, the TOF has a strong overlap initems with other measures in the ASEBAsystem. Thus, a Total Problem, an Internalizing,and an Externalizing compositecan all be obtained from the TOF. Also,it includes five narrow band scales: Withdrawn/Depressed,Language/ThoughtProblems, Anxious, Oppositional, andAttentional Problems.The TOF has a normative sample of3,943 children between the ages of 2 and18, most of which were obtained duringthe standardization of the Stanford-BinetIntelligence Scales-5th Edition (Roid,2003). In general, the TOF scales showgood test-retest reliability (.53–.87) over aperiod of 10 days, adequate interrater reliability(.42–.73), and good internal consistency(.74–.94) (McConaughy, 2005).Scores on the TOF are moderately correlatedwith corresponding parent completedASEBA (.27–.43) and teacher completedASEBA (.26–.38) scales (McConaughy,2005). Also, the TOF scales have differentiatedchildren with ADHD from normalcontrol children (McConaughy, 2005).In summary, the TOF and DOF areboth time-efficient observational systemsthat require minimal observer training andfit into a multimethod assessment system.Both observational systems have informationshowing some basic levels of reliability. Theyboth also provide norm-referenced scores,although the sample for the DOF norms isvery limited. Also, both systems have provento be useful for discriminating normal fromclinic-referred children. However, their abilityto differentiate within children with emotionaland behavioral problems is less clear.Behavioral AssessmentSystem for Children-StudentObservation SystemThe BASC-2-Student Observation System(BASC-2-SOS; Reynolds & Kamphaus,2004) is a commercially available, short (15-min) observational system that is designedfor use in a classroom setting. It is part ofthe comprehensive BASC-2 system, which

202 chapter 8 behavioral observationsincludes parent and teacher behavior ratingscales and a child self-report form, allof which have been discussed in previouschapters. The BASC-2-SOS defines 65specific target behaviors that are groupedinto 13 categories (4 categories of positive/adaptivebehaviors and 9 categories ofproblem behaviors). The 13 categories andexamples of target behaviors in each categoryare provided in Table 8.3.Table 8.3 Behavioral Categories from theStudent Observation System of BASC2Category/DefinitionCategory/DefinitionResponse to Teacher/Lesson (appropriateacademic behaviorsinvolving teacheror class)Peer Interaction(appropriateinteractions withother students)Work on School Subjects(appropriateacademic behaviorsthat studentengages in alone)Transition Movement(appropriatenondisruptivebehaviors whilemoving from oneactivity to another)InappropriateMovement (inappropriatemotorbehaviors that areunrelated to classroomwork)Example of SpecificBehaviorsFollows directionsRaises handContributes to classdiscussionWaits for help onassignmentPlays with other studentsInteracts in friendlymannerShakes hand with otherstudentConverses with others indiscussionDoes seatworkWorks at blackboardWorks at computerPuts on/takes off coatGets bookSharpens pencilWalks in lineReturns material usedin classFidgeting in seatPassing notesRunning aroundclassroomSitting/standing on topof deskCategory/DefinitionInattention(inattentive behaviorsthat are notdisruptive)Inappropriate Vocalization(disruptivevocal behaviors)Somatization(physical symptoms/complaints)Repetitive MotorMovements (repetitivebehaviors thatappear to have noexternal reward)Aggression (harmfulbehaviors directedat another personorproperty)Self-InjuriousBehavior (severebehaviors thatattempt to injureone’s self)Inappropriate SexualBehavior (behaviorsthat are explicitlysexual in nature)Bowel/BladderProblems (urinationor defecation)Example of SpecificBehaviorsDaydreamingDoodlingLooking around roomFiddling with objects/fingersLaughing inappropriatelyTeasingTalking outCryingSleepingComplaining of notfeeling wellFinger/pencil tappingSpinning an objectBody rockingHumming/singing tooneselfKicking othersThrowing objects atothersIntentionally rippinganother’s workStealingPulling own hairHead bangingBiting selfEating or chewingnonfood itemsTouching others inappropriatelyMasturbatingImitating sexual behaviorWets pantsHas bowel movementoutside toiletFrom C. R. Reynolds and R. W. Kamphaus (2004). Behaviorassessment system for children – 2nd edition(BASC-2). Circle Pines, MN: American GuidanceServices.The BASC-2-SOS was designed tobe completed during a 15-min observationof the child in an academic classroom.

chapter 8 behavioral observations203The behaviors during the observation arecoded in three parts. In Part A, each ofthe 65 behaviors are rated as being “NotObserved,” “Sometimes Observed” or“Frequently Observed.” Also, any behaviorjudged to be disruptive is noted. Part Buses a momentary time-sampling approachin recording data. The 15-min observationalperiod is divided into 30 intervals. Atthe end of each 30-sec interval the child’sbehavior is observed for 3-sec. A checklistallows the observer to mark each categoryof behavior that occurred during the3-sec observation interval. In Part C, theobserver is asked to describe the teacher’sinteractions with the student, focusing oncontingencies in the classroom that may beinfluencing the child’s behavior.The BASC-2-SOS is a simple and timeefficient observational system that assesses,through direct observation, many behaviorsthat are crucial for the clinical assessmentof children and adolescents. It is one ofthe few direct observational systems commerciallyavailable. Also, the BASC-2-SOShas an electronic coding format in whichbehavioral observations can be recordeddirectly into a PDA device.However, the BASC2-SOS lacks anumber of crucial psychometric elements.First, there is no information onthe reliability of the system provided inthe manual. Establishing inter-observeragreement is a crucial component indeveloping an observational system, toensure that observations are objective andrelatively free from bias. Second, there areno norms for the BASC-2-SOS, so normreferencedinterpretation of scores is notpossible. Third, there is limited informationon the validity of BASC-2SOS, suchas whether or not the BASC-2-SOS codecategories correlate with clinically importantcriteria (e.g., diagnoses, behaviorrating scales, response to intervention).Lett and Kamphaus (1997) reported thatscores on the BASC-2-SOS did differentiatechildren with ADHD from normalcontrol children. However, the limitedinformation on the psychometric propertiesof the BAS-2-SOS greatly limits itspotential contribution to many clinicalassessments.Behavioral Avoidance TestsBehavioral Avoidance Tests (BATs) havebeen used to observe a person’s behavioralresponse to anxiety-producing stimulisince the early 1900s (Jersild & Holmes,1935). Although there are many differentversions of BATs (e.g., Morris & Kratochwill,1983; Van Hasselt, Hersen, Bellack,Rosenblum, & Lamparski, 1979), they allinvolve exposing the child or adolescentto some feared stimuli (e.g., animal, dark,stranger, heights, blood), then requiringthe child to approach the feared stimuli ingraduated steps.BATs provide explicit and objective criteriafor observing a child’s behavioral reactionto the feared stimulus, such as howclosely the stimulus is approached, thenumber of steps in the gradual approachthat are taken, or the time spent touchingor handling the phobic stimulus. In quantifyingthese responses to anxiety-producingstimuli, BATs provide a measure of theseverity of a child’s anxiety and can helpdocument changes brought about by interventions(Vasay & Lonigan, 2000).Southam-Gerow and Chorpita (2007) providea good summary of the advantages anddisadvantages of BATs. They describe theprimary disadvantage of BATs as the absenceof a single standardized BAT. Instead, therehave been numerous different BATs developedthat vary widely on the number ofsteps in the graduated approach, the types ofinstructions given to the child, and how thefeared stimulus is presented. As a result, itis impossible to compare the findings acrossstudies and therefore it is impossible todevelop a significant body of knowledge onthe reliability, validity, and normative base

204 chapter 8 behavioral observationsfor any of the BATs. In addition, a hallmarkof the BATs is the rigorous control overhow the feared object is presented to thechild and how the child approaches it. Thisdegree of control may prevent the behaviorobserved in the contrived setting from generalizingto the child’s natural environment.One final limitation of BATs is that they aremost commonly used to assess specific fearsand are more difficult to design for assessingmore generalized anxiety (Vasay & Lonigan,2000).On the positive side, however, BATs arerelatively simple and time efficient in theiradministration. Many of the BATs havebeen shown to have good inter-observeragreement with minimal observer trainingand to be sensitive to treatment effects(Barrios & Hartmann, 1988; Southam-Gerow & Chorpita, 2007). Also, scoresfrom BATs are correlated with subjectiveratings of fear and with phobia diagnoses(Vasay & Lonigan, 2000). Most importantly,they provide one of the only methodsof assessing the behavioral components(i.e., a child’s avoidance of a feared stimulus)of childhood anxiety.Conflict Negotiation TaskThe Conflict Negotiation Task wasdesigned to assess peer interactions, especiallythose interactions that may be associatedwith childhood depression (Rudolph,Hammen, & Burge, 1994). Children areobserved with an unfamiliar partner ofthe same age and gender. The system usesa task involving three points of potentialconflict of interest between the child andhis or her partner. First, child dyads areplaced in a situation in which they areto build structures with colored blocksto match either of two models. Theyare informed that whoever constructs anidentical model would win a prize. Thedyad is given a set of blocks to share butthe number of blocks is only sufficient tobuild one complete model. Second, after10 min of observation, the dyad is askedto decide on how to distribute two prizesof unequal value. Third, the dyad participatesin a 5-min interview and the childwho received the less valuable prize isallowed to choose a new one.The interactions during this task arecoded by trained observers. Using an eventrecording system, behaviors are rated ona seven-point scale (0 = not at all present;4 = moderately present; 7 = to a largedegree present) and all ratings are scaledsuch that high scores indicate more negativepeer interactions. The behaviors aregrouped into four composites. Two compositesare related to broad dimensions ofsocial behavior displayed by the child beingassessed. Conflict-resolution competenceincludes persistence in problem-solvingefforts, positive assertiveness, positive conflictmanagement, and general social competence.Emotional regulation includesconflict exacerbation, positive affect, andnegative affect. The third composite consistsof a dyadic quality code based on ratingsof conflict or friction within the dyad,collaboration, problem-solving competenceof the dyad, and mutuality/reciprocity.Finally, a peer response code is basedon the ratings of the peer’s behavior towardto the assessed child, including generalresponse valence (negative or positive) tothe target child, discomfort and embarrassmentin response to the target child, andemotional state at the end of the interaction.Rudolph et al. (1994) reported on theuse of this system in a sample of 36 children(20 girls and 16 boys) between theages of 7 and 13. The four composites fromthis observational system showed very highcorrelations between raters (.88 to .92)and the behaviors within each compositewere highly intercorrelated (.82 to .97).Most importantly, when the 36 childrenwere divided into those high on a measureof depression and those low on thismeasure, depressed children were rated as

chapter 8 behavioral observations205significantly less competent on all four ofthe composites. Specifically, children highon depression were observed to be significantlyworse in their conflict-resolutioncompetence, in their emotional regulation,in mutuality and cooperation duringthe interaction, and in the negative valenceplaced on the interactions by their partners(e.g., their partner being more uncomfortablein the interactions). The authors pointout that these data support the contentionthat children who score high on depressionhave significant interpersonal difficulties.Furthermore, the findings suggestthat observational systems of children’sinter actions with peers can be useful forassessing important aspects of these socialdifficulties.Dodge’s Observation of PeerInteractionsDodge (1983) developed a direct observationsystem that also assesses several componentsof peer interactions. However, thisobservation system focuses on behaviorsassociated with acceptance in a peer group.Dodge developed his system in a sampleof 5-, 6-, 7-, and 8-year-old boys. Childrenwere observed in 60-min play groupsof eight boys each by three observers whowere stationed behind a one-way mirror.The boys wore numbered T-shirts to aid inquick identification by the observer. Eachobserver coded the behaviors of one boyfor a 6-min period and then coded a secondchild according to a prearranged schedule.There were 18 target behaviors of fivetypes that were defined for the observationsystem (see Table 8.4). A complex eventrecording system was used for the observations.Each time a target behavior wasobserved, the observer coded the time, thecontext (structured vs. unstructured), thetarget behavior observed, and the peer target(number of child). Observers receivedextensive training over a 4-week periodTable 8.4 Target Behaviors in the DodgeObservational System of Peer InteractionsBehavior CategorySolitary activeInteractive playVerbalizationsPhysical contactwith peersInteractions withadultTarget BehaviorsSolitary playWatching peersOn-task behaviorOff-task behaviorCooperative playAggressive playInappropriate play(e.g., standing ontable)Social conversationswith peersNorm-settingstatements (e.g., rulemaking)Hostile verbalizations(e.g., insults, threats)Supportive statements(e.g., compliments,offers of help)Exclusions of peersfrom playExtraneousverbalization(e.g., laughs, cheers)HitsObject possession (e.g.,grabbing an objectfrom peer)Physically affectionatebehavior (e.g., holdinghands, hugging)Social conversationwith adult leaderReprimanded by adultgroup leaderFrom K. A. Dodge (1983). Behavioral antecedentsof peer social status. Child Development, 54, 1386–1399.and, with this training, the observationalcodes showed quite high inter-observeragreement (Dodge, 1983). Across the 18target behaviors, 15 behaviors showedinter-observer agreement of 65% or better.

206 chapter 8 behavioral observationsThe three behaviors that showed poorinter-observer agreement were watchingpeers, norm-setting statements, and supportivestatements.Family Interaction CodingSystemOne of the most common uses of behavioralobservations is to observe parent–childinteractions, especially thoseassociated with childhood conduct problems(Frick & Loney, 2000). For example,Patterson and colleagues at the OregonSocial Learning Center have developeda direct observational system designed toassess children’s conduct problems in thehome and to assess the interactional patternsin which the conduct problems areoften embedded. The Family InteractionCoding System (FICS; Patterson, 1982)is composed of 29 code categories thatinclude both child behaviors and parentalreactions to the child behaviors. These categoriesare summarized in Table 8.5. TheTable 8.5 Target Behaviors in the FamilyInteraction Coding SystemApproval High rate a Physicalnegative aAttention Humiliate a PhysicalpositiveCommand Ignore a ReceiveCommandnegative a Laugh SelfstimulationCompliance Noncompliance a TalkCry a Negativism a Tease aDisapproval a Normative TouchDependency a No response Whine aDestructivenessPlay Yell aReproduced with author’s permission from G. R. Patterson(1982). Coerceive family process. Eugene, OR:Castalia.aDenotes aversive behaviors that are included in theTotal Aversive Behavior (TAB) score.goal of the FICS was to observe childreninteracting with family members in naturalhome settings. However, as describedby Patterson (1982), several restrictionshad to be made in the home for the observationalsessions. Specifically, to use theFICS, all family members must be presentduring the pre-arranged observation timeswith no guests present, and the family islimited to being present in two rooms ofthe house. There can be no telephone callsout (only brief answers to incoming calls)and no TV. Finally, there is to be no talkingto the observers during coding.The FICS was designed to have datacoded continuously and to provide asequential account of the interactionsbetween a child and other family members.The behavior of the child and the person(s)with whom the child interacts are coded insequence. After initial coding, many of thechild’s behaviors are summarized in a rateper-minutevariable that combines bothfrequency and duration of the behavior.However, the most frequently used scorefrom the FICS is the Total Aversive Behavior(TAB) score which is a sum of the numberof aversive behaviors which occurredduring the observational session.Patterson (1982) describes a moderatelevel of inter-observer agreement for mostcode categories of the FICS, with the categoriesof Negativism and Self-Stimulationshowing the most questionable levels ofagreement. One week test-retest reliabilityof the TAB was studied in a sample of 27boys and was found to be quite high (.78).The TAB was also found to discriminatebetween families of children referred forbehavior problems and nonreferred familiesand has proven to be sensitive to familyfocusedtreatment for children’s conductproblems (Patterson, 1982). Althoughmost of the individual code categories ofthe FICS can be coded consistently bytwo observers, psychometric informationis generally limited to the global index ofaversive behavior, the TAB.

chapter 8 behavioral observations207In summary, the FICS is an example ofan observation system designed to assess achild’s behavior in the home environmentand to assess parent–child interactionalpatterns in which the behavior is embedded.The FICS is generally most usefulfor younger children (10 and under). Also,most of the psychometric informationavailable for the FICS is for the aversivebehaviors assessed by the TAB. This is notas severe a limitation as one might think,however, given that these aversive behaviorshave proven to be quite important tounderstanding and treating children withconduct problems (Patterson, 1982).Structured Observation of Academicand Play SettingsMilich, Roberts, and colleagues (Milich,1984; Milich, Loney & Landau, 1982;Roberts, Ray, & Roberts, 1984) developedan observational system (StructuredObservation of Academic and Play Settings[SOAPS]) to assess behaviors associatedwith ADHD in a clinic playroomanalog setting. In this system, a clinicplayroom is designed with age-appropriatetoys, four tables, and a floor dividedinto 16 equal squares by black tape. Thechild is placed in two situations. Free Playinvolves the child being placed alone inthe room and allowed to play freely withthe toys. The Restricted Academic PlayroomSituation involves the child beingrequested (1) to remain seated, (2) tocomplete a series of academic tasks, and(3) not to play with any of the toys.Each observational situation lasts for 15min. A combination of event recording andtime sampling is used in this observationsystem. Event recording is used to determinethe total number of grids crossedfor the entire observational period. Thatis, the number of times that a child movescompletely from one square of the dividedroom into another is counted. Eventrecording is also used to determine thenumber of times the child shifts his or herattention from one task to another duringthe entire observational period. A 5-sectime sampling procedure is used to observeother target behaviors. These include theproportion of 5-sec intervals that the childis out of his or her seat, fidgeting, noisy,and on task. In addition, a 5-sec time samplingis used to determine the numberof intervals that the child was observedtouching forbidden toys. Also during theacademic task, the number of items completedis recorded.This observational system is useful inclinical assessments of ADHD for a numberof reasons. First, it is a relatively easy touse observational system. As a result, highinter-observer reliability has been obtainedfor most categories with minimal observertraining. Second, categories from this systemhave been correlated with clinicians’diagnoses of ADHD, they have differentiatedADHD children from aggressiveand other clinic-referred children (Milichet al., 1982), and they have been relativelystable over a 2-year period (Milich, 1984).Third, a modified version of this task hasbeen shown to be sensitive to treatmentwith stimulant medication (Barkley, 1988).ConclusionsIn this chapter, we have attempted to summarizeboth the advantages and disadvantagesof direct observations of behavior aspart of a comprehensive clinical assessment.Although there are some limitations,direct observations are a useful componentto many assessment batteries. Probablythe biggest limitation to the clinical utilityof direct observations is the time and costinvolved in conducting behavioral observationsappropriately. We have attempted tooutline some of the major considerations indeveloping and using observational systemsso that clinical assessors can (1) evaluate

208 chapter 8 behavioral observationsexisting observational systems appropriately,(2) develop their own observationalsystems as needed, and (3) recognize limitationsin the data provided by observationalsystems that are not developed andused in a sound manner.We concluded this chapter by providingan overview of several existing observationalsystems. This overview was notmeant to be exhaustive. The observationalsystems included were specifically chosento provide examples of the various domainsof behavior that observational systems canassess, the different settings in which observationscan be conducted, and the variousmethodologies that can be employed. Twoof the systems are commercially available(ASEBA-DOF/TOF; BASC-2-SOS)and cover a broad array of behaviors. Theother systems reviewed focus on more narrowlydefined dimensions of behavior suchas anxiety, social interaction, aggression, orADHD. Whether or not a clinical assessorchooses to use these specific systems,we feel that concrete examples of observationalsystems help to illustrate the uniquecontribution that behavioral observationscan make to an assessment battery.Chapter Summary1. Direct observations of children’s andadolescents’ behavior are an importantpart of an assessment because they providean objective view of behavior that isnot filtered through an informant.2. Direct observations are also helpful inassessing environmental contingenciesthat affect a child’s behavior.3. Direct observations also have a numberof limitations:a. Conducting observations in a waythat provides valuable information isoften an expensive and time-consumingprocess.b. Because of the cost of obtainingobservations, the development ofobservational systems often hasignored basic psychometric considerationssuch as testing the reliability ofthe system or developing an adequatenormative base.c. Even well-developed observationalsystems are subject to reactivity.That is, persons change their behaviorwhen they are aware of beingobserved, which reduces the validityof the observations.d. Other factors affecting the validityof observational systems include thedifficulty in observing an adequatesample of behavior and the inabilityto observe internal events.4. The basic components of observationalsystems include:a. What-defining target behaviors to beobserved.b. Where-selecting the most appropriatesetting in which to observe thebehaviorc. How-determining how the targetbehaviors will be codedd. Who-determining who shouldobserve the target behaviors5. In defining target behaviors, one mustconsider the level at which behaviorswill be defined and then clearly definethe behaviors to be observed.6. Observations conducted in a child’snatural environment have greater ecologicalvalidity but allow less controlover the observational setting thanobservations conducted in a laboratoryor analog setting.7. There are three basic ways in whichbehaviors can be recorded in an observationalsystem:a. Event recording-the number of timesa behavior occurred is recorded.

chapter 8 behavioral observations209b. Duration recording-the lengthof time from the initiation tothe desistance of a behavior isrecorded.c. Time sampling-behaviors arerecorded as to whether or not theyhave occurred during preset timeintervals.8. Observations can be conducted byoutside observers, people in a child’senvironment (e.g., parents, teachers),or by the child or adolescent himselfor herself.9. The ASEBA-DOF and ASEBA-TOFare two observational systems thatcan be used in conjunction with othercomponents of the ASEBA system.a. The DOF allows for a direct observationof a child’s classroom behaviorduring three to six 10-min observationalperiods.b. The TOF allows for a directobservation of a child’s behaviorduring individual standardizedtesting, and it has strong normativedata.c. Both the DOF and TOF have evidencesupporting their reliabilityand for differentiating referredfrom nonproblem children.10. The BASC-2-SOS is an observationalsystem used to assess classroom behaviorthat can be integrated with otherassessment components of the BASC-2system.a. It specifies 65 target behaviors anduses a momentary time-samplingprocedure for observations.b. There is no normative informationnor is there any information on thereliability or validity of this observationalsystem.11. BATs allow one to observe a child’sresponse to anxiety-provoking stimuli.12. Observational systems have also beendeveloped to assess peer interactions,parent–child interactions, and behaviorsassociated with ADHD.

C h a p t e r 9Peer-Referenced AssessmentChapter Questionsl What contributions can peer-referencedtechniques make to clinical assessmentsof children and adolescents?l What are some ethical concerns in theuse of peer-referenced assessments?l What are the different types of peer-referencedassessments?l What do sociometric exercises measureand why might this be an importantcomponent of clinical assessments?l Besides social status, what other areasof a child’s behavioral, emotional, andsocial functioning can be assessed bypeer nomination techniques?Peer-referenced assessment strategies areassessment techniques in which a child oradolescent’s social, emotional, or behavioralfunctioning is assessed by obtainingthe perceptions of the child’s peers. One of themost common types of peer-referencedassessment is the sociometric assessment,in which the child’s acceptance in or rejectionby his or her peer group is determined.We discuss sociometric techniques in moredepth later in this chapter. However, sociometricassessment should not be consideredsynonymous with peer-referenced assessment.There are many aspects of a child’sadjustment, not just peer social status, thatcan be usefully assessed through the perceptionsof a child’s peers. A sampling ofthe most common psychological domainssuitable for peer-referenced assessmentand the different measurement strategiesare the focus of this chapter.The main reason for using peer-referencedassessment is that it provides importantinformation that cannot be obtainedfrom other sources. The importance of peerperceptions has both an intuitive and empiricalbasis. A child or adolescent’s social milieuis considered a major influence on a child’spsychological adjustment in most develop-211P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_9, © Springer Science+Business Media, LLC 2010

212 chapter 9 Peer-Referenced Assessmentmental theories. Therefore, one of the mostdevastating effects of a child’s behavioraldisturbance is the effect that it may haveon his or her social environment (Mayeux,Bellmore, & Cillessen, 2007). Also, interventionsthat change many aspects of the child’sbehavior may not result in changes in thechild’s peer relationships (Hoza et al., 2005).For these reasons, understanding how achild is viewed by peers is critical in developinga complete picture of a child’s or adolescent’soverall psychological adjustment.Peer-referenced assessment, whether itfocuses specifically on a child’s social relationshipsor indirectly assess a child’s socialmilieu by determining how peers perceivea child’s emotional and behavioral functioning,allows for a better understandingof a child’s social network.The empirical literature supports thistheoretical emphasis on peer relationships.Parker and Asher (1987) conducted a metaanalyticreview of studies that have testedthe utility of peer relationships (primarilyacceptance and aggression) in predictinglater outcomes (dropping out of school,criminality, and psychopathology). Two ofthe major findings of this review were thatlow peer acceptance was consistently relatedto dropping out of school and that peerratedaggression was consistently relatedto delinquency. Similarly, in another studyof 445 girls who were first studied at ages10 to 13, rejected peer status was relatedto increased risk for criminal offending andalcohol abuse almost 40 years later (Zettergren,Bergman, & Wangby, 2006).This literature, therefore, clearly supportsthe importance of peer perceptions in predictingthe negative outcomes of a child and,hence, it illustrates the need to assess and tointervene in a child’s social milieu. Box 9.1summarizes several other interesting findingsfrom the Parker and Asher’s review that haveimplications for the use of peer-referencedtechniques in clinical assessments.The assertion that peer perceptionscannot be obtained by other methodscomes from the meta-analysis conductedby Achenbach, McConaughy, and Howell(1987). These authors calculated the averagecorrelations across the studies betweenpeer reports of social, emotional, or behaviorfunctioning with the reports of teachersand with the child’s self-report. Across23 studies reviewed, the average correlationacross all psychological domains was.44 between peer and teacher ratings, withthe correlation being somewhat higher forbehavioral (.47) problems than for emotional(.35) problems. Similarly, across 20 studies inwhich the correlation between peer ratingsand the child’s self-report of adjustment wasdetermined, the average correlation was .26,again with the correlation for behavioral difficulties(.44) being somewhat higher thanfor emotional difficulties (.31). These datasuggest that there are substantial differencesbetween how peers rate children and howteachers rate children and how childrendescribe themselves. Therefore, to understanda child’s peer network that is heavilyinfluenced by peer perceptions, these perceptionsmust be assessed directly.Ethics of Peer-ReferencedStrategiesDespite research on the unique and importantcontributions that peer-referencedtechniques can make to clinical assessments,these techniques are probably oneof the least used assessment techniquesof any reviewed in this book. The failureto include peer-referenced techniques inmany assessment batteries could be, inpart, due to the paucity of standardized,well-normed, and readily available assessmentprocedures. This exclusion couldalso be due to the time-consuming natureof many of the peer-referenced techniquesused in the research literature. However,as we discuss later in this chapter, there

chapter 9 Peer-Referenced Assessment213Box 9.1Research Note: Meta-Analysis of the Association Between Peer Relationshipsand Later Psychological AdjustmentParker and Asher (1987) completed a comprehensivemeta-analytic review of the predictiveassociation between poor relationshipsin early to middle childhood and later (adolescentand adult) psychological adjustment.As mentioned in the text, this review clearlysupported the importance of a child’s socialcontext in general, and peer perceptions of achild specifically, in terms of predicting lateradjustment.However, in addition to illustrating theoverall importance of peer relationships toclinical assessments of children, this reviewhad several other interesting results that canguide the assessment process. For example,in terms of predictive accuracy, a consistentpattern of errors emerged in which peer-referencedprocedures tended to make few falsenegativeerrors in predicting which childrenwould have poor outcomes, but there weremany false-positive errors. That is, most childrenwho have problems later in life had peerrelationship problems. However, a large numberof children with relationship problems donot show later difficulties. Knowledge of thistype of predictive relationship can be quitehelpful in interpreting peer-referenced assessments.The authors of this review also cautionusers of the literature to be aware ofthe fact that, despite knowing that thereis a predictive relationship between earlypeer relations and later adjustment, we donot know why this relationship exists. Forexample, it could be that because these childrenare excluded from normal patterns ofpeer interactions, they may also be excludedfrom normal socialization experiences anddeprived of important sources of support.However, it is also possible that early formsof a pathological process that may emergemore fully in adulthood may have a negativeinfluence on early peer relationships.In essence, peer relationships could be anaccidental by-product of a pathological processand not really have a causal relationshipwith later adjustment.A final relevant point made by the authors ofthis review is the fact that future research shouldattempt to obtain a more comprehensive pictureof children’s social relationships. For example,the authors found very limited data onshyness and social withdrawal in predicting lateroutcomes, with most studies relying on indices ofacceptance and aggression as the primary aspectsof social relationship to be studied. Other aspectsof peer relationships that could be studied systematicallyinclude impulsive/hyperactivebehavior, bossy and demanding behaviors, andbehaviors that define attributes that approximatehow children choose their friends (e.g., Is thischild fun to play with?).Source: Parker, J. G., & Asher, S. R. (1987). Peer relations and later personal adjustment: Are low-acceptedchildren at risk? Psychological Bulletin, 102, 357–389.are several relatively time-efficient peerreferencedprocedures that have been usedextensively in research which could add toa clinical assessment battery. Therefore,the low frequency of use is probably theresult of other considerations.One such consideration could be the intrusivenessof peer ratings. Peer-referencedassessment typically involves the use ofmany peer raters, making it more intrusivethan many other assessment techniques.For example, asking a teacher to complete abehavior rating scale adds only one person tothe assessment process. In contrast, havinga child’s class participate in a sociometricexercise involves 15–30 additional peoplein the assessment process.In addition to the sheer number ofpeople that must be involved, the peopleinvolved are typically children who may

214 chapter 9 Peer-Referenced Assessmentnot appreciate the need for discretionand confidentiality. Therefore, specialprecautions to limit the intrusivenessof this intervention are essential. Thereare several necessary precautions forthe use of any peer-referenced strategy.First, it is important that both the parentand the child being assessed are clearlyinformed and give their consent to thepeer-referenced assessment (Gresham &Little, 1993). Second, peer-referencedtechniques should be designed to ensurethat the child’s classmates do not knowthat the assessment is focused on anyone individual (see Box 9.2 for an exampleof instructions provided in a sociometricexercise). Finally, administration of peerreferencedtechniques should be carefullymonitored to ensure that answersare not shared, and those involved in theassessment should be instructed on theimportance of the confidentiality of theirresponses after the assessment (McConnell& Odom, 1986).One of the most controversial aspects ofpeer-referenced strategies is the use of negativeratings from peers (e.g., nominationsof children who are not liked or who areaggressive). Teachers and parents are oftenconcerned about the possibility that thesenegative ratings will lead to social rejectionand other negative reactive effects. Fortunately,over 50 years of research usingpeer-referenced strategies has not foundany evidence for these negative effects(McConnell & Odom, 1986). In fact, thepotential for negative effects has been specificallytested (Hayvren & Hymel, 1984;Mayeux, Underwood, & Risser, 2007). Forexample, Mayeux, Underwood, and Risser(2007) completed a sociometric exercisewith 91 third graders and then interviewedthem and their teachers. Their results indicatedthat children were not hurt or upsetby the procedures nor did the participantsfeel that their peers treated them differentlyfollowing the testing.Despite the lack of evidence for reactiveeffects of negative ratings, it would benice to be able to use only positive ratings.Unfortunately, research has clearly shownthat negative ratings are not simply theopposite of positive ratings. Negative ratingsadd crucial additional information tothe assessment. For example, rejected childrenare not simply unaccepted children interms of peer status, but they are activelydisliked by their classmates (Coie, Dodge,& Copotelli, 1982). Therefore, one cannotsimply assess peer acceptance andthen consider those low on acceptance asbeing rejected by peers. As we discuss laterin this chapter, this rejected status, whichrequires negative ratings to assess, is oneof the most important indices of a child’ssocial status. Therefore, negative ratingsappear to be an important part of a peerreferencedassessment strategy. However,clinical assessors should be sensitive toconcerns about negative ratings and assureparents and teachers that appropriatesafe-guards are being used in the assessmentprocedures and explain to them thecritical need for this information in theevaluation.Types of Peer-ReferencedTechniquesPeer NominationsPeer nominations are the oldest and mostcommonly used form of peer-referencedassessment (Asher & Hymel, 1981).The procedure involves asking childrenin a classroom to select one or more oftheir peers who display a certain characteristic(e.g., liked, fights, cooperates, shy,leader). Although this procedure is relativelysimple and straightforward, thereare several variations of peer nominationprocedures. For example, there canbe a predetermined number of childrenthat can be selected in each category (fixedchoiceformat). Alternatively, the numberof peers nominated in each category can

chapter 9 Peer-Referenced Assessment215Box 9.2Instructions for Two Peer-Referenced TechniquesAssessor-Administered Exercisewith Unlimited-Choice Format(Carlson, Lahey, & Neeper, 1984)After discussing the importance of confidentiality,sheets are provided to each child with the class roleand each of 30 nomination categories.Now we’re ready to begin. We have writtendown lots of things kids do, and we would likeyou to check which kids in your class do thesethings. Look across the list of names alongthe top of the page until you find your name.We’ll ask you to tell us about what things youdo later, but for now put a line through yourname and draw a line down the column yourname is in so you will remember not to checkthese things for yourself. Go to the other pageand do this every time you see your name.Now go back to the first page. See the numberone? After the number one, it says, “Thosewho are tall.” Now look across the names. Whois tall? Put an “X” under their name. Who isn’ttall? Put a “0” under their name. Go throughevery name one at a time and put an “X” underit if they are tall, and a “0” under it if they aren’ttall. Be sure to read every name on the top so youdon’t forget to check anyone. When you finishwith number one, you may turn back to the firstpage and wait for the rest of the group to finish.Following the completion of item one, allsubsequent items were completed in a similarmanner, with each item read aloud.Items1. Those who are tall2. Those who say they can beat everybody up3. Those who complain a lot4. Those who bother people when they are tryingto work5. Those who stand back and just watch otherswho are playing6. Those who get mad when they don’t gettheir way7. Those who start a fight over nothing8. Those who act like a baby9. Those who do not follow the rules when theyplay games10. Those who tell other children what to do11. Those who try to change the game when theyjoin in12. Those who help others13. Those who speak softly and are difficult tounderstand14. Those who do not pay attention whensomeone is talking to them15. Those who daydream a lot16. Those who keep talking even whensomeone is talking to them17. Those who ask a lot of questions whenthey join a group18. Those who always talk about themselveswhen they join the group19. Those who do not want your help even if youoffer it20. Those who do not know how to join in thegroup21. Those who show off in front of the class22. Those who share their things23. Those who give in to others too much24. Those who are afraid to ask for help25. Those who often change the subject26. Those who are honest27. Those who do not try again when they lose28. Those who can not wait their turn29. Those who never seem to be happy30. Those who let others boss them around a lotTeacher-Administered Exercise withFixed-Choice Format (Strauss et al.,1988)A. Pass out paper to all of the children in theclass.(Continues)

216 chapter 9 Peer-Referenced AssessmentBox 9.2 (Continued)B. Read the following statement to the children:“Class, I want us to do an exercise nowthat will help me learn more about you aspeople and about your friendships. It willjust take a couple of minutes and shouldbe fun, but if anyone does not want to takepart, you may feel free to sit quietly whilethe rest of us do the exercise. Because thequestions that I will ask are about privatefeelings, I want to ask you not to discussyour answers with each other and not to letyour neighbors see your paper.”C. Then ask the children to list the three childrenin the class that you…1. Like the most2. Like the least3. Think fight the most4. Think are the meanest5. Think are the most shyD. After the exercise is completed, collect thepapers. To protect the feelings of the children,you may wish to briefly look through the collectedpapers and state, “I haven’t had a chanceto look carefully at these, but it looks like everybodywas named as most liked by at least oneperson. That’s really nice.” Reiterate the needfor the children not to discuss their responseswith each other.E. Count the total number of children whoparticipated in the exercise, and the numberof times the child being evaluated wasnamed in response to each of the five questions.Please write this information on theback of the page and mail it in the selfaddressedstamped envelope provided.F. Thank you again for your assistance. If theparent gives us permission to do so, a copyof the evaluation results will be sent to you.It is important to note that the fact that thischild is being evaluated does not necessarilymean that he or she has psychological problems;many of the children that we evaluateturn out to be perfectly normal.Note: Procedures are provided with permission of the authors.be left entirely to the child providing thenominations (unlimited-choice format). Anexample of instructions for both a fixedchoiceand an unlimited-choice formatis provided in Box 9.2. Although someauthors prefer the unlimited-choice formatto avoid forcing a set number of childreninto categories, especially negativecategories (McConnell & Odom, 1986),there is little empirical evidence for any clearadvantage of one format over the other.A second dimension on which peer nominationprocedures can vary is the degree ofexplicitness that defines the nominatingpool. For example, some procedures simplyinstruct the children to consider any childin their class (Strauss, Lahey, Frick, Frame,& Hynd, 1988). In contrast, other nominationstrategies provide children with a rosterof names from which to choose (Coie &Kupersmidt, 1983) or may even providepictures of classmates from which the raterselects nominees for the individual categories(Moore & Updegraff, 1964). The rule ofthumb is that the younger the rater, the moreexplicit the definition of the nominatingpool should be. Also, if the pool of potentialnominees is not within a well-defined group(e.g., only part of a class is participating inthe procedure), then more explicit definitionsare required.Because of the difficulty and level ofintrusiveness involved in collecting peernominations from entire classrooms,Prinstein (2007) compared two alternativemethods for obtaining peer perceptions.That is, this author compared nominationsobtained from a full sample of 232adolescents ages 15–17 years old, withthose obtained (a) using only a randomlyselected subsample of 26 students and (b)using only two students in each class to

chapter 9 Peer-Referenced Assessment217rate other students. These students werechosen by their teachers as “social experts.”The correlations between nominationsobtained in the full classroom and thosecompleted by experts were generally quitehigh (r’s ranging from .55 to 93). Similarcorrelations were found between the fullclassroom procedure and the nominationscompleted by a random subsample(r’s ranging from .49 to 90). Thus, theseresults suggest that there may be someless cost intensive alternatives to obtainingpeer nominations, at least in classroomswith adolescent students.The typical level of interpretation of apeer nomination procedure is the numberof times a child was nominated in a givencategory. This number is then comparedto a normative base for that particularprocedure to see if the child was nominatedat a level that is atypical for childrenhis or her age. However, there are someinstances where more complex combinationsor adjustments of peer nominationsare desired. For some purposes it may beuseful to compare the number of nominationsobtained by a child in one area withthe number of nominations that the samechild received in another area. For example,in sociometric techniques, one oftencompares the number of times a child wasnominated as Liked Least with the numberof times he or she was nominated as LikedMost by classmates. This allows one todetermine the relative balance of two nominationcategories. However, to make suchcomparisons, the two scores should first beconverted to standard scores (e.g., Z-scores;Coie et al., 1982) to equate for possible differencesin the variance of the raw scores.A second type of conversion is warrantedif one wishes to compare nominationsof one child with the nominations ofanother child from another nominatingpool (e.g., different class). To make thiscomparison the number of nominationsmust be adjusted to equate scores for differingclass sizes. For example, 5 nominationsof Most Cooperative in a class of 12 shouldbe interpreted differently than 5 nominationsin a class of 30. As an example of thisconversion, Strauss et al. (1988) dividedthe number of nominations obtained by achild by the number of children participatingin the assessment. The quotient wasmultiplied by 23, so that the nominationswere all expressed in terms of a commonclass size of 23.SociometricsSociometric techniques focus on a specific,important aspect of a child’s peer relationships:a child’s social status. It answers thequestion of whether or not a child is likedand accepted by his or her peer group.Sociometrics do not assess specific behaviorsof the child. It answers the questionof whether the child is liked and not whatis the child like or why the child is liked(Asher & Hymel, 1981). Sociometric exerciseshave appeared in the research literaturesince the 1930s (see Gresham & Little,1993; Hughes, 1990), and the most commonlyused procedure has changed verylittle over this time. An example of thisbasic technique from Strauss et al. (1988)is provided in Box 9.2.Sociometric exercises can take the formof peer ratings, whereby peers rate a childon a Likert scale as to how well liked or dislikedhe or she is (Hamilton, Fuchs, Fuchs,& Roberts, 2000). However, the more commonmethod of obtaining sociometrics isthrough peer nominations. In this technique,children can be nominated by peersas a child who is Liked Most (sometimesdefined as Most like to have as a best friendor Most like to play with) and/or they can benominated by peers as a child who is LikedLeast (or alternatively, Least like to have as afriend or Least like to play with). Althoughthere is no definitive normative study thatspecifies exact cut-offs for when nominationsare considered indicative of problems,

218 chapter 9 Peer-Referenced Assessmentin a fixed response format allowing threenominations in each category in a class sizeof approximately 20 students, Liked Most(LM) nominations of less than two andnominations of Liked Least (LL) of greaterthan four are generally considered indicativeof problems in peer relations (Dodge, Coie,& Brakke, 1982; Green, Vosk, Forehand, &Beck, 1981; Strauss et al., 1988).A common way of interpreting sociometricnominations is by combining theLM and LL nominations into five distinctsocial status groups (Hughes, 1990). Asmentioned previously, when combiningLM and LL nominations, the nominationsshould first be converted to standardscores to equate for potential differencesin the variance of the two categories. Thefive groups are based on two differencescores. The social preference score is thedifference between LM and LL scores(LM − LL = Social Preference). High socialpreference scores indicate substantiallymore LM nominations than LL nominations.The social impact score is the sum ofthe LM and LL scores (LM + LL = SocialImpact). High social impact scores simplydetermine the number of nominations achild receives, regardless of whether theyare negative or positive. A combination ofthese scores leads to a child being consideredin one of several social status groups:Popular, Rejected, Neglected, Controversial,and Average. The method of combiningthese scores to determine a child’s socialstatus and two computational formulas thathave been used in research are provided inBox 9.3.Although there have been several variationsin the formulas for determining aBox 9.3Determining Social Status from Sociometric NominationsCategoryDescriptionComputational FormulaUsing Standard Scores(Coie, Dodge,& Coppotelli, 1982)ComputationalFormula Using RawScores (Strausset al., 1988)PopularHigh social preferencescores (LM − LL) but fewLL nominationsLow social preferencescores with few LMnominationsLow social impact scores(LM + LL) and few nominationsin LM categoryHigh social impact scoresand above-average LMand LL scoresAverage social preferencescores(1) ZLM − ZLL > 1(2) ZLM > 0(3) ZLL < 0(1) ZLM − ZLL < −1(2) ZLM < 0(3) ZLL > 0(1) ZLM + ZLL < −1(2) LM = 0(1). LM > 4.5(2) LL < 1.5Rejected(1) LM < 1(2) LL > 4.5Neglected(1) LM < 1.5(2) LL < 1.5Controversial(1) ZLM + ZLL > 1(2) ZLM > 0(3) ZLL > 0(1) ZLM − ZLL > −.5(2) ZLM − ZLL < .5AverageNote: .Both computational formulas are based on a fixed-choice format allowing for three nominations in boththe Like Most (LM) and Like Least (LL) categories. ZLM refers to LM nominations converted to standardZ-scores and ZLL refers to LL nominations converted to standard Z-scores. The unstandardized formulas arebased on scores adjusted to a class size of 23.

chapter 9 Peer-Referenced Assessment219child’s social status, the validity of thesegroupings has been consistently shown inresearch (see Gresham & Little, 1993),including showing good convergence withstatistical methods for clustering childrenbased on peer nominations (Zettergren,2007). Further research has suggested thatthese nominations are quite stable overtime. Jiang and Cillessen (2005) conducteda meta-analysis of 77 studies of over 18,339participants and reported that the averagelevel of stability for social status over periodsof less than 3 months ranged from r =.70 to r = .80. For periods of over 3 months,the average ratings ranged from .52 to .58.Importantly, research has suggested thatsociometric nominations can be influencedby the racial composition of a classroomwith social preference scores being higherwhen nominations are obtained from sameracepeers (Singleton & Asher, 1979). Jackson,Barth, Powell, and Lochman (2006)reported on 1,268 sociometric nominationsfrom children ages 9 to 11 years old across57 classrooms that ranged from 3 to 95%African-American. The results indicatedthat African-American students nominationswere more sensitive to the racial compositionof the classroom than for Caucasian students.Specifically, African-American students’social preference scores and nominations forfighting and being a leader improved as thepercentage of African-American studentsin the classroom increased. The effects forCaucasian students was less clear, althoughthere was a small effect of Caucasian studentshaving lower preference scores if theclassroom was less than 33% Caucasian.Social status has also been associatedwith emotional and behavioral characteristicsof the child. One of the most consistentfindings is that rejected children showhigher levels of aggressive and acting-outbehavior than nonrejected classmates(e.g., Dodge, 1983). However, neglectedstatus is also associated with problems inadjustment, most notably with anxiety(Strauss et al., 1988). Several behavioralcharacteristics are associated with popularchildren, including being more likely tocontribute to conversations during play,being more likely to engage in parallelplay, receiving and initiating more positivesocial behavior, and using effective peerentrystrategies (Gresham & Little, 1993).Children in the controversial status grouphave been less well studied. However, onestudy suggests that children in this socialstatus group tend to exhibit aggressive anddisruptive behaviors, like the rejected children,but also tend to be viewed as sociallyskilled and as leaders, like the popularchildren (Coie et al., 1982).A distinction that research has increasinglyshown to be important is betweenwhether a child is well liked by their peers(i.e., accepted) and whether the child isviewed as “popular” (Prinstein, 2007). First,these ratings have only been modestly correlatedwith each other (Prinstein, 2007; Vaillancourt& Hymal, 2006). Second, ratingsof greater peer acceptance have consistentlybeen related to more positive behavioral(e.g., less aggression; more prosocial behavior)and emotional (e.g., higher self-esteem)functioning (Gresham & Little, 1993; Sandstrom& Cillessen, 2006). However, childrenperceived by their peers as popular oftenshow more physical and relational forms ofaggression (McDonald, Putallaz, Grimes,Kupersmidt, & Coie, 2007; Vaillancourt& Hymel, 2006). Third, peer popularityseems to be more strongly related to peervaluedcharacteristics such as power, physicalattractiveness, athleticism, and dressthan peer acceptance (Prinstein & Cillessen,2003; Vaillancourt & Hymel, 2006).From this research, it is clear that achild’s social status is intertwined with hisor her behavioral and emotional functioning,both in terms of current and futurefunctioning. Therefore, sociometrics cancontribute important information to manyclinical assessments by providing a reliablemethod of assessing a crucial aspect of achild’s social functioning.

220 chapter 9 Peer-Referenced AssessmentMost sociometric exercises are conductedin school classrooms because theyprovide a very clear and well-defined referencegroup of peers from which to judgea child’s social acceptance or rejection.However, an important consideration inthe usefulness of data obtained from sociometricexercises conducted in classrooms is thelevel of student participation in the exercise.That is, there is evidence that evenmoderate declines in full classroom participationin the sociometric exercise canhave a major influence on the results. Forexample, Hamilton et al. (2000) comparedsociometric results using a peer-rating procedureacross varying levels of classroomparticipation. These authors reported that,even with a 75% rate of classroom participation,there were substantial differencesin the results compared to those obtainedfrom the full classroom. The results for25 and 50% participation rates were evenmore divergent from those obtained withfull participation. Importantly, the instabilityin the results across the varying levelsof participation was greatest for childrenwith adjustment problems compared towell-adjusted children. Therefore, for childrenwith problems, who are often of mostinterest in sociometric exercises, participationrates seem to be especially importantfor interpreting the results. These findingshighlight the need to interpret informationfrom sociometric techniques in the contextof the level of classroom participation inthe exercise. In Box 9.4 additional findingsfrom the Hamilton et al. (2000) study thathave potentially important implications forinterpreting information from sociometricexercises are summarized.Box 9.4Research Note: A Comparison of Sociometric Results Across Varying Levelsof Classroom ParticipationHamilton et al. (2000) investigated the effectsof different rates of classroom participationon peer ratings of social acceptance. Theseauthors reviewed 26 studies using sociometricratings to assess the social acceptance of childrenwith learning or behavioral disorders andfound that the vast majority of studies did notreport the rate of participation in the sociometricexercise and, for those studies that did, therates varied from 67 to 100%. The potentialeffects of these varying rates was investigatedin 14 classrooms (grades 3 through 6) with full,or nearly full, rates of participation (i.e., 92 to100%). The authors used a group sociometricprocedure in which each student was providedwith a class roster. Each student’s name wasaccompanied by four circles closing (1) a smilingface to indicate that the student is liked, (2)a straight-mouthed face to designate that therater is indifferent to the student, (3) a frowningface to indicate that the student is disliked, and(4) a question mark to indicate that the rater isunsure of the student’s likability. Each studentreceived a percentage score for each of the fourcategories by dividing the number of responses(e.g., smiles) by the number of student ratersminus one, thereby creating percentage scoresfor each type of acceptance rating.To assess the effects of classroom participationrates, the classrooms were then sampledrandomly and repeatedly three times for eachlevel) so that 25, 50, and 75% of studentswere involved in the exercise. At each participationlevel, peer ratings were comparedagainst those obtained at the full level ofparticipation. The results were quite consistentindicating that, as the rate of classroomparticipation decreased, ratings tended to beincreasingly divergent from those obtained atfull participation. Even the ratings at the 75%level were significantly different from theratings obtained at full participation (rangingfrom 3 to 18%). Importantly, the authorscompared the ratings of students with learning(Continues)

chapter 9 Peer-Referenced Assessment221Box 9.4 (Continued)disabilities and nondisabled students that wereunderachieving to average and high-achievingstudents. The decline in participation rateswas especially problematic for the ratings ofthe two groups with learning problems.The authors noted that the effects of classroomparticipation on this peer-rating proceduremay not generalize to the more commonpeer nomination procedure, although theymake the important point that participationeffects on nomination procedures need tobe tested in light of their results. They alsooffer an interesting explanation for why participationrates may affect the ratings of lesswell-adjusted children to a greater degree.The authors suggest that the better-adjustedchildren may be viewed more similarly acrossclassmates, with a general consensus across studentson their likeability. In contrast, studentsmay have more “polarized” views of childrenwith problems in adjustment, with some classmatesliking them, others disliking them, andstill others having a neutral view. As a result ofthis polarization, the ratings of these childrenmay be more dependent on which childrenare participating in the sociometric exercise.In actual practice, the problems in the accuracyin the peer ratings of disturbed childrenat lower levels of participation may be exacerbatedif these children tend to have disturbedfriends who may be less likely to volunteer toparticipate in the exercise. The authors concludeby noting that it is impossible to identifya specific “minimum rate” of participationthat should be obtained before the results ofa peer-rating sociometric exercise are uninterpretable.However, they suggest that thesedata clearly indicate that assessors using sociometricratings need to recognize the potentialeffect of participation rates on the results ofthese exercises, especially when assessing childrenwith adjustment problems.AggressionAnother common aspect of a child’s functioningthat is assessed by peer nominationsis aggression. The typical format is to havea class nominate the children in the classwho “Fights most.” As with other nominationtechniques, the format can either be ina fixed-choice or unlimited-choice format.Peer nominations of aggression have beenshown to be correlated with a psychiatricdiagnosis of conduct disorder and thereforecan be considered an important indicator ofthe impairment associated with this syndrome(Walker et al., 1991). However, one of themost interesting and troubling characteristicsof peer-nominated aggression is its stability.In their 5-year study, Coie et al., and Dodge(1983) found that peer nominations of “StartsFights” showed the most stability across the5years of any of the peer-nomination categoriesthat were obtained, exhibiting correlationsof .83 between third and fourth grades and.84 across fifth and sixth grades. Huesmann,Eron, Lefkowitz, & Walder (1984) provideeven more dramatic evidence for the stabilityof peer nominations of aggression.These authors found that peer nominationsof aggression at age 8 significantly predictedaggression 30 years later. Therefore, peernominations of aggression assess an aspect ofproblematic interpersonal functioning thatcan be highly stable for a child and is thus animportant target for intervention.HyperactivityAssessment of inattention and motor hyperactivity,behaviors typically associated withADHD, has relied primarily on information

222 chapter 9 Peer-Referenced Assessmentobtained from parents and teachers (Loeber,Green, Lahey, & Stouthamer-Loeber,1991). However, Schaughency and Rothlind(1991) provide some interesting datato suggest that peer nominations of inattentiveand hyperactive behavior could aidin the assessment of ADHD. Specifically,these authors reported that peer nominationsof “Can’t pay attention,” “Can’t waitturn,” and “Can’t sit still” correlated withteacher and observer measures of inattentionand hyperactivity. In a second study,these authors reported that peer nominationsof “Doesn’t pay attention” and “Can’tsit still” significantly discriminated betweenyoungsters diagnosed with ADHD fromother clinic-referred children.Although these results are promisingand suggest that peer nominations can aidin the assessment of ADHD behaviors,there is no evidence that these peer nominationsshould take the place of parent andteacher ratings as a primary informationsource for these behavioral domains. Also,it is unclear whether or not peer nominationsadd anything to the assessment of ADHDbehaviors over the information providedby other assessment techniques.DepressionLefkowitz and Tesiny (1980) developed thePeer Nomination Inventory of Depression(PNID) to aid in the assessment of childhooddepression. A list of 13 depressionrelatedcategories was developed by nineexpert judges. These nomination categoriesare provided in Box 9.5. Lefkowitz andTesiny found that the 2-month test-retestreliability of the individual depressionitems (mean r = .66) and the depressioncomposite for all 13 items (r = .79) wereacceptable. More importantly, the PNIDDepression composite was significantlyassociated with teacher (r = .41) and child(r = .23) ratings of depression. As was thecase with peer nominations of inattentionBox 9.5Depression Items on the Peer Nomination Inventory of DepressionProceduresChildren were provided a class roster. Eachitem was read aloud twice, and children wereinstructed to draw a line through all the nameson their class roster “which best fit the question”(Lefkowitz & Tesiny, 1980, p. 45).Self-nominations were not permitted, butchildren could choose not to nominate anyonein a category.Depression ItemsWho often plays alone?Who thinks they are bad?Who doesn’t try again when they lose?Who often sleeps in class?Who often looks lonely?Who often says they don’t feel well?Who says they can’t do things?Who often cries?Who does not play?Who does not take part in things?Who does not have much fun?Who thinks others do not like them?Who often looks sad?The 13 items are only the depression itemson the PNID. The PNID includes 20 items,with additional items measuring happinessand popularity.Source: Lefsowitz, M. M., & Testiny, E. P. (1980). Assessment of childhood depression. Journal of Consulting andClinical Psychology, 48, 43–50.

chapter 9 Peer-Referenced Assessment223and hyperactivity, peer assessment ofdepression is promising as a componentto a comprehensive evaluation but has notbeen sufficiently tested to determine itscontribution relative to more traditionalmeasures of depression.Other Peer-ReferencedTechniquesAlthough we have focused most of our discussionon peer nomination techniques, thereare other peer-referenced assessment strategiesthat have been used in research that maybe applicable to some clinical settings. Peerratings require children to rate on a Likerttypescale each member of their class or peergroup (Gresham & Little, 1993; McConnell& Odom, 1986). Like any rating scale, peerscales vary on the behavioral dimensionsincluded on the scale, the number of pointson the scale, and the behavioral descriptionsused as anchors on the scale. For example, arating scale for young children used a happyface to anchor the positive end of the continuum,a neutral face to anchor the middle,and a sad face to anchor the negative end ofthe continuum (Asher, Singleton, Tinsley, &Hymel, 1979).One of the most reliable forms of peerassessment, especially for very young childrenMcConnell & Odom, 1986), is thepaired comparison technique. In this procedure,photographs are taken of all thechildren in the class and photographs ofeach possible classmate dyad are paired.The rater is then asked to choose betweenthe two children in each dyad in referenceto some criterion (e.g., Fights, Liked,Shy, Cooperative, etc.). The salient cuesused in making the choices between peersmake this procedure much more reliablefor young children, especially of preschoolage. However, it is so labor-intensive thatit is not feasible for use in most clinicalassessments. As McConnell and Odom(1986) point out, for a class of 20 children,each child will have to make 171 selectionsfor each criterion.ConclusionsPeer-referenced assessment strategies sharethe common characteristic of having a childor adolescent evaluated on important psychologicaldimensions by his or her peers.Due to several practical and ethical considerations,peer-referenced strategies are notcommonplace in many clinical assessments.This is unfortunate because, if designedappropriately and with precautions taken toensure safe administration, peer-referencedassessment can provide an invaluable pictureof child’s social context. This picture ofa child’s social relationships is essential fortreatment planning given the importanceof social relationships for a child’s currentand future psychological adjustment.One of the most commonly used peerreferencedtechniques is the sociometricexercise. This peer nomination techniqueallows one to determine whether or nota child is accepted, rejected, or ignored(neglected) by his or her peer group. Thesedimensions of social status cannot be adequatelyassessed by other methods of assessment,such as teacher ratings or a child’sself-report. In addition, this crucial aspectof a child’s social context has been highlyrelated to emotional and behavioral disturbances.Therefore, sociometric assessmentis a type of peer-referenced assessment thatcould be an especially important componentof many clinical evaluations.Chapter Summary1. Peer-referenced assessments provideinformation on how a child or adolescentis viewed by his or her peers and,thereby, provide important insights intoa child’s social milieu.

224 chapter 9 Peer-Referenced Assessment2. Peer-referenced assessments mustbe conducted in light of two importantethical issues: the importance ofminimizing the intrusiveness of theassessment and the need to minimizepotential reactive effects of negativeratings by peers.3. Peer nominations are the most commonlyused forms of peer-referenced assessments.They involve having a child’s oradolescent’s peers select one or more childrenwho display certain characteristics.4. Sociometric exercises are a type of peernomination that determines whether achild is accepted, rejected, or neglected byhis or her peers.5. Aggression, hyperactivity, and depressionhave also been assessed through peernomination procedures. Unfortunately,the relative utility of peer-referencedassessments of these psychological dimensions,in comparison to other assessmentmodalities, has not been tested.6. In addition to peer nominations, perceptionsof a child’s peers can be obtainedby having children rate each other alongcertain dimensions on a Likert-typescale.

C h a p t e r 1 0Projective TechniquesChapter Questionsl What are some of the key issues in thedebate over whether and how projectivesshould be used in clinical assessments?l What are some of the strengths andlimitations of the clinical and psychometricapproaches to interpretation ofprojectives?l How would viewing projective techniquesfrom either a traditional projectionapproach or as a behavioral sampleinfluence the type of technique thatwould be used and the interpretationsthat would be made?l What are the basic interpretive strategiesfor inkblot techniques, thematic techniques,sentence-completion techniques,and projective drawing techniques?l What are some specific examples ofadministration, scoring, and interpretivesystems for each type of projective technique?The ControversySurrounding ProjectiveTechniquesNo type of assessment has engenderedas much controversy as projective techniques.For some, projectives are synonymouswith personality testing andprovide some of the richest sources ofclinical information on children andadolescents (Hughes, Gacono, & Owen,2007; Rabin, 1986; Weiner, 1986). Forothers, projective techniques typically225P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_10, © Springer Science+Business Media, LLC 2010

226 chapter 10 projective techniquesdo not meet even the minimum of basicpsychometric standards, and their use,therefore, detracts from the assessmentprocess and tarnishes the image thatpsychological testing has with other professionalsand with the general public(Anastasi, 1988; Gittelman-Klein, 1986;Hunsley & Bailey, 2001). In Box 10.1,we have attempted to summarize some ofthe major arguments made on either sideof this debate.Our philosophy in writing this chapterwas not to espouse either of the strong viewson projective testing. Instead, our goal wasto provide the reader with an overview ofthis method of assessment that would allowfor an informed view of the appropriate roleof projective techniques in clinical assessments.Too often in the past the debate overprojectives has focused on ideological arguments,or even on personal beliefs, withouta critical and scholarly examination of theactual issues involved. Therefore, the firstpart of this chapter focuses on what we feelare the major issues in the use of projectivesthat determine whether they should beused and how they should be used in clinicalassessments.Irrespective of one’s eventual stand onthe projective controversy, projective techniquesremain one of the most commonlyused methods of clinical assessment by psychologistsin general (Watkins, Campbell,Neiberding, & Hallmark, 1995) and by childpsychologists specifically (Hojnoski, Morrison,Brown, & Matthews, 2006). This fact isBox 10.1The Projective DebateProLess structured format allows clinician greaterflexibility in administration and interpretationand places fewer demand characteristics thatwould prompt socially desirable responsesfrom an informant.Allows for the assessment of drives, motivations,desires, and conflicts that can affecta person’s perceptual experiences but areoften unconscious.Provides a deeper understanding of a personthan would be obtained by simply describingbehavioral patterns.Adds to an overall assessment picture.Helps to generate hypotheses regarding a person’sfunctioning.Non-threatening and good for rapportbuilding.Many techniques have a long and rich clinicaltradition.ConThe reliability of many techniques is questionable.As a result, the interpretations are morerelated to characteristics of the clinician thanto characteristics of the person being tested.Even some techniques that have good reliabilityhave questionable validity, especiallyin making diagnoses and predicting overtbehavior.Although we can at times predict things wecannot understand, it is rarely the case thatunderstanding does not enhance prediction(Gittelman-Klein, 1986).Adding an unreliable piece of information toan assessment battery simply decreases theoverall reliability of the battery.Leads one to pursue erroneous avenues in testingor to place undue confidence in a finding.Detracts from the time an assessor couldbetter spend collecting more detailed, objectiveinformation.Assessment techniques are based on an evolvingknowledge base and must continuallyevolve to reflect this knowledge.

chapter 10 projective techniques227not cited to defend their use. It is cited simplyto indicate that projective techniques area firmly entrenched part of the clinical assessmentprocess that shows no signs of changingin the near future.Clinical Techniqueor Psychometric Test?Much of the debate over the use of projectivetechniques comes from a confusion asto the most appropriate criteria with whichto judge the usefulness of projectives. Traditionalmethods of evaluating psychologicaltests are grounded in measurementtheory, which, as was discussed in Chap. 2,relies primarily on indexes of the reliabilityand validity of the scores that result fromthe test (Anastasi, 1988). When evaluatedon these terms, most projective techniqueshave not fared well (Hunsley & Bailey,2001). As Rabin (1986) states:“An aspect of projective tests that is not tobe overlooked is the frequent disappointmentand disaffection with the adequacy,reliability, and validity of several projectivemethods. The psychologist, reared in the atmosphereof respect for science and for thepsychometric purity of his instruments, oftenfinds them wanting” (p. 8).One way in which these criticisms have beenaddressed has been through the developmentof standardized administration, scoring,and interpretive procedures for certainprojective techniques which are designedto provide scores that meet traditional psychometricstandards (Weiner, 2001). Twoexamples of such approaches that are frequentlyused for testing children and adolescentsare the Rorschach ComprehensiveSystem (Exner, 1974) and the RobertsApperception Test for Children (McArthur& Roberts, 1982). Both of these approachesto projective testing are discussed in moredetail later in this chapter. However, it isimportant to note that both systems sharethe goal of providing very clear and explicitguidelines on how the tests are to be given,scored, and interpreted. Such standardizationis a prerequisite to further psychometricevaluation.This method of addressing the criticismsof projective tests has not met with unanimousapproval. Instead, it has been arguedthat projective tests should not be evaluatedby traditional measurement theoryand that any attempt at standardization willlimit the clinical utility of the technique.For example, Haak (1990) has argued that:“The problem with all of these standardizationefforts is the amount of destructionthey wreak on the essential nature of projectives.All such approaches result in a hugeloss of the rich and complex information thatis obtained by using the technique in the firstplace” (p. 149).This argument is based on the contentionthat projectives are part of the olderclinical tradition that seeks to describe theindividual person in depth, capturing all ofhis or her unique dispositions, motivations,conflicts, and desires. This is an idiographicapproach that is not concerned with howthe individual differs from the norm orhow his or her scores compare to thoseof some other reference group (e.g., thosewith diagnoses of depression). Instead, thegoal is simply to understand the person’sunique qualities. In this conceptualization,“validity” takes on a very different meaningthan the one that is typically used inmeasurement theory. One is not concernedwith how a score compares to some objectivecriterion outside the person beingtested.For some psychologists this clinicalapproach might seem unscientific. However,all clinicians rely on intuition at somepoint in an evaluation to understand thenuances of an individual case. Our scienceof human behavior is not at a point

228 chapter 10 projective techniqueswhere every clinical decision can be guidedby well-established principles, and, giventhe complexity of psychological functioning,such pure empiricism may never bepossible. Therefore, the clinical view ofprojectives considers these techniques asa structured way of obtaining these intuitions.By using this structure, one can usethe judgments of other experienced cliniciansas a guide to making interpretations.The importance of understanding thisdebate is not to decide which view is right.What is more crucial is for one to recognizethe two disparate ways of using projectivetechniques and the unique strengthsand weaknesses of both. For example,using projectives as a psychometric techniqueallows one to compare a person’sscore with those from a normative group,or with those from some relevant clinicgroup, or with some other clinically importantcriterion (e.g., response to treatment).However, to use the scores in this way, onemust maintain rigorous standardization inprocedures and be willing to live within theconfines of the data that are available. Afrustrating aspect of clinical assessments isrealizing the limitations of what our assessmentscan provide.On the other hand, using projectives asa clinical tool allows one greater flexibilityin administration and interpretation.However, with this flexibility, the interpretationsthat result from the assessmentare much more susceptible to influencesthat are indiosyncratic to the assessor.Interpretations of the same case materialmay vary widely across clinicians. As such,interpretations should be clearly viewed asclinical impressions and not be evaluated inthe same way as empirically derived interpretations.In Box 10.2 we have provideda more detailed discussion of the importanceof clearly defining one’s approachto projective assessment and then recognizingthe limitations inherent in eithermethod.Projection or BehavioralSample?Even more basic than the debate over themethod of interpretation is confusion overwhat psychological processes projectivetechniques are supposed to measure. Thecritical nature of this question is obviousfrom a psychometric viewpoint. Validity isthe critical property of a test and it is oftendefined as evidence that the test is measuringwhat it is supposed to measure (Anastasi,1988). Therefore, if it is unclear what a test issupposed to measure then it will be unclearas to what are the most appropriate methodsof determining its validity.One dominant view of projective tests,which is the view that led to the name projective,is best described by Murray (1943):“There is the tendency for people to interpretan ambiguous human situation inconformity with their past experiences andpresent wants” (p. 1). This forms the basisof the projective hypothesis. The projectivehypothesis rests on the assumption thatpeople, in the absence of clear environmentaldemands, will project basic aspectsof themselves in their interpretations ofenvironmental stimuli. Freudian theory,which dominated clinical psychology fordecades, heavily emphasized unconsciousconflict as the basic element of humanpersonality. Projection is seen by manyas being a window to these unconsciousdynamics (Rabin, 1986).However, there is a second view of projectives.Rather than seeing them as windowsto hidden or unconscious motivesand drives, many assessors view projectivesas a behavioral sample. For example, Knoff(1983) writes:“A student completes an incomplete sentenceblank with ‘I hate myself’ or ‘My father beatsme up all the time,’ and these hypotheses areconfirmed through self-injurious behavioror a physically abusive father, is this a hiddenaspect of personality? Or how is a student’s

chapter 10 projective techniques229Box 10.2Two Approaches to Projective Testing: You Cannot Have It Both WaysThe divergent approaches to the interpretationof projectives can be descriptively labeledas the psychometric approach and the clinicalapproach. The problem that arises in the use ofprojectives is that many clinical assessors aren’taware of the approach that they are using andtherefore, do not recognize the limitations oftheir approach. To put it bluntly, many assessorswant the best of both worlds. They wantthe flexibility and the rich clinical informationafforded by the clinical approach, but they donot want to recognize the potential biases ininterpretation that are inherent in such usage.In contrast, many psychologists have foundnew promise in projective assessments withthe advent of standardized administrationand scoring procedures for some techniques.However, users of these systems are frustratedby the limited and often confusing data-baseson which to base interpretations and often slipback into making interpretations that are betterconsidered clinical intuitions. In this boxwe provide two examples of the confusionresulting from these differing approaches tointerpretation.In the clinical tradition, interpretations arebased on clinical judgment and experience.This is often considered bad practice, but wefeel that such clinical intuition is unavoidableand even desirable in any assessment enterprise.The problem arises when users fail torecognize the potential unreliability of theirclinical judgments. In fact, justification fortheir interpretation is often based on researchon the Exner Comprehensive System for Rorschachinterpretation, which has demonstratedacceptable levels of reliability for many scores(Hiller et al., 1999). Unfortunately, they usethis argument to justify the reliability of anyinterpretation they make from the Exner systemor to justify the reliability of their interpretationsfrom any projective technique. Thislatter practice would be analogous to assumingthat all self-report measures of anxietyhave the same psychometric properties andtherefore can be interpreted in the same way.A second example comes from a commonpractice in using one of the newer standardizedsystems, like the Exner system for Rorschachinterpretation. Psychologists have enjoyed theincrease in reliability that such systems provideand which sets the stage for more empiricallybased interpretations. However, studies havenot always been able to show empirical supportfor some of the interpretations that have beenwell established in the clinical tradition (Carter& Daccy, 1996; Finch & Belter, 1993; Stredney& Ball, 2005). This has led some to the conclusionthat the richness of the Rorschach recordis simply too complex for current methodology(Finch & Belter, 1993). This implies that thescores cannot be tested adequately with currentresearch methodology. Because of this fact,these authors and others have recommendedthat one should use both the psychometric andclinical method of interpretation of Rorschachwhen using the Exner system.The problem arises when assessors makea clinical interpretation that does not haveempirical support but place undue confidencein this interpretation because they are using a“reliable and valid system.” This goes back toa basic psychometric principle. Tests or interpretivesystems are not themselves reliableand/or valid. The individual interpretationsthat one makes from them can be reliable and/or valid. Unfortunately, the Exner system, likemost of the interpretive systems for the projectivetechniques, encourages interpretationsbased on the clinical tradition, some of whichhave been supported in research and others ofwhich have not garnered much research support.Users then are often unaware of the basisof their interpretations.response to a thematic approach which usesreal photographs depicting significant interpersonalsituations (i.e., peer group acceptance,attitudes toward school-work,reactions to new sibling) different from aninterview question asking how she/he is gettingalong with peers on the playground?”(p. 448).

230 chapter 10 projective techniquesIt is evident from this quotation that onecan view responses to projective tests assamples of behaviors from which onewould like to generalize to behaviors inother situations, outside of the testingenvironment. In fact, this type of interpretationunderlies Rorschach’s original developmentof the inkblot test and has been theguiding principle for Exner’s more recentsystem of interpretation. Rorschach, andlater Exner (see Exner & Martin, 1983;Exner & Weiner, 1994), describe the inkblottests as a “perceptual test,” meaningthat a person’s perception of the inkblot isused as a sample of behavior with whichto generalize to the person’s perception ofother, more clinically relevant, stimuli.These two competing views of what ismeasured by projectives have several importantimplications for the assessment process.As already mentioned, how one viewsthe process will determine what evidenceis used to establish the test’s validity. Forexample, if one views the test as a behavioralsample, then one would want evidencethat the behaviors obtained from the testare associated with behaviors outside thetesting situation. Alternatively, if one viewsprojectives as tapping unconscious conflicts,then the relationship to overt behavior isnot expected to be one-to-one, because thesame conflicts can be manifested in differentbehaviors (Koppitz, 1983). In this case,validity would be best established by showingthat responses on a projective techniqueare associated with other indicators ofunconscious conflicts.Implicit in this discussion is the importantpoint that the way one views the psychologicalprocess that is being measuredby projective tests will determine the typesof interpretations that will be made from achild’s or adolescent’s responses. A personviewing the results in terms of projectionwill make interpretations about drives andmotivations. It is these types of predictionsthat one wishes to make. In contrast, a personviewing the results in terms of a sampleof behavior will make interpretations aboutbehavioral tendencies that are likely to bemanifested in situations outside of the testingsituation. For example, if the Rorschachis used as a sample of perceptions, one wouldwish to make predictions about how theseperceptual tendencies will be manifested inother situations.The final impact of viewing projectivetechniques as either projection or a behavioralsample is its influence on the selectionof the type of stimulus used. Specifically,if one is operating from the projectivehypothesis, one would want as ambiguousa situation as possible. For example,the Thematic Apperception Test (Murray,1943) contains a blank card that has nopicture on it and the person is required tomake up a story about this card. This is anexample of a very ambiguous situation withfew demand characteristics, or very littlestimulus pull, that would guide a person’sresponse. This stimulus allows for the purestform of projection.In contrast, if one wishes to obtain abehavioral sample from the projective technique,high levels of stimulus pull may actuallybe beneficial. If one knows the demandcharacteristics that promoted the response,then one would have some clue as to whatsituations one might generalize (i.e., oneswith similar demand characteristics). Forexample, cards from the Roberts ApperceptionTest for Children (McArthur & Roberts,1982) were designed to pull for specificthemes (e.g., peer conflict, school problems,marital discord in parents). This is not adesirable property from the pure projectionviewpoint because it increases the demandcharacteristics of the stimulus and therebylimits the degree of projection required.SummaryOur approach to the debate over the useof projective techniques is that assessorsshould use or not use projective testing

chapter 10 projective techniques231based on a careful consideration of criticalassessment issues. Assessors are oftenunclear about what approach to measurementthey are using (i.e., clinical or psychometric),and often make inappropriateinterpretations based on this confusion.Further, assessors are often unclear aboutwhat they are trying to measure with projectives,and again, this leads to confusionin interpretations or to the selection of atechnique that is not well-suited for theirpurpose. The rest of this chapter will highlightcharacteristics of specific projectivetechniques. However, these general issuesare paramount in understanding and usingthese techniques; therefore, these issuesare revisited throughout this chapter withreference to specific techniques.Inkblot TechniquesOne of the most commonly used projectivetechniques is the ink-blot technique. Thestimulus is simply an inkblot, and the childis asked to interpret this ambiguous stimulusin some way. The best known of theinkblot techniques, the Rorschach, consistsof ten cards with standardized inkblots.Although the Rorschach is often consideredsynonymous with inkblot techniques,a notable alternative is the Holtzman InkblotTechnique (HIT). This technique wasdeveloped to overcome psychometric limitationsin the Rorschach by constructing acompletely new set of inkblots (Holtzman& Swartz, 2003). The HIT consists of twoparallel forms, each of which contains 45inkblots.Volumes have been written on the differentinterpretive systems for inkblottechniques for children (e.g., Ames et al.,1974; Exner & Weiner, 1994; Holtzman& Swartz, 2003). The primary variationsamong these systems are along the dimensionsdiscussed in the introduction to thischapter: whether the inkblot test is viewedas a projective approach or as a behavioralsample, and whether a clinical or psychometricapproach to interpretation is taken.Rather than giving a superficial overviewof the different approaches to interpretation,we will focus on one of the most commonlyused methods of interpretation forchildren: Exner’s Comprehensive System(ECS) for Rorschach interpretation. Thissystem attempts to integrate five majorapproaches to Rorschach interpretationinto a single Rorschach approach (Exner& Martin, 1983). Even limiting our focusto one system of interpretation does notallow us to do justice to the intricacies ofRorschach interpretation using the ECS;accordingly, the reader is referred to Exnerand Weiner (1994) for a more in-depth discussionof this system of interpretation forassessing children and adolescents.The Exner ComprehensiveSystem for RorschachInterpretationProcess MeasuredThe ECS treats the Rorschach as a perceptual-cognitivetask. When viewed in thisway:“The Rorschach becomes a task to whichpeople respond by exercising their perceptual-cognitiveabilities and preferences. To articulatetheir answers, they must select partsof these variegated stimulus fields to whichthey wish to attend, use some mixture of thefeatures of the stimulus and their own needsto guide formulations, and identify objectsthat will give substance to their impressions.In short, they decide how to scan the stimulus,how to translate the stimulus input, and whatto report” (Exner & Weiner, 1982, p. 3).Weiner (1986) outlines four basic factorsthat influence a child’s response to theinkblot. First, the nature of the stimulusitself may lead the child to classify a blot

232 chapter 10 projective techniquesin a certain way. Although the inkblotswere designed to minimize stimulus pull,there are clearly some common or typicalanswers that are based on the specific featuresof the blots. Second, responses maybe influenced by concerns about making aparticular impression which could lead tosome censoring of responses in a sociallydesirable manner. Third, responses areinfluenced by personality traits that predisposea person to perceive the blots inidiosyncratic ways. Fourth, responses arepartly a function of situational psychologicalstates that affect a person’s perceptualexperience. Each of these factors providesan important context for interpreting achild’s response to the Rorschach inkblots.Administration and ScoringIn contrast to the complexity of Rorschachinterpretation, administration of the task isrelatively simple. The subject is handed theindividual cards with the inkblot stimuli andmerely asked, “What might this be?” Theonly unacceptable response is, “It’s an inkblot.”If the child provides this answer, thenhe or she is encouraged to see it as somethingit’s not. All ten cards are administeredin this way, and the subject’s responses duringthis free association phase are recordedverbatim for later scoring. After the childresponds to all ten cards, the examiner entersthe inquiry phase. The assessor readministerseach card and reads to the child his or herinitial responses. The child is instructed toshow the examiner which part of the blotled to the response and what made himor her think it looked that way. The childis informed that the assessor would like tosee it “just the way you did” and severalstandardized prompts are provided (e.g.,“What in the blot makes it look like that toyou?” Exner & Wiener, 1994). The child’sresponses during this inquiry phase are alsocoded verbatim to use in later scoring.The heart of the ECS is the extensiveand detailed scoring procedure of a child’stest protocol (i.e., verbatim responses toRorschach cards). This system includesapproximately 90 possible scores. Thereare seven major categories of codes, whichare described in Table 10.1: Location,Determinants, Form Quality, OrganizationalActivity, Popularity, Content, andSpecial Scores. The ECS utilizes a StructuralSummary, which shows all of the possiblescores, plus various summary scoresthat are ratios, percentages, and other derivationsof the individual scores that provideimportant information for interpretation.NormingThe best normative data on the Exnersystem come from a large (n = 1,870)nationwide sample of children between theages of 5 and 16 (Exner & Weiner, 1994).At each age in this 12-year age range, therewere at least 105 children. There was alsofairly equal gender representation at eachage, and the inclusion of minority childrenwas at a proportion that approximatednational census data. The only weaknessevident in this normative data base wasthe overrepresentation of children fromhigher socio-economic strata (Exner &Weiner, 1994). From this normative sample,Exner and Weiner (1994) documentedseveral age-related trends in scores. Thesetrends are summarized in Table 10.2.Users of the ECS with children shouldbe aware of these developmental changesin the Rorschach responses and interpretscores within a normative perspective.The extensive normative database forchildren available with the ECS aids insuch inter pretations. This normative baseis one of the major reasons for the popularityof the ECS for use with children.However, it is important to note that theadequacy and consistency of the normativescores across different samples of childrenand adolescents has been questioned (Hunsley& DiGiulio, 2001; Meyer, Erdberg, &Shaffer, 2007).

chapter 10 projective techniques233Table 10.1 Summary of Scores Used in Exner’s Comprehensive System to Rorschach InterpretationCategories Description Examples of ScoresLocation Part of the blot used by respondent W = Whole blotD = Common areaDd = Uncommon AreaS = White spaceDeterminantForm QualityContentPopularOrganizationalActivitySpecial ScoresFeatures of the blot that contributed tothe formation of the responseMeasures the perceptual accuracy of theresponse (i.e., does the area of the blotreally conform to the child’s perception)Places into categories the variouspersons, places, and things that form thechild’s responseCodes the number of times the childgave a high-frequency (very common)response to a blotProvides an estimate of the efficiencyof a child ’ s organization of the stimulusfieldDenotes unusual verbal material in achild ’ s responseF = FormC = ColorT = Texture/ShadingM = Human Movement+ = Superior–overelaborated0 = Ordinary–commonU = Unusual–rare but easy to see- = Minus-distorted, arbitrary, andunrealisticH = Whole HumanAn = AnatomyBl = BloodFi = FireFd = FoodHh = Household itemsP = Number of popular responsesgiven in the entire protocolZ score = Higher scores indicategreater organizationaleffortINCOM = IncongruousCombination-mergesdetails or images inunrealistically wayMOR = Morbid-response includesreferences to death orclear dysphoric feelingAG = Aggressive Movementresponseincludes action thatis clearly aggressiveReliabilityBecause of its explicit and standardizedadministration and scoring procedure, it isnot surprising that the ECS has proven tobe more reliable than many other inkblotinterpretive systems, showing high interraterand high test-retest reliability (Hilleret al., 1999). For example, in a sample of 258-year-old children, one-week test-retestcoefficients for individual and summary

234 chapter 10 projective techniquesTable 10.2 Age Trends in the Normative Data for the Exner Comprehensive System forRorschach InterpretationScoreLength of RecordLocationDevelopmentalQualityMovementDeterminantsChromatic ColorDeterminantsForm DimensionReflection ResponsesPopular ResponsesSpecial ScoresAge TrendYounger children tend to give fewer responses than older children. Protocolsof less than 17 are not uncommon before age 15 and records of more than25 are unusual prior to age 13.Younger children (less than 11) give more responses that include the wholeblot rather than a specific area. More children, and especially very youngchildren, give at least one response that uses an infrequently identified areaof the blot.Younger children frequently give many vague responses in which diffuseimpressions of the blot or blot area are given without clearly articulatingspecific outlines or structural features. Such vague responses account forone-third of the responses of children ages 5, 6, and 7.Younger children give few human movement responses. It is not unusual forthe responses of 5-, 6-, and 7-year-olds to have few or none such determinants,whereas it is uncommon for this to occur after age 11.It is not uncommon for children to give color responses that are notcreated based on the form features of the blot. The presence of such purecolor responses is often interpreted as indicative of poor affective regulation.About 70% of 5-year-olds, 35% of 8-year olds, 23% of 12-year olds, and8% of 16-year olds give at least one pure color response.Answers that include the impression of depth, distance, or dimensionalityincrease with age such that, by age 8, they appear at least once in over halfof all subjects’ records.In contrast to responses of adults, images reported as reflections or mirrorimages, because of the symmetry of the blot are quite common in childprotocols. Such responses appear in about half of the protocols of childrenunder the age of 8. Although the incidence of such responses declines overtime, they are still found in about 25% of the protocols of 15-year olds.There is a steady increase in the number of popular responses with age,with adolescents giving approximately one-third more popular responsesthan children under age 8.Many of the special scores that document unusual verbalizations andcognitive slippage are more common in young children than in adolescentsand adultsscores ranged from r = .49 to r = .95, witha mean coefficient of r = .84 (Exner &Weiner, 1994). In fact, the only coefficientto drop below r = .70 was the coefficientfor Inanimate Movement (r = .49).Validity and InterpretationsWhile Rorschach scores are reliable overshort time intervals, the stability of thescores over longer periods is lower thanthe stability of adult scores. Specifically,Exner, Thomas, and Mason (1985) tested57 children at 2-year intervals from age 8to 16. In general, most scores showed onlymoderate consistency over each two-yearinterval until the interval between the agesof 14 and 16. Some notable exceptions werethe fairly stable coefficients for the use ofgood form, the use of popular responses,

chapter 10 projective techniques235the number of active movement responses,and the use of shading features for makingdepth and dimension responses.The modest stability of Rorschachscores is not unique to this type ofassessment, but is characteristic of mostassessment techniques in children (e.g.,McConaughy, Stanger, & Achenbach,1992). In fact, this is probably a positiveattribute of the scores because it suggeststhat the scores capture the rapid developmentalchanges experienced by childrenand adolescents. However, there is a tendencyto equate Rorschach responses withpersonality assessment, and to equate personalitywith stable dispositions. Thesefindings on the low stability of Rorschachscores clearly argue against making strongdispositional statements based on a child’sRorschach protocol.One common use of the ECS has beento assess childhood depression. Exner(1983) initially developed a DepressionIndex (DEPI) based on six scores froma child’s protocol. Unfortunately, theDEPI, based primarily on research inadults, showed very poor agreement withother measures of depression in children,which led Exner to revise the DEPI(Exner, 1990) in an effort to increase itscorrespondence with other measures ofdepression. Tests of the revised DEPIindex have also failed to find consistentassociations with other measuresof depression (Archer & Krishamurthy,1997; Ball et al., 1991; Carter & Dacy,1996). These findings could be a functionof inadequate methods of assessingchildhood depression in general, whichresults in the failure to have an appropriatestandard with which to judge the Rorschach.Alternatively, Weiner (1986) hasargued that the Rorschach: “is a measureof personality processes, not diagnosticcategories… it can help to identify formsof psychopathology only to the extentthat they identify personality characteristicsassociated with the types of disorder”(p. 155). However, these findings suggestthat users of the DEPI from the ECSshould not expect the scores to be highlyrelated to other indexes of depression.An even more extreme caution is inorder for the Suicide Constellation forChildren, a set of scores based on an indexused to assess for suicidal tendencies inadults (Exner, 1978). The suicide constellationwas developed by selecting the eightbest predictors of suicide from the ECSin a small sample (n = 39) of children whohad attempted or committed suicide withinfewer than 60 days after the Rorschach wastaken (Exner & Weiner, 1994). Unfortunately,the predictive validity of this indexhas not been replicated in other samples,making the interpretation of this indexquestionable at present (Allen & Hollifield,2003).Another common use of the Rorschachis in the detection of cognitive and perceptualirregularities that could be associatedwith schizophrenia (Weiner, 1986). Exnerand Weiner (1994) outline four sets ofRorschach scores that can aid in this detection.First, disordered and illogical thoughtprocesses are the focus of several specialscores in the Exner system. For example,the Incongruous Combination score identifiesresponses that condense blot detailsor images into a single incongruous perceptin which the parts or attributes do notbelong together: “a person with the headof a chicken” (Weiner, 1986, p. 217). Second,perceptual inaccuracies are suggestedwhen the child has a protocol with manyresponses that do not correspond closelyto the form structure of the blot (i.e., poorform quality) or protocols with few commonor popular responses. Third, interpersonalinadequacies that are often associatedwith schizophrenia can be assessed inresponses involving human movement.Movement responses with poor form qualityare considered indicative of inaccurateor unrealistic interpretations of interpersonalsituations (Exner & Weiner, 1994).

236 chapter 10 projective techniquesFourth, the irregular content of a protocol,such as one with very violent (e.g., two boysstabbing each other in the chest) or verybizarre (e.g., flowers squirting poisonousgas) content can be considered suggestiveof disturbed ideation that is often associatedwith schizophrenia.Exner and Weiner (1982) reported dataon 20 children (ages 9–16) reliably diagnosedwith schizophrenia and 23 nonschizophrenicchildren. These data indicated thatthe use of these indexes produced a highcorrect classification rate (90.7%). Thesepositive findings must be interpreted withthree cautions, however. First, indicatorsof perceptual disturbances for the Rorschachhave not always shown high correlationswith other behavioral indicators ofthought disorders (Smith, Baity, Knowles,& Hilsenroth, 2001). Second, while theseindexes appear to be correlated with a diagnosisof schizophrenia, it is unclear howmuch utility these indexes possess over theactual behavioral symptoms of the disorder(Gittelman-Klein, 1986). In other words,there is no evidence to suggest that childrenwho show elevations on these indexes, butwho do not show overt behavioral manifestationsof the disorder, are at risk for developingschizophrenia. Second, Gallucci(1989) studied a sample of 72 intellectuallygifted children and found elevated rates onthese indexes, compared to age norms, butno other signs of maladjustment in the children.Intellectually superior children mayprocess the Rorschach stimuli in nonconventionalways, but these differences shouldnot be considered indicative of a psychoticprocess. Similarly, Holaday (2000) reportedthat children and adolescents diagnosedwith post-traumatic stress disorder alsoreported significantly higher scores on theRorschach indicators of schizophrenia.Thus, indicators of schizophrenia on theRorschach appear to have a high rate offalse-positives for thought disorders (i.e.,many children score high who do not haveother indicators of psychosis).EvaluationThese examples are only a small sampleof the common uses of the ECS in clinicalassessments of children and adolescents.However, these examples illustrate two issuesthat are important for using the Rorschachin the assessment of children and adolescents.In general, Rorschach responses donot typically correspond closely to behaviorallybased diagnoses; therefore, use ofthe Rorschach for diagnostic purposes is notrecommended. Second, many of the Rorschachscores and indexes were developedand validated on adults. Unfortunately, theextension to children has not met with greatsuccess, as is evident from studies on theChildren’s Depression Index and SuicideConstellation of Children. Therefore, usersshould be wary about using adult-orientedsystems for interpreting the Rorschachresponses of children.Thematic (Storytelling)TechniquesThe second type of projective techniquefor children is the storytelling or thematicapproach. In this technique a child is showna moderately ambiguous picture or photographand asked to tell a story about it. Forexample, the instructions to the RobertsApperception Test for Children (McArthur& Roberts, 1982) are:“I have a number of pictures I am going toshow you one at a time. I want you to makeup a story about each picture. Please tell mewhat is happening in the picture, what ledup to this scene, and how the story ends. Tellme what the people are talking about andfeeling. Use your imagination and rememberthat there are no right or wrong answersfor the picture” (p. 7).Thematic tests have been a popular type ofprojective technique with children because

chapter 10 projective techniques237the storytelling format is usually nonthreateningand fun for them. However,it does require a significant level of verbalability in the child.There are many different thematic techniquesthat have been used with childrenand adolescents, which vary in the types ofpictures that are used to promote children’sstories (see Kroon, Goudena, & Rispens,1998 for a review). One of the most commonlyused techniques for children andadolescents is the Thematic ApperceptionTechnique (TAT; Murray, 1943). The TATconsists of 31 cards with black-and-whitepictures of primarily adult figures involvedin some relatively ambiguous action orinteraction. Because the TAT was designedprimarily for use with adults and containedadult pictures, the Children’s ApperceptionTest (CAT; Bellak & Bellak, 1949)was developed especially for children. Theoriginal CAT contains ten cards with picturesdepicting animal figures, although alater CAT-H was developed with humanfigures. The pictures on both the CAT andthe CAT-H were designed to elicit typicalchildhood conflicts that are predicted frompsychodynamic theory (e.g., sibling rivalry,oedipal urges, toileting concerns). Anotherthematic test heavily influenced by psychodynamictheory is the Blacky Picture Test(Blum, 1950). The Blacky Picture Testconsists of 11 cartoons whose central figureis a dog named Blacky. Like the CAT, thepictures were designed to depict psychosexualconflicts common in children.Two apperception tests, the SchoolApperception Method (Solomon & Starr,1968) and the Michigan Picture Test-Revised (Hutt, 1980), were designed morespecifically for use in educational settings.Another apperception test designed specificallyfor use with children is the RobertsApperception Test for Children (RATC;McArthur & Roberts, 1982). The RATC isquite explicit in the themes assessed by thestimulus pictures. Unlike many other thematictechniques, the themes the pictureswere designed to assess are not specific topsychodynamic theory. Also, the RATC isone of the few thematic techniques thatincludes an explicit scoring system. TheRATC is reviewed in greater detail later inthis chapter.Several thematic approaches havespecific sets of pictorial stimuli for specificgroups of children. This is based onresearch showing that children providegreater verbalization on thematic apperceptiontechniques when the stimulusmaterial more closely matches their ethnicity,gender, and age (Constantino &Malgady, 1983). For example, the TATand RATC contain some pictures that aregender-specific and the RATC containssupplementary pictures depicting AfricanAmerican children (McArthur & Roberts,1982). Of particular note, the Tell-Me-A-Story technique (TEMAS: Constantino,Malgady, & Rogler, 1988) is a thematicapperception test that was specificallydesigned to be a culturally sensitive testfor inner-city children and adolescents.The TEMAS involves 23 brightly coloredcards depicting inner-city themes involvingpeer and family interactions. Thereare 11 sex-specific cards and two parallelsets for minority (depicting Hispanicand African-American characters) andnonminority children. Also, unique tothe TEMAS are separate norms for Caucasian,African-American, Puerto Rican,and other Hispanic children across threeage groups (5–7, 8–10, 11–13). However,these norms are based on a rather limitedsample (n = 642) of children from publicschools in New York City (see Flanagan &DiGiuseppe, 1999).General Interpretation ofThematic TechniquesOne important issue in the use of thematictechniques is the lack of standardizedadministration or scoring procedures for

238 chapter 10 projective techniquesmost systems. Of the 12 thematic apperceptiontests used for children and adolescentsreviewed by Kroon et al. (1998), only5 had standardized and objective methodsof scoring children’s responses. For example,assessors administering the TAT oftenselect certain cards to administer. Further,there are many different systems forobtaining scores but most assessors do notuse any systematic scoring system. Giventhe lack of consistency in administrationand scoring, it is not surprising that evidencefor the reliability and validity ofthematic techniques is limited. Thematictechniques are often interpreted within anidiographic or clinical tradition in whichclinical impressions of an individual childare obtained through an analysis of thechild’s stories.Clinical interpretation of a child’s storyis typically based on two broad aspects ofa child’s response. The first step is a processinterpretation. In this part of the interpretation,one notes such characteristics of thestories as how elaborate the stories were,whether the stories were coherent and tiedto the stimulus card, and whether therewere any specific cards for which the childhad difficulty formulating a story. This typeof interpretation can be used to determinehow invested the child or adolescent was inthe assessment process, whether there wereany potential disturbances or idiosyncrasiesin thought processes, and whether therewere any specific types of stimuli that eliciteddefensive reactions from the child.The second part of the interpretiveprocess is a content analysis. Children’sstories are typically analyzed for (1) thecharacteristics of the hero or main character(e.g., motives, needs, emotions, selfimage),(2) forces that affect the hero inhis or her environment (e.g., rejection bypeers, punitiveness from parents, frighteningforces, support by parent, affectionfrom sibling), (3) the coping or problemsolvingstrategies used by the hero (e.g.,aggression, compromise, nurturance), and(4) the outcomes of the story (e.g., positiveor negative, outcomes brought aboutby hero or someone in his or environment,outcomes are realistic). The content analysisshould determine whether there are anyconsistent themes in a child’s story, especiallythemes that transcend the stimuluspull of a card. For example, an aggressivestory provided for a card that shows twochildren fighting is less diagnostic than astory with an aggressive theme based on apicture of two people sitting next to eachother in a park.Roberts ApperceptionTechnique for ChildrenThe Roberts Apperception Techniques forChildren (RATC; McArthur & Roberts,1982) is one example of a thematic techniquethat was explicitly designed for usewith children and is one of the few storytellingprocedures with an explicit and standardizedscoring system. This instrumentillustrates some major components in theinterpretive process of thematic techniques.ContentThe RATC is intended for use with childrenand adolescents of ages 6–15. Thereis a standard set of 27 stimulus cardsdepicting common situations, conflicts,and stresses in children’s lives (McArthur& Roberts, 1982). Eleven cards have parallelmale and female versions, and there is asupplementary set of stimulus cards featuringAfrican-American children. A descriptionof the RATC cards and the themes thecards were designed to elicit are providedin Table 10.3.Administration and ScoringThe administration procedures of theRATC are quite simple (the instructionsgiven to the child were provided earlierin this chapter). The RATC provides

chapter 10 projective techniques239Table 10.3 Depictions in the Stimulus Cards from the Roberts Apperception Test forChildrenCard Number Description Common Themes1 (B & G) Both parents discussing something withchildElicits themes of family confrontationand stories in which parents aregiving advice or punishing a child2 (B & G) Mother hugging child Elicits themes of maternal supportand dependency needs in relation toa material figure3 (B & G) Child working on homework Elicits themes related to child’sattitude to school, teachers, tests,and homework4 One child standing over another childin prone position5 (B & G) Parents are shown in an embrace withchild looking on6 (B & G) Two white children are shown interactingwith a black childElicits themes with aggression,accidents, and illnessesElicits themes related to a child’sattitude toward parental displays ofaffectionElicits themes related to peerinteractions and racial attitudes7 (B & G) Child sitting up in bed awake Elicits themes of anxiety and baddreams8 Both parents speaking to male and femalechild9 Child standing with clenched fists over achild sitting on the ground10 (B & G) Mother holding baby with childlooking onElicits themes related to familydiscussions, such as around disciplineor planning a family activityElicits themes related to peeraggressionElicits themes of sibling rivalry andattitudes toward the birth of a newsibling11 Child cowering with hands in front of face Elicits themes of fear and anxiety12 (B & G) Adult male glaring at a distressed adultfemale with child looking on13 (B & G) Child preparing to throw chair onto theground14 (B & G) Child with paint on hands has put handprintson wall with mother looking on indistress15 Adult female in bathtub with male childlooking through door16 (B & G) Child and father in a discussion whilefather looks at a paperElicits themes of parental conflictand parental depressionElicits themes of anger and aggressivefeelingsElicits themes of maternal limit settingand child wrongdoingElicits attitudes toward sexuality andnudityElicits themes of father-child relationshipsand paternal approvalNote: B & G = Separate cards for girls and boys.Source: McArthur, D. S., & Roberts, G. E. (1982). Roberts Apperception Test for Children. Los Angeles: WesternPsychological Service.

240 chapter 10 projective techniquesexplicit instructions for scoring the storiesprovided by a child. Each story is scored on16 coding categories. There are 8 Adaptivecategories, 5 Clinical categories, and 3 categorieslabeled Indicators. A description ofthese categories is provided in Table 10.4.As evident from this box, the RATC codingcategories are quite similar to traditionalcontent areas used to interpret otherthematic techniques. Scores used in interpretationsfrom the RATC are the totalnumber of times a given code was presentacross all stories. This allows one to determineconsistent themes (high scores withina category) across stories.NormingOne objective of the authors of the RATCwas to develop a standardized scoringsystem so that normative data could begenerated and used by other users of thesystem (McArthur & Roberts, 1982). Theimportance of age-specific normative datain the interpretation of projective testswas already discussed in the previous sectionon the Rorschach. Unfortunately, thenormative data provided in the RATCmanual are minimal. The normative sampleon which norm-referenced scores arebased consisted of 200 school children:20 boys and 20 girls in the age ranges of6–7 and 8–9 and 30 boys and 30 girls inthe age ranges of 10–12 and 13–15. Notonly is the size of the sample small, butits representativeness is also questionable.The sample was taken from three schooldistricts in southern California. Althoughthe manual states that an effort was madeto select children from lower, middle, andupper socioeconomic family backgrounds(McArthur & Roberts, 1982), there is noevidence to show that this goal was met,nor is there any information given onthe ethnic makeup of the sample. Finally,comparisons of scores from this normativesample to other non-referred samplesof children have shown very differentdistributions of scores (Bell & Nagle,1999). Therefore, the norm-referencedscores provided in the RATC manual areof questionable utility.ReliabilityA positive outcome of the explicit scoringsystem was an increase in reliabilitycompared to other thematic approacheswithout standardized administration orscoring procedures. The manual reportsthat 17 doctoral-level raters averaged 89%agreement on three RATC protocols, and8 master’s-level clinicians reached 84%agreement. Evidence for the split-halfreliability of the RATC was less impressive,however. Acceptable reliability(above .70) was found for only 6 of the 13adaptive and clinical scales: Limit Setting,Unresolved, Resolution 2, Resolution 3,Problem Identification, and Support.Validity and InterpretationsThe increase in reliability afforded by theRATC scoring system has set the stage forthe development of a database with which tojudge the instrument’s validity. However, theextent of this database is presently quite limitedand the findings mixed. For example,the manual reported comparisons between200 clinic-referred children and the 200well-adjusted children in the standardizationsample (McArthur & Roberts, 1982). In thisbroad test, all eight of the adaptive scales andall three indicators differed between the twogroups. In contrast, the only clinical scaleto show differences between groups wasthe Rejection scale. Thus, the clinical scalesfailed to differentiate maladjusted from welladjustedchildren, which does not bode wellfor the likelihood of these scales accomplishingthe more difficult task of differentiatingtypes of problems within clinic-referredchildren.

chapter 10 projective techniques241Table 10.4 Profile Scales and Indicators from the Roberts Apperception Test for ChildrenDescription ofScale Purpose Scoring CriteriaReliance onOthers (A)Designed to measure the adaptive capacity to usehelp to overcome problemMain character (1) seeks assistance for handling problem or completingtasks, (2) asks permission, or (3) asks for approval or material objectsSupport-Other (A) Reflects tendency to support others Main character (1) gives object or does something requested or (2)gives emotional support or encouragementSupport-Child (A) Measures self-sufficiency, maturity, assertiveness,and positive affectLimit Setting (A) Measures the extent to which parent placesreasonable and appropriate limits on childProblemIdentification (A)Measures child ’ s ability to articulate problemsituationsResolution-1 (A) Measures child’s tendency to seek easy orunrealistic solutions to problemsResolution-2 (A) Measures a child’s tendency to use constructiveresolutions to problemsResolution-3 (A) Measures a child’s tendency to use constructiveresolutions to problemsAnxiety (C) Measures a child’s tendency to interpret situationsas dangerous and fearfulMain character (1) shows appropriate self-confidence, assertiveness,self-reliance, perseverance, or delay of gratification or (2) experiencespositive emotionsStory describes some appropriate disciplinary action or constructivediscussion following child wrongdoingCharacter in story states a problem or obstacle or experiences contradictoryfeelingsStory involves a situation in which a problem is solved without anyclear mediating process or by some unrealistic or imaginary processStory involves constructive resolution of internal feelings, externalproblem, or conflicted interpersonal relationship; solution is limitedto present situation without an explanation of the process involved inworking through the problemSame as Resolution-2, except the story explains how the characterworked through problemStory involves (1) character showing apprehension, self-doubt, or guiltor (2) themes of illness, death, or accidentsAggression (C) Measures a child’s aggressive impulses Story involves angry feelings, physical or verbal attack, or destructionof objectsDepression (C) Measures a child’s tendency to depression Story involves dysphoric feelings, giving up, or vegetative symptoms ofdepression(Continues)

242 chapter 10 projective techniquesTable 10.4 (Continued)Description ofScale Purpose Scoring CriteriaRejection (C) Measures issues that a child might haveconcerning separation and rejectionUnresolved (C) Measures a tendency toward having an externallocus of control or an inability to control one’semotionsAtypical Response Measures distortions in a child’s thoughtprocesses or unusual ideationMaladaptiveOutcome (I)Measures poor problem-solving abilities or thepresence of a pessimistic or hopeless cognitive styleRefusal (I) Measures a lack of investment in the task orextreme defensiveness to certain stimuliStory involves physical separation, rejection, dislike of another person,needs unmet by others, or racial discriminationStory involves an emotional reaction without a resolutionStory involves a distortion of stimulus card, is an illogical story,involves homicidal or suicidal ideation/action, involves death, orinvolves child abuseStory involves characters that behave inappropriately to solve a problemor when the story ends with the main character dyingChild refuses to respond or stops in the middle of a story and refuses togo onNote: Descriptions of scoring criteria are not full criteria and should not be used in place of the explicit criteria provided in the RATC manual. Manual also provides concreteexamples of each criterion. (A) = Adaptive scale, (C) = Clinical scale, (I) = Indicator.Source: McArthur, D. S., & Roberts, G. E. (1982). Roberts Apperception Test for Children. Los Angeles: Western Psychological Services.

chapter 10 projective techniques243EvaluationBecause of the lack of validity evidence,the RATC should not be used in diagnosticdecision-making, as is the case for otherthematic approaches. Instead, the RATCshould be used as a method of obtainingclinical impressions, with a considerationof all the strengths and weaknesses of thismethod of interpretation. However, unlikemany other thematic techniques, theexplicit scoring system of the RATC has ledto a reliable scoring procedure that sets thestage for further validation to guide interpretations.Another caution in interpretingthe RATC stems from the poor normativebase from which norm-referenced scoresprovided in the RATC are derived (Bell& Nagle, 1999). These scores should heregarded as suspect until further informationbecomes available from larger andmore representative samples of childrenand adolescents.Sentence CompletionTechniquesAnother type of projective technique thatis frequently used in the clinical assessmentof children is the sentence-completiontechnique (SCT). The sentence-completionmethod involves providing the child,either orally or in writing, with a numberof incomplete sentence stems such as, “Myfamily is…” or “I am most ashamed of….”As is evident from these examples, thestimulus employed in sentence-completiontechniques is much less subjective than theother projective methods reviewed in thischapter. That is, the sentence stems have ahigh degree of stimulus pull in promptingcertain types of answers. As a result, manyhave debated whether sentence-completionmethods should even be considered projective,given their more objective nature(Hart, 1986).Although SCTs clearly require a differentlevel of inference than other projectives,they probably are closer to other projectivesin design and interpretation than toself-report rating scales. However, a decisionas to whether or not to use an SCT ina clinical assessment goes back to our initialdiscussion of projective techniques in general.If one wishes to interpret projectivesas a behavioral sample, then the objectivenature of the SCT and the lower level ofinference required is a distinct advantage.In contrast, if one wishes to enhance projection,then the strong stimulus pull of theSCT is less desirable.To illustrate the diversity in how SCTsare used, Holaday, Smith, and Sherry (2000)surveyed a random sample (n = 100) of membersof the Society for Personality Assessmentand they obtained a 60% responserate to their survey. On questions related towhy and how they used SCTs in their clinicalpractice; the only response endorsed bythe majority of respondents (67%) was as“part of a more comprehensive assessmentbattery.” A substantial minority were splitin endorsing “to determine personalitystructure” (i.e., as a projective test) and inendorsing “as a structured interview” (i.e.,as a behavioral sample) to describe theiruse of SCTs, with 30% and 25%, respectively,endorsing these uses. Interestingly,28% endorsed the use “to obtain quotablequotes” as a justification for inclusion ofSCTs in their assessment battery, suggestingthat the information provided by theSCT is often used to obtain examples toillustrate findings from other assessmentprocedures (e.g., clinical diagnoses).ContentFeatures of SCTsDespite a common format, there arenumerous SCTs available that vary in theircontent, length, complexity, and purpose.The Rotter Incomplete Sentence Blank

244 chapter 10 projective techniques(Rotter & Rafferty, 1950) is one of the oldestand most common of the SCTs. It wasoriginally developed for use with adultsand consists of 40 items designed to elicitinformation on psychosexual conflicts. Itis available in three forms, and the authorsprovide a quantitative scoring procedurethat can be used to determine the degree ofconflict present in each response. Anothercommonly used SCT is the Hart SentenceCompletion Test for Children (HSCT).The 40-item HSCT was developed specificallyfor use with children, and the contentwas designed to elicit children’s perceptionsof family, peers, school, and self (Hart,1986). There are numerous other SCTprocedures that are beyond the scope ofthis chapter to discuss in detail (see Haak,2003; Holaday et al., 2000 for reviews).AdministrationAdministration of SCTs is straightforwardand typically includes instructions whichinform children that they are to completethe sentences in whatever mannerthey choose, and that there are no right orwrong answers.There are three important dimensionson which the administration proceduresof SCTs differ (Hart, 1986). First, SCTsand users of SCTs can differ on whetherthe sentence stems are to be read aloudto the child or adolescent or whether thechild being assessed is to be asked to readthe questions and respond privately. Thechoice of which administration format touse is partly a function of the child’s readinglevel and age, with assessors tendingto read sentence stems to children moreoften than to adolescents (Holaday et al.,2000). However, the verbal interchangethat results from the reading of questionsto a child can also provide an assessor withadditional information (e.g., a child’s affectiveresponse to a sentence stem, a child’sapparent motivation toward the task) onwhich to evaluate his or her responses.Second, some users of SCTs request thatthe child answer as quickly as possible bysaying the first thing that comes to his orher mind in an effort to elicit spontaneousand unguarded responses. In contrast,other users attempt to promote deliberationby telling the subjects that they cancomplete the sentences in any way theylike and that the purpose of the test is tobetter understand their real feelings. Thefirst use of the SCT is typically preferredif the goal of administration is projection.The second administration procedure istypically preferred if the goal is to obtain abehavioral sample.Third, SCTs differ in whether or notthey include an inquiry process. In theinquiry phase, children are asked to explaintheir responses in more depth. This questioninghelps the assessor determine why achild may have completed the sentence ina particular way. This information is especiallyuseful for responses that are unusual,ambiguous, or diagnostically important(Hart, 1986). Because of the importantclinical information obtained by thisinquiry, it is often an integral part of theadministration of SCTs for many assessors(Haak, 2003).InterpretationAs with most projective techniques, thereis great variability in how SCTs are scoredand interpreted. Many SCTs do not haveexplicit scoring or interpretive guidelinesand, even for those that do, many users donot follow them in practice. For example,of the 60 respondents from the survey ofusers of SCTs conducted by Holaday et al.(2000), only 17% of those respondentswho said they use SCTs in the assessmentof children stated that they score the testaccording to a manual or according to theauthors’ instruction, and 27% of thoserespondents who said they use SCTs withadolescents reported doing so. In fact,25% of respondents did not even know

chapter 10 projective techniques245the name of the SCT that they used inpractice and 13% reported that they writetheir own sentence stems to address theirclient’s needs.However, interpretation of SCTs typicallyrelies on an analysis of the manifestcontent of a child’s responses. As was thecase with the thematic techniques, anassessor would analyze a child’s responsefor consistent themes that might provideclues to the child’s emotional adjustmentor his or her perceptions of certain personsor situations. For example, positiveresponses to stems designed to assess perceptionsof parents (My father is the best;What I like best about my father is he isnice) are thought to be an indication of apositive father-child relationship. This isan example of the low level of inferencethat is often applied to the interpretationof SCTs.Some assessors also analyze the process ofa child’s responses, such as whether they arecomplex, whether they are perseverative,whether they are expressive and imaginative,and whether they are coherent and related tothe sentence stem (Haak, 2003). This type ofanalysis can provide the assessor with insightas to how invested a child was in the task andsome possible clues about a child’s thoughtprocesses. Box 10.3 provides an overview ofthe most common approaches to interpretingresponses on SCTs.EvaluationAs is evident from the discussion of SCTsto this point, most systems do not haveexplicit and standardized administration,scoring, and interpretive procedures.These decisions are often left to the judgmentof the assessor, who can be guidedby the advice of the authors of the SCT orother experienced clinicians (e.g., Haak,2003). As a result, most SCTs can be consideredto fall in the clinical tradition ofprojectives, with most techniques failingto have well-established psychometricproperties (Anastasi, 1988). Of particularconcern is the lack of a normative base toguide the interpretation of SCTs in childrenand adolescents. Also, most SCTswere initially developed for adults, so thecontent is often inappropriate for children.Drawing TechniquesA final popular approach to projective testingwith children is through the interpretationof children’s drawings. The popularityof drawing techniques in the assessment ofchildren can be attributed to two factors.First, unlike other projective techniquesthat require substantial verbal ability oftenexceeding the capacity of some very youngchildren, drawing techniques are primarilynonverbal. Second, most children are familiarand comfortable with drawing, so it isan enjoyable assessment context for a child.Koppitz (1983) writes:“Drawing is a natural mode of expression forboys and girls. It is a nonverbal language andform of communication; like any other language,it can be analyzed for structure, quality,and content” (p. 426).From this description, it is evident thatthe interpretation of children’s drawings isbased on the same assumptions that underliethe interpretation of other projectiveapproaches; namely, that drawings containnonverbal clues and symbolic messagesabout a child’s self-concept, motivations,concerns, attitudes, and desires (Cummings,1986).Knoff (2003) provides a general frameworkfor administering and interpretingdrawings. Specifically, administrationinvolves two phases. During the performancephase, the child is provided with thenecessary materials to complete the task(e.g., paper; crayons) and asked to draw

246 chapter 10 projective techniquesBox 10.3Research Note: Interpretive Approaches to Sentence Completion TechniquesThe author of the Hart Sentence CompletionTest for Children (Hart, 1986) providedan interesting summary of various interpretiveapproaches to sentence completion techniques(SCTs) in the clinical assessment ofchildren and adolescents.Strategy 1The most common interpretive approachto SCT is to review each item’s content andobtain clinical impressions about a child’s personalitydynamics. The assessor searches forpatterns, clues, and thought processes and generateshypotheses consistent with the assessor’sview of human behavior. This approach oftenleads to different interpretations of the sameset of responses by different clinicians. What isviewed as important or diagnostic will dependon the assessor’s theoretical orientation.Strategy 2The next approach places sentence stems intoclusters with similar item content that arejudged to elicit similar psychological information.The assessor determines if there areimportant patterns of responses within a clusterof items. However, like the first strategy,the interpretations are heavily dependent onthe assessor’s orientation. Which items determinea meaningful cluster and what constitutesan important pattern of responses withina cluster are based on an assessor’s theoreticalorientation.Strategy 3The third approach is exemplified by theRotter Incomplete Sentence Blank (Rotter& Rafferty, 1950) and is based on psychodynamictheory. Each response is analyzedaccording to the degree of intrapsychic conflictevident in the response. It is a quantitativescoring system in which the severity ofthe conflict is rated as negative, neutral, orpositive.Strategy 4The fourth approach involves comparingresponses on an SCT to some predeterminedcriteria. This approach attempts to limit theunreliability inherent in the other strategiesby minimizing the reliance on the assessor’stheoretical orientation. An example of thisapproach is the Hart Sentence CompletionTest for Children (Hart, 1972). In a standardizationsample, a large pool of responseswas obtained for each sentence stem. Theresponses were placed into positive, negative,and neutral categories by expert judges. Ineach rating category, representative responseswere identified for each sentence stem to aidassessors in making their determination of thevalence of a child’s response.Source: Hart, D. H. (1986). “The Sentence Completion Techniques,” in II. M. Knoff (Ed.), The Assessment ofChild and Adolescent Personality, New York: Guilford.specific pictures. During the inquiry phase,a series of questions are asked to clarifythe persons and objects in the picture, tounderstand their actions and motives, andto have the child describe in more detailwhy he or she chose to draw the picturein the way he or she did. Both the drawingitself and the child’s description of it areused to generate hypotheses about potentialthemes that may provide insight into thechild’s emotional (e.g., anxiety) and social(e.g., family relations) functioning.Draw-a-Person TechniqueOne of the most popular drawing techniquesfor children is the Draw-a-PersonTechnique (DAPT), made popular by aseminal publication by Koppitz (1968).In this technique a child is simply given a

chapter 10 projective techniques247paper and lead pencil and asked to draw apicture of a whole person. It is left up tothe child the type of person to be drawn(e.g., age, gender, race, context of figure).After finishing this first drawing, the childis given another sheet of paper and askedto draw another person of the opposite sexfrom the first drawing.Koppitz (1968) provides one of the mostexplicit guides to interpreting children’s figuredrawings. She organizes her approacharound three basic questions. The firstquestion is, How did the child draw the figure?Such content analysis is the focus notonly of the Koppitz system, but of mostinterpretive systems of children’s drawing.In the Koppitz system, the figure is viewedas reflecting a child’s self-concept. Koppitzdeveloped a series of 30 Emotional Indicators(EI) that were rare in children’s drawings,that were independent of age, and thatdifferentiated undisturbed from maladjustedchildren. Examples of EI in the Koppitz systeminclude poor integration of parts, slantingof figure by 15° or more, omission ofmouth, body, or limbs, and monster or grotesquefigures. Figure size is another EI thatis not only a part of the Koppitz system, butis included in many interpretive systems andis considered to be a key indicator of a child’sself-esteem. Small figures are interpreted asindicating low self-esteem (2 in. or less inheight) and large expansive figures (9 in. ormore in height) are interpreted as indicatinghigh levels of self-esteem. Box 10.4 summarizes,in more detail, Koppitz’s EI.Box 10.4Further Discussion of the Koppitz Emotional Indicators for Human FigureDrawingsAs mentioned in the text, the writings of Koppitz(1968, 1983) have been quite influentialin the interpretation of human figure drawingsfor children and adolescents. A key element toher projective interpretation of drawings is thepresence or absence of 30 Emotional Indicators(EI). Koppitz’s El were chosen based on(1) their utility in differentiating disturbedfrom nondisturbed children, (2) their lowprevalence (less than 6%) in the drawings ofnondisturbed children, and (3) their occurringindependent of age.The EIs can be divided into three broad categories: Quality Signs, Special Features, andOmissionsQuality Signs Special Features OmissionsPoor integration of parts Tiny head No eyesShading of face Crossed eyes No noseShading of hands & neck Presence of teeth No mouthAsymmetry of limbs Short arms No bodySlanting figures Long arms No armsTiny figure Arms clinging to body No legsBig figure Big hands No feetTransparencies of major body parts Hands cut off No neckLegs pressed togetherGenitalsMonster/grotesque figuresMultiple figures drawn spontaneouslyClouds(Continues)

248 chapter 10 projective techniquesBox 10.4 (Continued)Koppitz explains that the EI are notscores but are clinical signs that mayreveal underlying attitudes and characteristicsof the child (Koppitz, 1983). Thereis evidence that the EIs occur at a greaterfrequency in the drawings of emotionallydisturbed than nondisturbed children (seeFinch & Belter, 1991). However, Koppitzdescribes the difficulty in interpreting EIsfor the individual child:“There is no relationship between an EIand overt behavior. For instance, long armsand big hands both reflect aggressivenessand anger, yet children who show these twoEI on their drawings may act very differently.One boy may reveal his anger by refusingto do his homework or by truantingfrom school, another child may be physicallyaggressive to peers, while a third childmay withdraw and soil himself when angry.The Human Figure Drawings indicate thatall three pupils are angry; the youngsters’behaviors demonstrate how they expressthis anger. It is also important to recognizethat different EI can reflect the same attitude.Thus, a girl may show acute anxietyby shading the body and face of her HumanFigure Drawing and by omitting thearms. When she makes another drawingsome time later, she may omit the figure’snose and hands and may draw a dark cloudabove the figure. Similarly, a single EI mayhave different meanings depending on thesituation. For example, a tiny figure mayreflect underlying timidity or shyness, orit may indicate withdrawal or depression.The true meaning of a given EI can only bedetermined by other aspects of the personalitybattery, from observing the child indifferent settings, and from studying his orher developmental and social background”(Koppitz 1983, p. 423).The second question around whichthe Koppitz interpretation is organized is,Whom does the child draw? Most childrentend to draw figures that are of the samegender as themselves (Cummings, 1986;Finch & Belter, 1993). Based on these findings,Koppitz considered a child’s drawingan opposite-sex figure on his or her firstdrawing to be diagnostic, either of problemsin gender identity or as a reflectionof loneliness and isolation. There is also atendency to view the figure as an indicatorof the child’s image of his or her own body(Cummings, 1986).The final question in the Koppitzinterpretive system is, What is the childtrying to express via the drawing? A child’sself-figure may reflect his or her self-perceptions,or a drawing of someone elsemay reflect attitudes or conflicts towardthis person. Koppitz notes that a child’sdrawing may (1) be a reflection of a child’swish, fantasy, or ideal; (2) be an expressionof real attitudes or conflicts; or (3)be a mixture of both. To help clarify thisissue, many assessors note either a child’sspontaneous verbalizations about a figureor ask the child to tell a story about thefigure. As noted above, the assessor mayfollow up with an inquiry phase in whichhe or she asks specific questions about thefigure such as, Who is he/she? or Whomwere you thinking about while you weredrawing? or What is he/she thinkingabout? or How does he/she feel?House–Tree–PersonA second projective drawing technique isthe House-Tree-Person (HTP) technique(Cummings, 1986). In this technique thechild is asked to draw a house, a tree, anda person. The order is always the same andthe drawings are done on separate sheetsof paper. After the drawing, children are

chapter 10 projective techniques249asked a series of questions to give them anopportunity to describe and interpret theobjects that were drawn (Cummings, 1986;Koppitz, 1983). According to one of theoriginators of the HTP technique, the threefigures give insight into different facets of achild’s functioning (Hammer, 1958). Thehouse is thought to elicit feelings associatedwith the child’s home situation and familialrelationship. In contrast, the tree is thoughtto elicit deeper and unconscious feelingsabout the child and his or her relationshipswith the environment. Unlike the self-portrait,the tree is thought to have less pull forconscious self-descriptions and therefore toinvolve a greater level of projection. And,finally, the drawing of a person is thoughtto reflect more of a conscious or semiconsciousview of the child’s self, the child’sideal self, or a significant other.Kinetic Family DrawingA third common projective drawing techniquethat is used in the assessment of childrenand adolescents is the Kinetic FamilyDrawing. In this technique a child is askedto “Draw a picture of everyone in yourfamily, including you, doing something”(Burns & Kaufman, 1970, p. 5). Theseinstructions emphasize the family engagingin some activity, hence the term kinetic.As was the case for the HTP technique,there is an inquiry phase in which a childis asked to describe and explain his or herdrawing (Cummings, 1986). The first partof the inquiry typically involves the childexplaining who each figure is (e.g., name,relationship to the child, age). The child isthen asked to describe all the figures, whatthey are doing in the picture, how they arefeeling, and what they are thinking about.After these initial descriptive questions, thechild is asked to tell a story about the drawing,saying what happened immediatelybefore the actions depicted in the drawingtook place and what happens next. Finally,the child is asked to describe anything thathe or she would change about the pictureif he or she could.The popularity of the KFD lies in itsability to assess a child’s perceptions of hisor her family in a fun and nonthreateningway. Burns and Kaufman (1970) outlinea three-part interpretive process that isheavily dependent, not just on the drawing,but on the inquiry phase that follows.The first part of the interpretive processis the analysis of the actions portrayed inthe drawing. They not only refer to themovements between people but the energy(e.g., avoidance, conflict, nurturance) andemotion (e.g., love, anxiety, anger) capturedin the picture. The next part of theinterpretive process deals with the styleof the family drawing. Style refers to thepatterns of interactions among significantfamily members and often reflects a child’sdefense system (e.g., denial, isolation). Thefinal stage of the interpretation is the symbolicinterpretation, which is analogous tothe content interpretations of other projectivedrawing techniques.Psychometric Cautions forDrawing TechniquesAs with most other projective techniques,the best method of validating projectivedrawings has been hotly debated. Ina review of the psychometric propertiesof drawing techniques, Cummings (1986)found that the lack of explicit scoringand interpretive guidelines for projectivedrawings has caused most systems to havepoor reliability. Even for those systems inwhich high reliability estimates have beenobtained, correlations between drawingsand other measures of a child’s adjustmenthave not been consistently shown (Joiner,Schmidt, & Barnett, 1996). Most studieshave found that clinicians are unable todistinguish clinically identified childrenfrom nonclinical controls using projective

250 chapter 10 projective techniquesdrawings. Of great concern is the use ofprojective drawings to detect child sexualabuse. Summaries of this research have notbeen able to find indicators from drawingsthat have consistently and reliably differentiatedabused from non-abused childrenacross multiple samples (Garb, Wood, &Nezworski, 2000).This inability to demonstrate thevalidity of projective drawings has led someauthors to suggest that the use of drawingsin clinical assessments of children is unethical(Martin, 1983). This strong stancehas sparked a lively debate (Knoff, 1983).It is clear that content analyses of drawingshave rarely been shown to predictovert behavior, yet many users still try tomake behavioral predictions (e.g., aggression,anxiety, history of sexual abuse) fromdrawing techniques. A quote from Cummings’s(1986) review of projective drawingresearch seems to summarize a sensibleway to view assessment with projectivedrawings and possibly a good way to viewprojective testing in general:“The greatest value associated with projectivedrawings does not lie in the graphicsymbols represented on the paper. Rather,the value of the technique may be in thepractitioner’s opportunity to observe the examinee’sbehavior while drawing. Drawingsprovide a nonthreatening beginning pointwhich should lead to an in-depth explorationof attitudes, feelings, and beliefs via thesynthesis of direct interviews, third-partyinterviews, observations, and test data” (pp.238–239).ConclusionsIn this chapter we have outlined some of themajor issues in the debate of when and howto use projective testing in the clinical assessmentof children and adolescents. As withmost assessment techniques, the problemwith projective testing lies not in the techniquesthemselves, but in the inappropriatepurposes for which they are often employed.This issue is exacerbated in the use of projectivetechniques because of wide-spreaddisagreement over the basic nature of thesetechniques. There is considerable debateover which psychological processes theyare designed to measure, and there is lackof agreement over what method of interpretation(e.g., clinical or psychometric) ismost appropriate for a given technique. Themost important goal of this chapter was toprovide the reader with a clear discussion ofthese issues so that projectives can be usedappropriately, with the assessor clearly recognizingthe limitations of whichever interpretiveapproach is used.We have also summarized some of themajor methods of projective testing that areused with children. We have discussed inkblottechniques, story-telling techniques,sentence-completion techniques, and projectivedrawings. Space limitations preventan exhaustive review of specific techniquesand interpretive systems. However, we haveattempted to provide selected examples ofeach type of projective method as a basis fordeveloping greater expertise in the use andinterpretation of these techniques throughfurther didactic and clinical training.Chapter Summary1. Projective techniques have been thefocus of much controversy. However,they remain the most frequently usedmethod of psychological assessment.2. Much of the debate over projectivetechniques stems from the confusionover what the projective techniqueswere designed to measure and how tobest evaluate their usefulness.(a) The first area of confusion is whetherprojective techniques obtain samplesof behavior or whether they assessunconscious personality dynamics.

chapter 10 projective techniques251(b) The second focus of debate iswhether projectives are ways ofobtaining highly individualizedclinical impressions or whether theyare psychometric tests that shouldbe evaluated by traditional indexesof reliability and validity.3. The Exner Comprehensive Systemprovides a structured method foradministering, scoring, and interpretingresponses to Rorschach inkblots.(a) The inkblots are administered intwo phases: a free association phaseand an inquiry phase.(b) Detailed scoring of responses provides90 possible scores to be used ininterpretation.(c) Normative samples of children havedocumented several age trends inchildren’s Rorschach responses.(d) Scores from the Exner system haveproven to be reliable.(e) The validity of scores for children hasnot been well established, although ithas been difficult to determine themost appropriate way of testing thevalidity of Rorschach scores.4. Thematic story-telling techniques providea child with a relatively ambiguouspicture and require that the child “makeup a story” about the picture.5. Most interpretive systems of thematictests use a two-part interpretation of achild’s stories. The first step interpretsthe process of a child’s stories (e.g.,coherence of stories) and the second stepinterprets the content of the stories.6. A popular thematic test for use withpreadolescent children is the RATC.(a) The RATC contains pictures depictingcommon situations that childrenexperience and provides a standardscoring system for children’sresponses.(b) The explicit scoring instructionsallow for reliable scoring of children’sresponses.(c) The standardization sample onwhich norm-referenced scores arebased is quite small and its representativenessis questionable.(d) Existing evidence for the validity ofRATC scores is quite limited.7. Sentence-completion techniques providethe child with a sentence stemand require the child to complete thesentence.8. Most sentence-completion techniquesdo not have standardized scoring proceduresfor interpreting children’sresponses. It is left to clinical judgmenthow to interpret the content of theresponses.9. Drawing techniques, such as theDraw-a-Person Technique, the House-Tree-Person, and the Kinetic FamilyDrawings are popular for assessingchildren because drawing is a familiarand enjoyable exercise for children.10. Despite their popularity, scoresderived from children’s drawingshave not been highly associated withother indicators of a child’s emotional,behavioral, or social functioning.

C h a p t e r 1 1Structured Diagnostic InterviewsChapter Questionsl What are the major differences betweenstructured and unstructured interviews?l What are the major similarities anddifferences among the most commonstructured diagnostic interviews used toassess children?l What information can diagnostic interviewsprovide that cannot be obtainedfrom other assessment techniques?l What are some important guidelines forthe appropriate use of diagnostic interviewsin the assessment of children andadolescents?HistoryClinical interviews have a prominent placein the history of psychological assessment.The face-to-face verbal dialog betweenassessor and client is the prototypicalformat for most clinical enterprises. Historically,the most common type of clinicalinterview has been the unstructuredinterview. In the unstructured interview,the interviewer determines what questionsshould be asked, how the questions shouldbe framed, what follow-up questionsshould be asked, and what are acceptableresponses from the client. This unstructuredformat is quite consistent with theclinical approach to assessment, which wasdiscussed in the previous chapter on projectivetechniques. It allows the assessmentto be tailored to the individual needs of theclient and relies heavily on the individualclinician’s orientation and expertise.However, the unreliability inherent inunstructured interviews generates somesignificant problems. The results and interpretationof such interviews tend to behighly idiosyncratic to the clinician conductingthe interview. This unreliability isespecially problematic for research. Hence,many clinical assessors have developed more253P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_11, © Springer Science+Business Media, LLC 2010

254 chapter 11 structured diagnostic interviewsstructured diagnostic interview schedulesthat provide a clear and standardized formatfrom which to conduct the clinical interview.This standardization helps address thekey problems associated with unstructuredinterviews; however, one must recall thatthese interviews are valid only for certainpurposes (e.g., collecting data on specificsymptoms) and relatively less valid for others(e.g., treatment planning for a specificclient; Mash & Hunsley, 2005).Initially, most of these instruments weredesigned for use with adults and were primarilyused in research. Two of the betterknown early interview schedules were theFeighner Research Diagnostic Criteria(Feighner et al., 1972), which later becamethe NIMH Diagnostic Interview Schedule(DIS; Robins et al., 1981), and the Schedulefor Affective Disorders and Schizophrenia(SADS; Endicott & Spitzer, 1978).Over the past two decades, structureddiagnostic interviews have moved frombeing strictly research instruments to beinga part of many clinical assessments. Inaddition, several interview schedules havebeen developed for use with children andadolescents. This chapter focuses on theseinterview schedules and their potential rolein the clinical assessment of children andadolescents.OverviewStructured diagnostic interviews consist ofa set of questions that the assessor asks theinformant (e.g., parent or child). There areexplicit guidelines on how responses are tobe scored. Interview questions generallystart with a stem question (e.g., Have youbeen involved in many physical fights?).If the stem is answered affirmatively, thenfollow-up questions are asked to determineother relevant parameters such asfrequency (e.g., How many fights have youbeen in, in the past year?), severity (e.g.,Have you ever used a weapon in a fight?),onset (e.g., When was the first time yougot into a fight?), and impairment (e.g.,Has fighting caused problems for you atschool, home, or with kids your age?). Anexample of the question format from theNIMH Diagnostic Interview Schedulefor Children-Version 4 (DISC-IV; Shafferet al., 2000) is provided in Box 11.1.Box 11.1Example of the DISC-IV QuestionFormatThe following is an example of the stem/follow-up question format used by mostdiagnostic interview schedules. This questionwas taken from the questions assessing MajorDepressive Disorder from the DISC-IV(Shaffer et al., 2000).“I’m now going to ask some questionsabout feeling sad and unhappy.”1. In the last year, was there a time whenyou often felt sad or depressed for a longtime each day?IF YES,A. Was there a time in the last year whenyou felt sad or depressed for a longtime each day?IF NO, GO TO QUESTION 2B. Would you say that you felt that wayfor most of the day?C. Was there a time when you felt sad ordepressed almost every day?IF NO, GO TO QUESTION 2IF YES,D. In the last year, were there twoweeks in a row when you felt sad ordepressed almost every day?IF NO, GO TO QUESTION 2E. When you were sad or depressed,did you feel better if something goodhappened or was about to happen toyou?F. Now, what about the last four weeks?Have you felt sad or depressed?

chapter 11 structured diagnostic interviews255Commonalities Across InterviewSchedulesIn Table 11.1, we summarize the basiccharacteristics of some of the most commonlyused diagnostic interviews for children.It should be noted that the interviewsare continuously updated to keep up withchanging diagnostic criteria or modifiedfor some specific application (e.g.,outcome measure for treatment of childhooddepression, screening device for anepidemiological study of child disorders).Therefore, many of the interviews havemultiple versions and are continually beingrevised. The contents of these interviewsare all based on the criteria specified in theDiagnostic and Statistical Manual of MentalDisorders, starting with the third edition(DSM-III; American Psychiatric Association,1980) and continuing through itsmost recent revision (DSM-IV-TR; AmericanPsychiatric Association, 2000).Which DSM disorders are assessed isgenerally quite similar across the differentinterview schedules. All of the interviews canbe used to assess for the disruptive behaviordisorders, affective disorders, and anxietydisorders in children and adolescents. Eachof these interviews also allows for at leasta brief screening for schizophrenia. Themajority of structured interviews also allowfor an assessment of substance use, elimination,and eating disorders. Tic disordersare covered exclusively by CAPA, DISC-IV,and K-SADS, whereas the ADIS containsunique screening questions for mental retardation,learning disorder, and somatoformsymptomatology. The CAPA is unique in itsdetailed assessment of sleep disorder symptoms.Most interviews organize questions bydiagnosis. However, the Interview Schedulefor Children and Adolescents (ISCA;Sherrill & Kovacs, 2000) provides a symptom-orientedinterview format with itemsclustered by content (e.g., impaired concentration)and topic area (e.g., mental status)rather than specific diagnostic criteria.In order to promote multi-informantassessments, most interviews contain parallelforms to ask identical questions toboth the child and a parent. There is evenan experimental teacher version of theDISC 2.3, an earlier version of the DISC,that was used in the DSM-IV field trialsfor the disruptive behavior disorders(Frick et al., 1994).A recent trend in the most widely usedstructured interview schedules is the developmentof computer-administered versions.Currently, the DISC-IV, the DICA,the CAPA, and the Dominic-R all haveformats that can be administered by computer.The computer format was designedto enhance the reliability of the interviewsby increasing the ease of administrationand data collection. Similar to the penciland-paperversion, the computer format isdesigned to be administered by an examiner.The examiner reads items from thecomputer screen and enters the patient’sresponses. The computer quickly scoresand stores responses, selects the appropriatefollow-up questions, and skips out ofdiagnostic sections when the respondentfails to meet a certain threshold of severity. Also,the computer-administered interviews haveprograms that quickly score the interviewand provide various summary scores (e.g.,symptom indexes, number of diagnosticthresholds met) that aid in the interpretationof the results.Most of the interviews were designedto assess children and adolescents betweenthe ages of 8 and 17. Some interviewsreport applicability to younger childrenwith parents as informants (e.g., Vallaet al., 2000). However, there is someevidence that the reliability of children’sself-report on diagnostic interviews islow before age 9 (Edelbrock et al., 1985;Hodges & Zeman, 1993). The length oftime that it takes to administer a diagnosticinterview is heavily dependent on thechild being assessed. Because of the stem/follow-up question format, children with

256 chapter 11 structured diagnostic interviewsTable 11.1 Overview of Structured Diagnostic Interviews for ChildrenInterview Time Ages Time Frame Informants CommentsAnxiety Diagnostic Interviewfor DSM-IV (ADIS;Albano & Silverman,1996; Silverman, Saavedra,& Pina, 2001)Child and AdolescentPsychiatric Assessment(CAPA; Angold & Costello,2000)Children’s Interview forPsychiatric Symptoms(ChIPS; Weller et al.,2000)Diagnostic Interview forChildren and Adolescents(DICA; Reich, 2000;Reich et al., 1982)Diagnostic InterviewSchedule for Children(DISC-IV; Shaffer et al.,1993, 2000)60–90min60–120min30–60min60–120min90–120min7–17 Present Child and Parent Semi-structured interview focusing on childhoodanxiety disorders. Categorical and dimensionalsymptom assessment with diagnostic summary thatplaces diagnoses in hierarchical format. Support forreliability and validity for anxiety disorder diagnoses.Limited validity data for other diagnostic categories.9–17 Present Child and Parent Semi-structured interview with flexibility in wordingand in probing for relevant symptoms. Can beused by trained lay interviewers. Includes Likerttypesymptom severity scores. Separate section forimpairment. Good reliability except for modestestimates for ODD and CD. Some validity evidencefor diagnoses relative to external criteria.6–18 Present Child and Parent Highly structured format. Can be used by trained layinterviewers. Brief with broad symptom coverage.Simplified language for use with younger children.Limited research, but some evidence of good sensitivityand specificity.6–18 Lifetime Child and Parent Structured with changes toward more flexible format.Good reliability with some evidence of validity.Evidence supporting use categorically and dimensionally.Used cross-culturally and with a variety ofpopulations.6–17 Present (separatelifetime module)Child and parent (experimentalteacher versionavailable)Highly structured format designed for use by trainedlay interviewers. Good reliability at symptom anddiagnosis level. Covers many diagnostic domains.Frequently used inepidemiological studies. Diagnoses from parentversion related to greater impairment.

chapter 11 structured diagnostic interviews257Dominic-R (Valla,Bergeron, & Smolla,2000)Pictorial Instrument forChildren and Adolescents(PICA-III-R; Ernst,Cookus, & Moravec,2000)Schedule for AffectiveDisorders and Schizophreniafor School-AgedChildren (K-SADS;Ambrosini, 2000; Chamberset al., 1985; Tillmanet al., 2003)15–25min40–60min6–11 Present Child Highly structured pictorial interview with verbalsymptom descriptions. Designed for use by lay interviewerswith minimal training. Promising reliabilityand validity data. Should be interpreted with cautionif used with young children.6–16 Present Child Semi-structured pictorial interview with verbalsymptom descriptors. Designed for use by experiencedclinicians. Categorical and dimensional symptomassessment of 14 diagnostic categories. Shouldbe interpreted with caution if used with youngchildren. Limited psychometric data. Authors notethe need for update to reflect DSM-IV criteria.90 min 6–18 Present and lifetime(in separateand combinedinterview formats)Child and Parent Semi-structured interview originally designed tofocus on depression. Variety of versions that varyin length. Low level of structure necessitates useby clinicians. Includes ratings of symptom severity.Widely used in a variety of populations with goodpsychometric evidence.

258 chapter 11 structured diagnostic interviewsmore symptoms will require more interviewtime because of the need to ask morefollow-up questions. However, as is evidentfrom Box 11.1, the average time toadminister the interviews does not varymuch across the different schedules andlasts typically from 60 to 90 min.Major Sources of VariationAcross Interview SchedulesFrom the previous discussion it is clear thatthe various interview schedules probablyhave more similarities than differences.However, one of the major differencesacross schedules is the degree of structureinherent in the interview format. All of theinterviews provide some degree of structureand give guidelines for standardizedadministration and scoring. However, thereis substantial variation in the amount of“leeway” given the assessor across the variousinterviews. For example, the K-SADSis one of the least structured of the interviews.The original manual for administrationincludes the following instructions:“The K-SADS supplies a series of questionsaddressed to the child for each item to berated. The aim is not to oblige the rater toask all of the questions. They serve as a guidefor questions which have been found mosthelpful and informative. The rater shouldask as many questions as necessary to arriveat a well-documented rating. Needless tosay, probing should be as neutral as possibleand leading questions should be avoided”(Puig-Antich & Chambers, 1978, p. 2).In contrast, the DISC was designed to havea high degree of structure in administration.The manual for its administrationincludes the following instructions:“The DISC symptom questions are designedto be read exactly as written. There is verylimited scope for independent questioning.DO NOT deviate from the prescribedquestion sequence. DO NOT make up yourown questions because you think you have abetter way of getting at the same information,or because you think the question ispoorly worded” (Fisher et al., 1992, p. 31).The trade-off between leeway and structureis obvious. Less structure allows theassessor to tailor the interview accordingto the needs of the individual client. However,these interviews generally require agreater degree of experience to administerand often have lower levels of reliability(Gutterman, O’Brien, & Young, 1987;Hodges & Zeman, 1993).Another major variation among thestructured interviews for children is thetime frame used to assess symptoms anddiagnoses. All of the interviews assesswhether problems are currently evident.This is called a present episode frame of reference.Most interviews consider present episodesto be within the previous six months,although in some instances the time framemay be as short as within the last two weeks(e.g., ISCA for emotional disorders) or aslong as within the last year (e.g., DISC-IVfor Conduct Disorder). Of note, the CAPArestricts the assessment of symptoms tothe previous three months due to concernswith the reliability of memory in childrenand adults over longer time intervals. Similarly,the Dominic-R does not obtain anytemporal information, such as onset andduration of symptoms, given concerns as tothe validity of such information in youngchildren. Nevertheless, the major sourceof variation is whether or not interviewschedules are limited to present episodes.A number of interviews restrict the focusof assessment to the present episode timeframe (e.g., CAPA and ChIPS). However,an increasing number of interviews providefor the assessment of both present and lifetimediagnoses. For example, the DISC-IV providesa more recently incorporated wholelifemodule assessing for whether or not a

chapter 11 structured diagnostic interviews259child has exhibited symptoms of diagnosessince age five but prior to the current year(Shaffer et al., 2000). Similarly, there arelifetime formats for both the ISCA andK-SADS. The DICA is unique in its exclusivefocus on lifetime diagnoses.A third source of variation within theinterview schedules is the answer format. Formost interviews, the interview responses arecoded into a categorical format (yes, no).This categorical format is consistent withthe DSM orientation in which symptomsare considered either present or absent.In contrast, the ADIS, CAPA, ISCA,PICA-III-R, and K-SADS have answerformats that can be placed on a Likert-typescale that allows one to rate the severity ofa symptom. While this format makes itmore difficult to translate responses intoDSM diagnoses, it does not create an artificialdichotomy between the presence orabsence of a symptom and allows symptomscores to reflect gradations in severity.Evaluation of DiagnosticInterviewsAdvantagesStructured interviews share with behaviorrating scales the goal of obtaining adetailed description of a child’s emotionsand behaviors from multiple informants.The logical question is: what advantages dothe time-consuming structured interviewsoffer in comparison to the more time-efficientbehavior rating scales? Some of themore important advantages of structureddiagnostic interviews follow.(1) Structured interviews are useful inobtaining important parameters of achild’s behavior that are not typicallyassessed by most behavior rating scales.Specifically, most interview schedulesprovide questions that elicit informationon the duration of a child’s behavioraldifficulties and the age at which theproblems began to emerge. This temporalinformation allows one to take adevelopmental perspective in understandinga case, a perspective that hasproven to be crucial for assessing manyforms of childhood psychopathology(e.g., Silverthorn & Frick, 1999).(2) Interviews also allow one to determinethe temporal sequencing among behaviors.For example, it is important in theassessment of childhood depression todetermine whether periods of sadnessoccurred contiguously with other behaviorsassociated with depression, such assleep disturbances, eating disturbances,or thoughts of death (Kazdin, 1988).(3) Most structured interviews assess thelevel of impairment associated withbehaviors being reported. Most interviewshave questions that elicit informationon the degree to which a child’sdifficulties are affecting his or herfunctioning in major life arenas (e.g., athome, at school, and with peers).(4) Diagnostic interviews also enhancethe correspondence between assessmenttechniques and diagnostic criteria. As mentionedpreviously, the most commonstructured interviews were specificallydesigned and revised to correspond tothe changing DSM system. Therefore,the usefulness of the interview is, inpart, dependent on the usefulness ofthe diagnostic definitions that are beingassessed.This tie between assessment and diagnosiscan be advantageous for several reasons.First, it promotes revisions of the interviewsto correspond with advances in ourknowledge of the basic characteristics ofchild and adolescent psychological disorders.Second, it allows one to make adiagnosis based on strict adherence todiagnostic criteria. Either due to theoretical,empirical, or practical reasons

260 chapter 11 structured diagnostic interviews(e.g., insurance reimbursement), manyclinicians attempt to make DSM diagnosesas a result of their assessments.Too often, diagnoses are made based oninformation (e.g., rating scales, projectivetests) that do not directly assess the diagnosticcriteria or through techniques thatare unsystematic in their assessment ofsymptomatology and associated features(e.g., unstructured interviews; Klein et al.,2005). As a result, the meaning of thediagnosis is ambiguous. Structured diagnosticinterviews do not have this problem.They tend to result in more reliablediagnoses (Silverman & Ollendick, 2005)and can actually provide important informationthat may alter diagnostic impressionsand treatment recommendations(Kashner et al., 2003).(5) Diagnostic interviews are also helpfulin training clinical assessors. As assessorsare developing their competence ininterviewing, it is often helpful to havean explicit format from which to conductthe interview. It gives the assessora good way to learn the basic characteristicsof childhood emotional andbehavioral disorders. After being trainedin administration procedures and afterconducting several interviews withactual clients, assessors often begin tointernalize the diagnostic criteria for themost common disorders of childhood.This knowledge can then be applied insituations in which a structured interviewis not possible.DisadvantagesDiagnostic interviews also have a numberof weaknesses of which the clinician shouldbe aware.(1) The time consuming nature of theinterviews coupled with some questionas to whether they provide incrementalvalidity in the assessment of somedisorders such as ADHD (Pelhamet al., 2005) suggest that structuredinterviews may not be practical oruseful in many situations.(2) In addition, structured interviewsdepend on DSM criteria, which is astrength for assessing well-validatedsyndromes but a weakness for assessingdisorders with a weak empirical basis.(3) Diagnostic interviews are subject tothe same potential reporter biases thatwere discussed in previous chapters onself-report inventories and parent andteacher behavior rating scales.(4) Making differential diagnoses with theassistance of structured interviews stilldoes not directly translate to plans forintervention (Mash & Hunsley, 2005).(5) There is great difficulty in makingnorm-referenced interpretations frominterviews. Clinically significant levels ofsymptoms are often based on DSM criteriarather than based on a comparisonwith a representative normative sample.Therefore, the appropriateness of clinicalelevations for a given age dependson the appropriateness of the diagnosticcriteria for that age. For example, Barkley(1997) has questioned the validityof applying the same diagnostic criteria(i.e., the same number of symptoms) forADHD across the early-childhood andadolescent years based on a documenteddecline in hyperactivity and impulsivitysymptoms as children move into adolescence.As a result, the uniform diagnosticthreshold, while appropriate forelementary school-aged children, maybe too conservative for adolescents.Several community studies using anearly version of DISC (e.g., Anderson,et al., 1987; Costello, 1989) allow one toview the base rates of disorders assessedby the DISC in community samplesof school-age children. Such work hascontinued with the current versionwith various populations (e.g., Roberts,

chapter 11 structured diagnostic interviews261Roberts, & Xing, 2007). Although thesestudies provide some information onhow a child who meets criteria wouldcompare to others in the general population,this type of normative informationis still much more limited than thetype provided by other assessment techniques,most notably behavior ratingscales.(6) Another weakness of most diagnosticinterviews is the failure to provide aformat for obtaining information from achild’s teacher. This source of informationis crucial in the clinical assessmentof elementary school-aged children(Loeber et al., 1991). As a result, informationfrom teachers must be obtainedby some other method, thereby makingit difficult to determine if discrepanciesbetween a teacher’s report and thereport of others are due to real differencesin a child’s classroom behavior orwhether they are due to differences inthe assessment format.Recommendations for Useof Structured InterviewsBased on the strengths and weaknesses ofstructured interviews, several recommendationscan be made on their appropriateuse in the clinical assessment of childrenand adolescents. First, like any assessmenttechnique, the diagnostic interviewshould never be used alone in aclinical evaluation. It should be one partof a comprehensive assessment battery.For example, information from diagnosticinterviews should be supplemented byassessment techniques that provide betternorm-referenced scores (e.g., behavior ratingscales) and by assessment techniquesthat provide information on a child’s classroomfunctioning (e.g., behavioral observationsin the school).In addition, the diagnoses derived fromthe diagnostic interviews should be viewedwithin the context of the overall assessment.A diagnosis can be viewed similarly to theway an elevation on a behavior rating scaleis interpreted. Specifically, it is one piece ofinformation that needs to be integrated withother sources of information to develop agood case formulation. Stated simply, diagnosesbased solely on diagnostic interviewsshould not be considered final clinical diagnoses.Such final diagnoses should be basedon an assessor’s integration of multiplesources of. A case example in which a diagnosticinterview was used as part of a comprehensiveassessment battery is provided inBox 11.2.The child’s age is also an importantconsideration in the use of structuredinterviews. Generally, the reliability formost interview schedules is low beforethe age of 9 for child self-report (Hodges& Zeman, 1993). It seems that the structured,face-to-face dialogue is not appropriatefor assessing very young children.Several interviews have been developedusing pictorial stimuli, rather than relyingpurely on question-and-answer format, inan effort to increase the reliability of theinterview for younger children.A description of two such pictorial interviewsis provided in Box 11.3.Another important issue in interpretinginformation obtained from standardizedinterview schedules is the consistent findingthat, if the interview is repeated, parentsand children uniformly report fewer symptomson the second administration (Jensen,Watanabe, & Richters, 1999; Piacentini,et al., 1999). For example, Piacentini, et al.(1999) found that when the DISC-IV wasreadministered to the same sample of 245parent–child pairs (age 9–18) 12 days later,parent-reported symptoms dropped 42%,and child-reported symptoms dropped 58%,The reasons for this symptom attenuationhave not been conclusively shown. However,they can include (1) sensitization to clinical

262 chapter 11 structured diagnostic interviewsBox 11.2Case Example: The DISC-IV in the Evaluation of a 9-Year-Old Girl withAttention-Deficit Hyperactivity DisorderAlexis who is 9 years and 3 months old wasreferred for a comprehensive psychologicalevaluation by her parents upon the recommendationof her teachers. Her teachers hadreported to Alexis’ parents that she was havingdifficulty paying attention and was daydreaming,interrupting others, and making carelessmistakes in her work. Her parents requesteda comprehensive evaluation to determine theseverity and possible cause of these difficultiesand to get recommendations for possibleinterventions to aid in her school adjustment.Alexis’s background, developmental, andmedical history were unremarkable. Duringthe testing, Alexis had great difficulty concentratingand was easily distracted. She was alsovery fidgety and restless. Intellectually, Alexishad much better verbal comprehension abilities,especially in the area of verbal reasoning,than nonverbal perceptual-organizational abilities.Consistent with her verbal abilities, Alexisscored in the above average range on measuresof reading and math achievement.Alexis’s emotional and behavioral functioningwere assessed through the use of structuredinterviews con-ducted with Alexis, her parents,and her teachers as well as through ratingscales completed by her parents and teacher.The structured interviews were the parentversion of the DISC-IV and the experimentalteacher version used in the DSM-IV fieldtrials (Frick et al., 1994). The child version wasgiven to Alexis.The following is an excerpt from the reportof Alexis’s evaluation that illustrates how informationfrom the DISC-IV was integrated withother assessment information.The only problematic areas that emergedfrom this assessment of Alexis’ emotional andbehavioral functioning were significant problemsof inattention, disorganization, impulsivity, andoveractivity that seem to be causing her significantproblems in the classroom. On the DISC-IV, Alexis’s parents agreed that such difficultieswere noticeable by the time she was age 5. Alexisherself endorsed many symptoms of inattention,but not overactivity, on the DISC-IV. Alexis’steachers described her as being very restless andfidgety, being easily distractible, having very disorganizedand messy work habits, having a hardtime completing tasks, and making many carelessmistakes. Results from teacher rating scalesfurther suggested that these behaviors are moresevere than would be typical for children Alexis’sage. These behaviors are consistent with a diagnosisof Attention-Deficit Hyperactivity Disorder(ADHD). Also consistent with this diagnosis werethe age of onset of these symptoms as reported byAlexis’s parents. These behaviors associated withADHD seem to be causing significant problemsfor Alexis in school, affecting the amount andaccuracy of her schoolwork.issues leading to a heightened threshold forsymptom reporting, (2) a circumscribedfocus on only the interval between assessmentperiods, (3) statistical regression tothe mean, and (4) knowledge that brevityof responding will shorten the duration ofthe interview (Piacentini et al., 1999). Thissymptom attenuation is not much of an issuein most clinical uses of structured interviewsthat do not involve multiple administrations.However, there is also evidence thatthe number of symptoms reported declineswithin an interview schedule, such thatparents and child tend to report more symptomsfor diagnoses assessed early in theinterview, even if the order of assessment isvaried (Jensen et al., 1999).This type of symptom attenuation withinan interview is of much greater concernbecause it clearly can influence the resultsfrom a typical clinical use of structuredinterviews. In an attempt to solve this problem,Edelbrock, Crnic, and Bohnert (1999)modified the administration of DISC-2.3

chapter 11 structured diagnostic interviews263Box 11.3Research Note: Developing Structured Interviews for Young ChildrenTo increase the usefulness of structured interviewsfor assessing younger children, severalauthors have attempted to develop interviewsthat involve pictorial content to eitherreplace or augment the typical question-andanswerformat for structured interviews. Oneexample of this approach is the Dominic-R byValla et al. (2000). The Dominic-R is designedto assess DSM-III-R criteria for anxiety disorders,mood disorders, and the disruptivebehavior disorders in children ages 6–11.The interview involves pictures of a childnamed Dominic facing situations that arecommon in children’s daily lives. The picturesare accompanied by questions about the visualimage (e.g., “Do you feel sad and depressedmost of the time, like Dominic?”). There is aversion of the Dominic-R, the “Terry Questionnaire,”that employs an African Americancharacter named “Terry,” and there are translationsof the Dominic-R in French, Spanish,and German. There is also an InteractiveDominic Questionnaire that is a CD-ROMbasedinteractive cartoon. The Dominic-Rtakes about 15 to 25 minutes to completeand is highly structured, which allows it to beadministered by lay interviewers. The reliabilityof the Dominic-R was assessed in a sampleof 340 community children aged 6 to 11, andit revealed reliability coefficients that weremuch improved over other structured interviewswith very young children. An adolescentversion has also been developed (Smolla et al.,2004).For the child version, test-retest reliabilityover 7 to 12 days for diagnoses from the Dominic-Rranged from a kappa of .44 to a kappaof .69, with most being above .60 (Valla et al.,1997). Also, the diagnoses based on the Dominic-Rwere strongly associated with diagnosesmade by experienced clinicians with kappa valuesranging from .64 to .88 (Valla et al., 2000).Furthermore, research has shown the Dominic-Rto accurately designated children meetingDSM-IV criteria for Conduct Disorder(Arseneault et al., 2005). Therefore, it appearsthat the combination of pictorial stimuli andverbal stimuli can enhance the reliability andvalidity of responses in young children.A similar assessment system, described byErnst, Cookus, and Moravec (2000), is calledthe Pictorial Instrument for Children andAdolescents (PICA-III-R). The PICA-III-R isa semi-structured interview that includes 137pictures assessing anxiety disorders, mood disorders,disruptive behavior disorders, psychoticdisorders, and substance abuse. Therefore,the content of the PICA-III-R is somewhatbroader than that of the Dominic-R. Like theDominic-R, the PICA-III-R combines pictorialstimuli with verbal questions. However,the verbal questions on the PICA-III-R aremore extensive, often including many followupquestions (e.g., “Do you ever get like him(e.g., sad)?” “How much?” “Do people tellyou that you look sad?”). Also, the PICA-III-R verbal questions are not meant to be readverbatim. The type and degree of questioningis left somewhat up to the interviewer. Becauseof all of these characteristics, the PICA-III-Rmay be more useful for older childrenwith more severe forms of psychopathologycompared to the Dominic-R, and itmust be administered by a experienced clinician.Ernst et al. (2000) reported that, ina sample of 51 inpatient children and adolescents(aged 6 to 15), the PICA-III-Rscales were generally internally consistent(i.e., most over .80), although the internalconsistency of the Mania (.69) and Obsessive-Compulsive symptoms (.54) were somewhatlow. There is minimal information on thevalidity of the PICA-III-R other than the factthat scores from this interview changed overthe course of hospitalization for the inpatients,which presumably reflected improvementbrought about by treatment.Both the Dominic-R and PICA-III-Rprovide examples of ways to enhance theusefulness of structured interviews in obtainingchild self-report. It is important to note,however, that both of these interviews arein the very early stages of development, andmuch more information is needed on their(Continues)

264 chapter 11 structured diagnostic interviewsBox 11.3 (Continued)reliability across samples and their validity forassessing DSM diagnoses.Finally, another example of a uniqueapproach to the assessment of young childrenis the Berkeley Puppet Interview (see Measelle,Ablow, Cowan, & Cowan, 1998). Thisinterview calls for two puppets to make statementsabout themselves and then ask the askthe child a question with two choices basedon the same attribute (e.g., “Are you good atmaking friends or are you not good at makingfriends?”). This interview includes academic,social, and symptom-related (i.e., aggressionhostilityand depression-anxiety) domains.Some evidence of convergent validity has beenfound based on ratings by parents and teachers(Measelle et al., 1998) and the interviewhas been able to differentiate clinic-referredfrom community samples of children (Ablowet al., 1999). However, much more researchon the validity, feasibility, and utility of thisassessment tool is needed.to produce more stable symptom endorsementsthat are less susceptible to attenuationeffects. They included a more detailedintroduction with an overview of all areasto be covered on the interview, and theyprovided definitions of key concepts usedthroughout the interview. Furthermore,they employed a flexible order of administrationthat allowed parents to select theorder in which the various diagnoses wereassessed. Using this methodology in a sampleof 24 parent-child pairs with childrenranging in age from 6 to 15, there was veryminimal reduction in the number of symptomsreported when the interviews wererepeated one week later. However, noneof the standardized administration proceduresthat accompany structured interviewschedules provide for this type of administration,and therefore, the possible dropin number of symptoms reported for disordersassessed later in the interviews mustbe considered when interpreting the resultsfor individual children and adolescents.A final consideration in using structuredinterviews concerns when to administerdiagnostic interviews in the assessment battery.There is no research on this issue, andthese recommendations come from clinicalexperience. Diagnostic interviews shouldnot be the first assessment administeredto parents. The structured format doesnot facilitate the development of rapportbetween the interviewer and parent, andsome parents become frustrated in tryingto fit their main concerns and descriptionsof their child’s behavior into the confines ofthe interview. Therefore, it is often helpfulto precede diagnostic interviews with lessstructured questions that allow parents toexpress, in their own words, their concernsfor their child. However, for children andadolescents, we actually find that the structuredformat enhances rapport in manycases. Children often enter the assessmentsituation nervous because they are unsureabout what is expected of them. The clearand explicit response format of diagnosticinterviews makes the demands of the situationapparent for the child and therebyreduces his or her anxiety in many cases.Up to this point, we have tried to give anoverview of structured interviews, lookingat the various formats that are available,highlighting some of the advantages anddisadvantages of using interview schedulesin a clinical assessment, and providingguidelines for appropriate use. In the nextsection, we provide a more in-depth lookat one particular interview schedule, theDISC-IV. We chose the DISC-IV as a prototypicalexample of a structured interview

chapter 11 structured diagnostic interviews265because it is one of the most widely usedinterview schedules for children and adolescents,and it has been one of the mostsystematically developed. However, onemust be aware that the DISC-IV is one ofthe most structured interview schedules,and therefore, it has all of the advantagesand disadvantages that accompany a highdegree of structure.Focus on the NIMHDiagnostic InterviewSchedule forChildren (DISC-iv)DevelopmentThe original version of the DISC (DISC-1; Costello, 1983; Costello et al., 1984)was designed to be a downward extensionof the adult-oriented Diagnostic InterviewSchedule (Robins et al., 1981). TheDISC-1 was developed as part of an initiativeby the NIMH Division of Biometryand Epidemiology that focused onobtaining a greater understanding of theprevalence of childhood mental disorders(Shaffer et al., 2000). The DISC-1 wasdesigned for use in epidemiological studiesand was explicitly tied to the version ofthe DSM being used at the time (DSM-III;American Psychiatric Association, 1980).In 1985, Dr. David Shaffer at the NewYork State Psychiatric Institute and his colleaguesundertook a revision of the interviewto (1) improve its reliability for usewith children and for use by lay interviewersand (2) provide diagnostic compatibilitywith the DSM-III-R (American PsychiatricAssociation, 1987) and anticipated DSM-IV and ICD-10 criteria (Fisher et al., 1992).Modifications to the DISC have beengreatly informed by field-testing conductedas part of a large NIMH-funded multisitestudy titled the Methods for the Epidemiologyof Child and Adolescent MentalDisorders (MECA) study.The current version of the DISC, theDISC-IV (Shaffer et al., 2000), assessesapproximately 30 diagnoses of childhoodand adolescence and is fully compatiblewith the DSM-IV (American PsychiatricAssociation, 2000) and ICD-10 (WorldHealth Organization, 1993) classificationsystems. The development of the DISChas expanded to other languages and toalternative administration formats (i.e.,computerized).Structure and ContentThe DISC-IV (Shaffer et al., 2000) contains358 core questions and approximately1,300 questions that are asked contingenton a child’s responses to the core questions.There are two parallel versions ofthe DISC-IV, the youth version (DISC-Y),to be administered to children of ages of9–17, and the parent version (DISC-P), tobe administered to the parents of childrenof ages 6–17. There is also an experimentalteacher version (DISC-T), which wasdeveloped for use in the DSM-IV field trials(Frick et al., 1994). The DISC-IV wasdesigned with a primary focus on currentpsychological functioning. It assesses forsymptoms occurring within two overlappingtime intervals: the past twelve monthsand the past four weeks. The DISC-IV containsan optional whole life module designedto measure symptoms occurring as early asage 5. An alternative present-state versionhas been developed targeting only the fourweektime interval (see Shaffer et al., 2000).The DISC-IV is organized in “diagnosticmodules.” There are six modulesthat comprise sets of related disorders. Asummary of the modules is provided inBox 11.4. For each diagnosis, DISC-IV isdesigned to obtain information about thepresence of symptoms included in DSM

266 chapter 11 structured diagnostic interviewsBox 11.4Organization and Content of theDISC-1VModuleAnxietyMoodPsychosisDisruptiveBehaviorSubstanceUseMiscellaneousSource: Shaffer et al., 2000.Disorders CoveredSocial Phobia SeparationAnxiety Specific PhobiaPanicAgoraphobiaGeneralized AnxietySelective MutismObsessive-CompulsivePosttraumatic StressMajor Depressive/DysthymicMania/HypomaniaSchizophreniaAttention-Deficit/HyperactivityDisorderOppositional Defiant DisorderConduct DisorderAlcohol Abuse/DependenceNicotine DependenceMarijuana Abuse/DependenceOther SubstanceAbuse/DependenceAnorexia Nervosa/BulimiaNervosaElimination DisordersTic DisordersPicacriteria. If a certain threshold, usuallybelow DSM diagnostic threshold, is met,the questions regarding the age of firstonset, impairment, and past treatment areasked. (See Box 11.1 for an example of theDISC-IV question format.)AdministrationThe DISC-IV was designed to be administeredby interviewers without clinicalexperience after approximately two to sixdays of training. Use of the computerizedDISC-IV is accompanied by less stringenttraining requirements. Training includes (1)instruction on standard DISC-IV administrationprocedures, (2) viewing an actualadministration of the DISC-IV, and (3)supervised practice in administration witha confederate in a controlled situation.At the beginning of the interview, theinterviewer completes an introductorymodule consisting of several pieces ofdemographic information (e.g., age andsex of child) that are necessary to properlyadminister the interview. The intervieweralso establishes a time line with the intervieweeto assist in his or her recall for theonset and duration criteria contained in theinterview. The time line establishes salientevents (e.g., birthdays, vacations, start of theschool year, holidays) that occurred in theyear preceding the interview. These anchorshelp the child or parent remember the timeframe for diagnostic questions.The verbal instructions given to therespondent are semi-structured. That is,several points that must be covered areprovided, but verbatim instructions arenot required. The points include:1. There are no right or wrong answers.The best answer is the one that tells themost about the child.2. The informant should try to answer“yes” or “no” to each question.3. The time frame is within the last year,unless otherwise specified.4. Some of the questions on the form willbe left out.5. Some questions maybe asked morethan once.6. It is possible to take breaks, if needed.Unlike the instructions, the administrationof the actual DISC-IV questionsis quite structured. The questions aredesigned to be read exactly as written andin the sequence prescribed. Interviewersare explicitly instructed not to makeup their own questions or to ask for anexample unless it is requested in the inter-

chapter 11 structured diagnostic interviews267view format. If a respondent does notunderstand the question, the interviewershould repeat the question, emphasizingthe words that seemed to cause confusion.The interviewer is not allowed to answerinterpretive questions for the respondent(e.g., “What do you mean by often?” or “Isone or two times considered frequent?”).The interviewer is instructed to simply askthe respondent to interpret the question“whichever way s/he thinks is best.”ReliabilityThe DISC-IV (and other interview schedules)relies heavily on psychometric dataderived from prior versions of the instrument.Initial reliability data on the DISC-2.3was obtained from a series of articles by Shafferand colleagues (Piacentini et al., 1993;Schwab-Stone et al., 1993; Shaffer et al.,1993). These authors tested the psychometricproperties of the DISC-2.3 in a sampleof 75 clinic-referred children ages 11–17. In41 cases, the child and/or parent were reinterviewedone to three weeks later by a secondinterviewer. There were sufficient casesto calculate the test-retest reliability for fiveDSM-III-R diagnoses (i.e., Attention-DeficitHyperactivity Disorder, Oppositional DefiantDisorder, Conduct Disorder, MajorDepression, and Separation Anxiety Disorder).The kappa statistics for the test-retestagreement are reported in Table 11.2. Alsoreported in this table are the intraclass correlations,which provide an estimate of thetest-retest reliability of the symptom clustersthat form the criteria for DSM-III-R diagnoses.Importantly, the tendency for parentsand children to report substantially fewersymptoms on repeated administrations ofstructured interview can substantially reducetest-retest coefficients. Therefore, Table 11.2also includes another index of reliability, theCronbach’s alpha, as an estimate of the internalconsistency of the symptom clusters.On a diagnostic level, all diagnosesshowed relatively high test-retest agreementTable 11.2 Test–Retest and Internal ConsistencyEstimates from the DISC-2.3Kappa ICC AlphaParent only (n = 39)Attention-Deficit .55 .87 .87HyperactivityOppositional.88 .82 .75DefiantConduct .87 .86 .56Major Depression .72 .82 .88Separation Anxiety .77 .61Child Only (n = 41)Attention-Deficit.72 .83HyperactivityOppositional.16 .44 .67DefiantConduct .55 .60 .59Major Depression .77 .68 .85Separation Anxiety .72 .66 .71Combined Parentand Child (n = 37)Attention-Deficit .56 .84HyperactivityOppositional.59 .65DefiantConduct .50 .80Major Depression .71 .66Separation Anxiety .80 .72Note: Kappa is the agreement between diagnosesat Time 1 and diagnoses at Time 2 with a 1- to3-week interval between interviews. ICC is the intraclasscorrelation between symptoms at Time 1 andTime 2. Alpha is Chronbach’s alpha calculated forthe symptom at Time 1.Source: M. Schwab-Stone,P. Fisher, J. Piacentini, D. Shaffer, M. Davies, &M. Briggs (1993). “The Diagnostic InterviewSchedule for Children-Revised Version (DISC-R): II.Test–Retest Reliability,” Journal of the American Academyof Child and Adolescent Psychiatry, 32, 651–657.for the parent interview, the child interview,and the combined parent and child interview.When combining a parent and child report,a symptom was considered present if eitherparent or child endorsed it. The one exceptionto the generally good reliability was thelow reliability of the Oppositional DefiantDisorder diagnosis by child report. Another

268 chapter 11 structured diagnostic interviewsexception was the low internal consistencyof the Conduct Disorder symptoms. However,the low internal consistency of thesesymptoms is not surprising for two reasons.First, only three symptoms are required forthe diagnosis of Conduct Disorder and mostof these symptoms tend to have a relativelylow base rate. Second, these symptoms tendto be indicative of discrete—perhaps independent—problembehaviors or delinquentacts, making internal consistency estimatesa less than optimal method of determiningthe reliability of this symptom domain.Although these reliability data are basedon an earlier version of the DISC, the DISC2.3, Shaffer et al. (2000) provide initial reliabilityestimates for the DISC-IV from asample of 84 parents and 82 children, ages9–17, selected from several outpatient psychiatricclinics. These data were derivedfrom the computer-administered version ofthe DISC-IV, the C-DISC-IV. Interviewswere conducted by lay interviewers with anaverage retest interval of seven days. Thepreliminary findings are consistent with theresults for the DISC-2.3, with kappa coefficientsranging from .43 (Conduct Disorder)to .96 (Specific Phobia) for the parent reportand from .25 (Simple Phobia) to .92 (MajorDepressive Episode) for the child report.ValidityThe validity of diagnostic interviews isoften assessed by comparing the results ofstructured interviews to diagnoses made byexperienced clinicians. For example, Piacentiniet al. (1993) reported moderate tostrong agreement between the results ofDISC interviews and clinician diagnoseswhen the Parent DISC-2.3 was used (averagekappa = .50) but low agreement based onthe Child DISC-2.3 (average kappa = .34).Combining the two interviews gave agreementestimates between those of eitherinformant alone (average kappa = .41).These authors reported that most of thecases with disagreements between cliniciandiagnoses and the DISC-2.3 were cases thatwere close to the diagnostic threshold. Forexample, several disagreements emergedin which children had seven symptoms ofADHD (rather than the required eightsymptoms in DSM-III-R criteria) and werenot given the diagnosis according to theDISC-2.3 but were given the diagnosis ofADHD by the clinician.However, another study (Lewczyk et al.,2003) found relatively poor correspondencebetween diagnoses based on the DISC-IVand diagnoses by clinicians. The DISC-IVyielded higher rates of diagnoses on anxietydisorders, ODD, CD, and ADHD, butclinician diagnoses of depression were morecommon.Friman et al. (2000) provided a uniqueinvestigation of the predictive validityof the DISC by comparing interviewdata to behavioral observations recordedin a residential treatment program. Theresearchers examined both convergent anddiscriminant validity across a lengthy timeinterval (i.e., one year). Validity data wereobtained on 369 children, aged 9–17, whowere administered a computerized versionof the DISC-Y 2.3, the C-DISC-Y 2.3,upon enrollment in the residential programand at a one year follow-up. Diagnoses ofboth Oppositional Defiant Disorder andConduct Disorder were compared to dailyobservations of disruptive behavior thatwere coded by program staff and summedto form monthly behavior ratings of bothoppositional and conduct problem behaviors.Youth meeting criteria for a DISC-2.3diagnosis of ODD or CD upon enrollmentexhibited significantly greater observedbehavioral difficulties on program entrythan youth not meeting a diagnosis foreither disorder. Furthermore, change indiagnostic status across the two assessmentperiods predicted changes in observeddisrupted behavior across the same timeinterval. For example, youth who met criteriafor an ODD/CD diagnosis at Time

chapter 11 structured diagnostic interviews2691 but not at Time 2 were characterizedby a downward pattern of observed antisocialbehavior in the months separatingthe interviews. This is contrasted withyouth whose observed antisocial behaviorincreased as they moved from no diagnosisat Time 1 to diagnosis at Time 2.As further evidence for the validity of theDISC interview, Edelbrock and Costello(1988) found strong associations betweenthe diagnoses of ADHD, Conduct Disorder,and Depression/Dysthymia from theDISC-P and the Hyperactive, Delinquent,and Depressed scales of a previous versionof the Child Behavior Checklist (Achenbach& Edelbrock, 1983) in a sample of270 clinic-referred children between theages of 6 and 16. High rates of agreementwere also found between the originalDISC-P and the CBCL in another studyof 40 psychiatric referrals and 40 pediatricreferrals (Costello, Edelbrock, & Costello,1985). In contrast, the relation between theCBCL and the child version of the DISCtended to be much lower. However, it isimpossible to determine whether the lowcorrelations with the DISC-C were dueto differences in informants (i.e., parentcompletedCBCL and child-respondentDISC) or to differences in the assessmentinstruments themselves.ConclusionsStructured diagnostic interviews havebecome an important part of many clinicalassessments of children and adolescents.Like behavior rating scales, diagnostic interviewsprovide a reliable means of assessing achild’s emotional and behavioral functioning.In this chapter, we have attempted tohighlight the advantages and disadvantagesof using diagnostic interviews. Diagnosticinterviews enhance clinical assessments byproviding a format for determining howlong a child’s problems have been occurring,for determining the temporal sequencing ofbehaviors, and for estimating the degreeof impairment associated with a child’semotional or behavioral problems. Theseimportant parameters of a child’s emotionaland behavioral functioning are often notassessed by other assessment modalities.In addition, diagnostic interviews are typicallytied to the most recent revisions of theDSM, which closely links assessment withthis system of classification.On the negative side, diagnostic interviewsare often time intensive, and they typicallydo not provide any norm-referencedinformation on a child’s functioning abovethat which is accorded by DSM criteria. Inaddition, diagnostic interviews typically donot include a format for obtaining informationfrom a child’s teacher, and their reliabilityin obtaining self-report information foryoung children (i.e., below age 9) is somewhatquestionable, although pictorial orother formats provide a promising methodof enhancing their usefulness in this youngage group. As a result of these weaknesses,diagnostic interviews are best used as partof a more comprehensive assessment battery.We have attempted to provide guidelinesfor their use in this capacity. We havealso attempted to provide an overview of themost commonly used diagnostic interviewsfor assessing children and adolescents, highlightingthe major commonalities and differencesacross interviews. We concludedthe chapter with a more detailed discussionof the DISC-IV as an example of a typicaldiagnostic interview schedule designed foruse with children and adolescents.Chapter Summary1. Structured diagnostic interviews consistof a set of questions to be asked of achild or adolescent with explicit guidelineson how the youth’s responses areto be scored.

270 chapter 11 structured diagnostic interviews2. Most of the commonly used structuredinterviews are designed to assess diagnosticcriteria from one of the recent versionsof the Diagnostic and Statistical Manualof Mental Disorders. Therefore, if a goalof the evaluation is to make or clarify aDSM diagnosis, structured interviews arean important assessment tool.3. Interviews vary on the degree of structureinherent in the interview formatand whether or not the interview assessesonly current episodes of the disorders.4. Structured interviews, like behavior ratingscales, obtain detailed descriptionsof a child’s emotions and behaviors frommultiple informants.5. Unlike rating scales, however, structuredinterviews allow for the assessmentof important parameters of a child’sbehavior, such as the duration of thebehavioral difficulties, the temporalsequencing among problems, and thedegree of impairment associated withthe difficulties.6. Most structured interviews are timeconsuming and are not good for makingnorm-referenced interpretations.7. Diagnoses derived from structured interviewsshould be viewed within the contextof other assessment instruments.8. Child self-report from diagnosticinterviews is typically not reliablebefore age 9.9. The NIMH Diagnostic InterviewSchedule for Children-Version IV(DISC-IV) is a prototypical structureddiagnostic interview for use with childrenand adolescents.a. The DISC is highly structuredso that it can be administered bytrained lay interviewers.b. There are child, parent, and experimentalteacher versions of theDISC.c. The DISC contains 358 core questionsand 1,800 follow-up questions that areasked contingent on a child’s responsesto the core questions.d. The DISC is organized in six modules:(1) Anxiety Disorders, (2) MoodDisorders, (3) Psychosis, (4) DisruptiveBehavior Disorders, (5) Alcoholand Substance User Disorders, and(6) Miscellaneous Disorders (i.e.,Eating Disorders, Elimination Disorders,Tic Disorders).10. Diagnostic interviews should only beused as part of a more comprehensiveassessment battery.

C h a p t e r 1 2Assessing Family ContextChapter Questionsl Why is assessing a child’s family contextimportant in the clinical assessment ofchildren’s emotional and behavioral functioning?l What dimensions of a child’s family contextare most important to the assessmentprocess?l How can one assess these importantdimensions of family functioning?IntroductionIn the summary of research on childhoodpsychopathology provided in Chap. 3, oneof the more important findings was thatchildren’s and adolescents’ emotional andbehavioral functioning was heavily influencedby the demands and stressors theyexperienced in their environment. As aresult, to truly understand a child’s or adolescent’spsychological adjustment, onemust not limit the assessment to obtainingcharacteristics of the youth but must alsoassess the important contexts that shape achild’s or adolescent’s behavior. There is nocontext more important to understanding achild or adolescent than the family context.Causal Role of FamilialInfluencesResearch on childhood psychopathologyconsistently suggests that factors within thefamily play a major causal role in the developmentof personality and psychopathology(Erickson, 1998). At times the causal role of271P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_12, © Springer Science+Business Media, LLC 2010

272 CHAPTER 12 assessing family contexta child’s family has been overemphasized,either by ignoring the potential effects thata child or adolescent can have on the family(Pardini, 2008), or by ignoring the factorsthat can influence both the family and child(Frick & Jackson, 1993). These caveatsindicate that a child’s family context is onlyone piece of a very complex puzzle in understandingthe psychological adjustment of achild or adolescent. However, research alsosuggests that it is a very important piece ofthe puzzle. Family factors play an integralrole in the causal theories of many typesof problems across many theoretical orientations.Box 12.1 provides examples ofthree different types of childhood problemsfrom three different theoretical perspectives,all of which emphasize the familycontext as a causal agent in a child’s adjustmentproblems. Therefore, if the goalof an assessment is to uncover the possiblecauses of a child’s or adolescent’s emotionalor behavioral difficulties and, thereby, pointthe way to important treatment goals, theassessment of family functioning is criticalto the assessment process.Box 12.1Childhood Problems in a Family Context: Examples of Three Problems and ThreeTheoretical OrientationsChildhood Anxiety: An OperantPerspectiveOne way of conceptualizing the causes ofchildhood anxiety has been from an operantperspective. In this model a child receivesreinforcement from his or her environmentthat functions to maintain and increase theanxious behavior. Often this reinforcementis provided in the family context. A goodexample is provided by Ross (1981) of an8-year-old girl, Valerie, with school phobia.Valerie refused to attend school, and an assessmentof her home context revealed that theconsequences of her refusal to go to schoolwere indeed quite positive. Instead of goingto school, Valerie was able to sleep an hourlater than her three siblings who were attendingschool. Until her mother left for work,she was allowed to follow the mother aroundthe house and then was taken to a neighbor’shouse for the day. At the neighbor’s houseValerie was free to do whatever she pleasedfor the rest of the day, such as playing gamesand making occasional trips to a corner storewhere she bought candy, gum, and soft drinks.It was clearly a comfortable routine for bothchild and mother, who thereby avoided Valerie’stemper tantrums.Childhood Aggression: A Social LearningPerspectiveGerald Patterson (Patterson, 1982) andhis colleagues at the Oregon Social LearningCenter have developed a social learningmodel for the development of aggression.In this model, family interactions provide atraining ground for a child to learn coercivemethods of controlling interactions with others.Through analyses of micro-social interactionsbetween parents and children, Pattersonoutlines the development of a coercive cyclethat develops between parent and child andwhich escalates through aversive conditioning.The cycle starts when a parent makes ademand of a child and a child reacts aversively(e.g., whines, becomes defiant). Rather thanpushing the child, the parent withdraws thedemand, which reinforces the child’s aversiveresponse. During the next phase of the cycle,the parent again makes a demand of the childbut decides not to give in to the child’s aversivebehaviors. As the child becomes more aversive(e.g., temper tantrum), the parent becomesmore aversive (e.g., yelling, spanking). As theparent becomes more aversive, the child eventuallycomplies, which reinforces the parent’sincrease in aversive behavior.(Continues)

CHAPTER 12 assessing family context273Box 12.1 (Continued)This cycle repeats itself over and overagain, leading to each party reinforcingincreasing levels of aversiveness in the other.This training in coercive responses is thencarried over by the child into other settingswith other people (e.g., teachers, peers). It isevident that the cycle is transactional: that is,both the child and the parent contribute tothe escalating cycle. However, for the purposeof the current discussion, it is evident howimportant the child’s family environment,especially parent–child interactions, is to thedevelopment of aggression within this theoreticalframework.Eating Disorders: A Family SystemsPerspectiveSargent, Liebman, and Silver (1985) describefamily characteristics that provide a context inwhich the psychological features of anorexia nervosafit and are adaptive.Families of a person with anorexia have beenfound to have parents who are overinvolved intheir child’s life. This overinvolvement preventsthe child with anorexia from perceiving her ownsensations, including hunger. It also prevents thechild from developing a sense of self-competenceand the ability to use problem-solving skills. Asthe anorexia worsens, the family becomes moreprotective and involved and further inhibits theaffected child from acting more maturely andadaptively. Families of a child with anorexia alsotend to have difficulty resolving conflict. As aresult of unresolved marital conflict, the parentshave difficulty collaborating to handle thechild’s symptoms and actually counteract eachother in their attempts. These are just a few ofthe family dynamics that family system theoristshave proposed to explain the developmentand maintenance of anorexia nervosa in a child.However, it clearly illustrates the primary roleof the family context for understanding a childwith an eating disorder.Family History and DifferentialDiagnosesTwo other facts emerge from research onchildhood psychopathology that point tothe importance of assessing a child’s familycontext in clinical assessments. First,childhood emotional and behavioral problemstend to be rather amorphous, lackingclear boundaries, more so than is the casein adult psychopathology (see Lilienfeld,2003). Stated another way, children withproblems often have multiple types of problems,and it is often difficult to know whatis primary and what might be secondary.Second, there seems to be a parent–childlink to many types of psychopathology,with parents and children showing similarpatterns of adjustment (McMahon &Dev Peters, 2002). Taking these two factstogether, assessment of the adjustment ofparents in whom the type of problem maybe more clearly defined may provide cluesto the primary problem of the child.An example from research on childhoodaffective disorders illustrates the useof family history data in making a differentialdiagnosis. Prior to adolescence, thediagnosis of a bipolar affective disorder isdifficult to make. But research suggeststhat a significant proportion of childrenwith a depressive disorder will develop abipolar disorder later in life (Geller, Fox,& Fletcher, 1993). Geller et al. foundthat obtaining a family psychiatric historyhelped to predict which children with adepressive disorder were at most risk fordeveloping a bipolar disorder. Specifically,the presence of a family history of a bipolardisorder significantly predicted which ofthe children with depression would laterbegin to cycle between manic and depressivestates. This study also illustrates theimportant treatment implications for mak-

274 CHAPTER 12 assessing family contexting differential diagnoses. Children whowere depressed and had a family historyof bipolar disorders were more likely tohave manic behaviors develop followingtreatment with anti-depressant medicationthan were the depressed children withouta family history of bipolar disorder. A casestudy in which family history informationwas used in making a differential diagnosisis provided in Box 12.2.Interpreting InformationProvided by the ParentMany of the assessment techniques discussedthroughout this book rely on thereport of family members in the assessmentof child or adolescent adjustment.As a result, another important reason forassessing a child’s family context is thatfactors within the family can affect theBox 12.2Family History and Differential Diagnosis: A Case Study of a 7-Year-Old Girl withSocial PhobiaClaire is a 7-year, 2 month-old girl who was inthe middle of the first grade when her teacherrecommended that she be tested at an outpatientmental health clinic. Her teacher was concernedthat she might have Attention DeficitHyperactivity Disorder. Claire seemed brightand capable of learning and, in fact, performedquite well in one-on-one situations with theteacher or teacher’s aide. However, in the generalclassroom setting, Claire rarely finished herwork. She was often noted to be staring off intospace and she had to be constantly redirectedback to her work. Her teacher emphaticallystated that Claire was not a behavior problem.In fact, Claire was quite quiet and reserved andeven had difficulty asking for help when it wasneeded.There were several differential diagnosesthat were considered in the psychologicalevaluation of Claire. A psychoeducationalevaluation that included an intelligence testand an academic screener indicated that Clairewas quite capable of learning at or above alevel expected for her age. Therefore, herproblems in school did not seem to be causedby the presence of an intellectual deficit or alearning disability. However, the differentialdiagnosis between an attention deficit disorderand an anxiety disorder was more difficult,as she exhibited many behaviors consistentwith both types of problems. Several piecesof information helped make the decision thatClaire’s primary problem was one of anxiety,and particularly, social anxiety.First, Claire’s attentional difficulties tendedto be much milder than would be expected forchildren with attention deficit hyperactivitydisorder, as indicated by structured interviewsconducted with Claire’s mother and teacherand rating scales completed by her motherand teacher. Second, Claire showed a numberof other symptoms of anxiety in social situations.For example, she refused to go to Sundayschool at church and to other social activities(e.g., parties) and she had one good friend in theneighborhood but would only spend time withher if they were alone. Third, Claire’s motherhad a history of agoraphobia that had led toseveral lengthy periods in which she could notleave the house because of her fears.In this example a family history of anxietywas just one piece of the assessment thathelped to make the differential diagnosis.However, it seemed to be an important piece.The diagnosis itself ended up being importantbecause rather than treatment focusingon Claire’s attentional problems, a treatmentstrategy that used systematic desensitizationto social situations was implemented, withClaire’s teacher reporting dramatic improvementsin Claire’s school performance by theend of the year.

CHAPTER 12 assessing family context275information provided by family memberson the child’s adjustment. To appropriatelyinterpret the information obtained fromparents and other family members, onemust understand those factors that couldinfluence a parent’s accuracy in providinginformation on a child. For example,a noncustodial parent involved in a custodydispute may try to inflate the problemsof a child in an effort to get a morefavorable court decision. In contrast, thecustodial parent may have motivations topresent the child in a more positive light.A second example would be parents whoare trying to have their child placed in aresidential treatment center and who mayinflate problems in an effort to justify thisplacement. These are just two examples ofa myriad of familial factors that can affecthow one interprets the information providedby family members.De Los Reyes and Kazdin (2005) providea review of several family factors thatcan influence parent ratings of the childincluding family stress, the parent–childrelationship, and the level of marital discord.De Los Reyes and Kazdin (2005) alsonote several aspects of parental adjustmentthat can influence how parents rate theirchildren’s adjustment. One area that hasbeen the focus of substantial research is onthe effects of depression on parents’ reportof their child’s adjustment. There havebeen numerous studies that have called intoquestion the accuracy of depressed mothers’reports about their children’s behavior (seeDe Los Reyes & Kazdin, 2005 and Richters,1992 for reviews). This research hassuggested that depressed mothers reportmore problems in their children than arereported by nondepressed parents andteachers and more than are detected usingdirect behavioral observations of the children.These findings have led many authors toconclude that parents’ depression leads toa distorted view of the children’s behavior.We summarize Richter’s (1992) review andcritique of this literature in Box 12.3.Box 12.3Research Note: Depressed Mothers as Informants About Their Children: ACritical ReviewJohn Richters (1992) conducted a critical reviewof the research on the effects of depression ona parent’s rating of a child’s behavior. Richterscited 17 studies that have been published callinginto question the accuracy of depressed mothers’reports. In general, depressed mothers havetended to report more behavior problems intheir children than the level reported by teachers,fathers, or children and greater than thatobserved in behavioral observations. All of thesestudies led researchers to the conclusion thatdepressed mothers’ perceptions of their children’sbehavior were biased by their own level ofdepression. However, Richters’s critical reviewof several methodological and interpretive problemsthat have plagued this body of researchcalls into question this depressive bias theory.The first major problem in these studies wasthe fact that most of the comparisons betweenmothers and other informants used measuresthat were discordant on either the types ofbehaviors assessed or the situation in whichthe behaviors were assessed. The best examplewas the frequent comparison between mothers’and teachers’ ratings on a behavior rating scalethat had a different item content. In this case,both the behaviors assessed and the situationin which the behaviors were being observedwere discordant. As a result, it is unclearwhether the differences between mothers andthe other raters were due to maternal depressionor to differences in the behaviors and/or situations being assessed. Only 27% of thecomparisons between depressed mothers and(Continues)

276 CHAPTER 12 assessing family contextBox 12.3 (Continued)other informants used ratings that were bothbehaviorally and situationally concordant.The second pervasive problem in this literatureis the fact that most of the studies (94%)did not demonstrate that the mothers’ overreportingwas systematically related to maternaldepression. Twenty-four percent of the studiessimply documented that depressed mothersreported more behavior problems in theirchildren than did nondepressed mothers.Seventy-one percent indicated that maternaldepression predicted variance in mothers’ ratingsof their children that was not accountedfor by criterion ratings provided by otherinformants. Richters argues that the mostdirect evidence for the depression distortionhypothesis would be if mother-criterion disagreementswere systematically related (correlated)with measures of maternal depression.The third problem discussed by Richters isthat most of the studies (94%) focused onlyon maternal depression. It is well establishedthat depression is related to other factorswithin the individual (e.g., other forms ofpsychopathology) and the environment (e.g.,marital satisfaction). Therefore, it is unclearwhether or not mothers’ disagreements withinformants were due to the depression or toother aspects of the mothers’ adjustment and/or concomitant stressors in the family environment.As a result of these problems, Richters suggestedthat we must be cautious in acceptingthe depression distortion hypothesis until morerefined research is conducted. In fact, Richterscites five studies that used better methodologyand found that depressed mothers agreedwith other informants as well or even betterthan nondepressed mothers. However, theseare only a few studies, and they are not withoutflaws themselves. At this point, however,clinical assessors should at least be aware of theissues, many of which are unresolved, in thisvery important body of research.Source: Richters, J. E. (1992). Depressed mothers as informants about their children: A critical review of theevidence for distortion. Psychological Bulletin, 112, 485–499.Like parental depression, there is alsoevidence that parental anxiety may influencea parent’s report of childhood problems(Briggs-Gowan, Carter, & Schwab-Stone,1996). Frick and colleagues (Frick, Silverthorn,& Evans, 1994) found that in a sampleof 41 clinic-referred children between the agesof 9 and 13, mothers tended to report moresymptoms of anxiety disorders than did thechild. This overreporting was systematicallyrelated to anxiety in the mother. Specifically,the more anxious the mother, the greater theoverreporting of anxiety in the child. Theseauthors also reported that maternal anxietywas not associated with overreporting ofother types of maladjustment but seemed tobe more specifically related to anxiety. Thispattern of results would be consistent withthe possibility that anxious mothers projecttheir anxiety symptoms onto their reports ofanxiety in their children.Assessing Family Functioning:General IssuesTo this point we have discussed several reasonswhy assessing the family is an importantpart of clinical assessments of a child oradolescent. In this section we discuss morespecifically what areas of family functioningshould be assessed and how this can beaccomplished. However, before discussingspecific areas and techniques, two generalpoints deserve mention.First, many of the behavior rating scalesthat were reviewed in previous chaptershave subscales that assess various aspects ofa child’s family context. For example, theparent-completed Personality Inventoryfor Children-2 (PIC-2; Lachar & Gruber,2001) and the child self-report PersonalityInventory for Youth (PIY; Lachar & gruber,

CHAPTER 12 assessing family context2771994) both include a Family Relations scale.Included in this scale are items assessingmarital stability, consistency in discipline,emotional tone of family, community connectedness,and parental adjustment. TheMMPI-A contains a supplementary contentscale, the Adolescent-Family Problems scale(Archer, 1992), which includes 35 itemsassessing an adolescent’s perceptions of familyconflict, level of love and acceptance inthe home, family communication, and emotionalsupport provided by the family. TheBASC-2 Self-Report Scale (Reynolds &Kamphaus, 2004) contains a Relations withParents scale that assesses a child’s perceptionsof being important in the family, thequality of parent–child interactions, and thedegree of parental trust and concern. All ofthese scales provide a time-efficient screeningof many important aspects of a child’sfamily environment. The main drawbackis that each scale combines many differentaspects of a child’s family environment, makingit impossible to uncover specific areas ofstrength and/or dysfunction that could beimportant in understanding a child and inmaking treatment recommendations.This criticism leads to our next generalcomment for assessing a child’s family context.What areas to assess and how rigorousthe assessment should be within theseareas will vary depending on the purposeof the evaluation. In the sections that followwe make the case that several aspectsof the family should be routinely assessed:parenting style and parenting practices, parentingstress, marital conflict, and parentaladjustment. The depth of the assessmentin each area and which additional areas offamily functioning should be assessed willvary depending on the individual case. Forexample, the assessment of a child by aschool psychologist to document emotionaland behavioral factors that might be impairingacademic performance may include onlyminimal assessment of the child’s perceptionof the family environment and only as it mayinfluence his or her behavior in the classroom.In contrast, an assessment designedto assess a child’s adjustment to a recentparental divorce in order to make treatmentrecommendations on factors that could aidin the child’s post-divorce adjustment mayinclude a substantial family component. Thisassessment will most likely include obtainingextensive information on the level of parentalconflict and level of parental cooperation inchild-related issues, as these factors are crucialto understanding a child’s adjustment todivorce (Amato & Keith, 1991).Another type of assessment that requiresvery detailed and somewhat specializedassessment of family functioning is in thecase of known or suspected child abuse.A recommended assessment strategy forcases of child abuse is summarized in Box12.4. These examples illustrate the pointthat how intensive the assessment of familyfactors will be and which familial factorswill be assessed may vary somewhat fromcase to case.General Considerationsin Assessing FamilyFunctioningIn the subsequent sections, we reviewseveral critical areas of family functioningthat we feel are particularly importantin the clinical assessments of children andadolescents. In each case, we provide abrief overview of the research supportingthe importance of each aspect offamily functioning for understanding achild’s adjustment. This is followed by asummary of some commonly used measuresto assess that domain of functioning.It is important to note that most ofthe assessment methods that were chosenfor review were parent-report or childreportmeasures of family functioning.This was done for several reasons. First,these methods typically are the most time

278 CHAPTER 12 assessing family contextBox 12.4Research Note: A Child Abuse and Neglect Assessment StrategyCrooks and Wolfe (2007) outline a conceptualmodel to guide assessments of child abuse andneglect. Their model emphasizes the need tounderstand, not only the abusive behavior ofthe parent, but the family context in which theabuse takes place. They note that “the impactof maltreatment depends on not only theseverity and chronicity of the abusive eventsthemselves but also how such events interactwith the child’s individual and family characteristics”(p. 646). As result, an assessmentmust focus on a myriad of individual, familial,and cultural factors that research has relatedto child abuse and neglect, as well as the possibleprotective factors that can reduce theimpact of these risk factors.On the basis of this view of child abuseand neglect, Crooks and Wolfe suggest thatmost assessments need to be comprehensiveand need to address the following generalpurposes: (1) identify the general strengthsand needs of the family system; (2) assessparental responses to the demands of childrearing;(3) identify the needs of the child; and(4) assess parent–child relationship and abusedynamics. A summary of the important assessmentobjectives that follow from these overallgoals is provided below. Interested readers arereferred to the Crooks and Wolfe chapter inwhich they provide recommendations for specifictechniques to accomplish each of thesegoals. Many of these techniques are reviewedin other chapters of this text.Goal 1: Identify General Strengths andProblem Areas of Family SystemA. Family Background1. Parental history of rejection and abuseduring own childhood.2. Discipline experienced by parents duringown childhood.3. Family planning and effect of childrenon the marital relationship.4. Parents’ preparedness for and sense ofcompetence in child rearing.B. Marital Relationship1. Length, stability, and quality of maritalrelationship.2. Degree of conflict and physical violencein marital relationship.3. Support from partner in child rearing.C. Areas of Perceived Stress and Supports1. Employment history and satisfaction ofparents.2. Economic stability of family.3. Social support for parents, both within andoutside the family (e.g., number and qualityof contacts with extended family, neighbors,social workers, and church members).D. Parental Physical and Mental Health1. Recent or chronic health problems2. Drug and alcohol use3. Emotional disturbance and social dysfunctionGoal 2: Assess Parental Responsesto Child-Rearing DemandsA. Emotional Reactivity of Parent1. Parents’ perception of how abused childdiffers from siblings and other children.2. Parents’ feelings of anger and loss ofcontrol when interacting with child.3. Typical methods of coping with arousalduring stressful episodes.B. Child-Rearing Methods1. Appropriateness of parental expectationsfor child behavior, given child’sdevelopmental level.2. Typical methods used by parents forcontrolling/disciplining the child.3. Willingness of parents to learn newmethods of discipline.4. Parents’ perception of effectiveness ofdiscipline strategies.5. Child’s response to discipline attempts.(Continues)

CHAPTER 12 assessing family context279Box 12.4 (Continued)Goal 3: Identify Needs of the ChildA. Child Social, Emotional and BehavioralFunctioning1. Behaviors that may place this child atrisk for abuse.2. Problems in adjustment resulting fromabuse and living in family with multiplestressors.B. Child Cognitive and Adaptive Abilities1. Identify child’s developmental leveland coping capacity to determine mostappropriate method and level of intervention.2. Determine if abuse or chronic familystressors have led to cognitive delaysor delays in the child’s development ofadaptive behaviors.3. Child’s attributions for the abuse andreaction to family difficulties.Goal 4: Assessing Parent–Child Relationshipand Abuse DynamicsA. Risk of parent for future abuse andneglect.B. The quality of the parent-relationships.C. Parental empathy toward children’s feelingsSource: Crooks, C. V., & Wolfe, D. A. (2004). Child abuse and neglect. In E. J. Mash & R.A. Barkley (Eds.),Assessment of childhood disorders (4th ed., pp. 639–684). New York: Guilford Press.efficient method for collecting informationon the child’s family. Second, thesemeasures are often standardized and easilyobtainable. Third, these rating scalestend to have the best normative data thatallow for interpretations of scores basedon some comparison group. Thus, theserating scales tend to be the most useful inmany clinical assessments.However, such assessment methods arenot without limitations. Morsbach andPrinz (2006) reviewed eight measures thatuse parent-report of their own parentingbehaviors. Their evaluation of these measuressuggests that most demonstratedacceptable internal consistency (.70 andabove). Further, most measures showedmoderate concordance between parent andchild ratings of parenting (.23–37) withsomewhat higher concordance betweenreports of the two parents and betweenparent report and observations of parentingbehavior. However, these authorsnoted that parents are often asked to makeestimates of high-frequency behaviors (e.g.,yelling) over long periods of time (e.g., 6months) which may make accurate reportsdifficult. Further, they noted that many ofthe questions deal with sensitive issues thatmay not be socially desirable and may beconsidered intrusive by parents. As a result,such questions could result in biases intheir responses. Finally, these authors alsonoted that many scales often include vaguequantifiers (e.g., frequently, sometimes,never) that may also influence the accuracyof parents’ responses.Thus, although we have chosen to focuslargely on ratings scales for the assessmentof family functioning for the reasons notedabove, it is important to recognize the limitationsin the information obtained by thisassessment format. As noted by McMahonand Frick (2007), information obtainedby these measures should be interpretedwith other assessment information, suchas interviews and behavioral observations,whenever possible.

280 CHAPTER 12 assessing family contextParenting Stylesand PracticesThere is broad consensus that parentingbehaviors exert a significant influence onchild development. There is less consensusregarding the specific aspects of parentingthat are most crucial to child adjustment.However, Darling and Steinberg (1993)provide a good context for conceptualizingparenting and its effect on child and adolescentadjustment.Darling and Steinberg (1993) divide parentinginto two main components: parentingstyles and parenting practices. Theseauthors define parenting styles as “a constellationof attitudes toward the child thatare communicated to the child and createan emotional climate in which the parents’behaviors are expressed” (p. 493). Theseauthors use Baumrind’s (1971) typologyto exemplify parenting style. Baumrinddivides parenting styles into three types.The authoritarian style is characterized by arule-adherence orientation that de-emphasizesautonomy and emotional support. Thepermissive style is a child-centered style inwhich child autonomy is of primary importanceand rules and demands are minimal.The authoritative style is characterized byemotional support and respect for appropriateautonomy in the child but in the contextof clearly defined and consistently enforcedrules. It is this last parenting style, Authoritative,that research has consistently linkedto healthier child adjustment.In contrast to parenting style, parentingpractices are defined as the techniquesused by the parent to socialize their childand enforce rules. For example, a specificdiscipline practice (e.g., degree of corporalpunishment), use of positive parenting strategies(e.g., praise and reward for appropriatebehavior), consistency in parenting, andappropriate supervision and monitoring ofa child’s behavior by a parent are all examplesof parenting practices that have beenlinked to child adjustment (Frick, 1994).The unique contribution of Darling andSteinberg’s model of parenting is not onlyits explicit distinction between parentingstyle and parenting behaviors but also itsclear specification of how these factorsinteract to influence child development.Specifically, parenting style provides a contextin which parenting behaviors influencea child’s development. As a result, the sameparenting behavior may have differenteffects on a child depending on the parentingstyle. For example, there is a generallyaccepted association between adolescents’school performance and their parents’involvement in their schooling. However,the effectiveness of parents’ school involvementin facilitating academic achievementhas been found to be greater among parentswho have an authoritative parentingstyle than among parents who show anauthoritarian parenting style (Steinberg,Lamborn, Dornbusch, & Darling, 1992).The implications of this model of parentingare important for clinical assessmentsof children and adolescents. Itsuggests that, to understand the effects ofparenting on a child or adolescent’s development,one must assess both parentingstyle and parenting practices. In the followingsections, we provide a review of somemeasures that have been used to assessboth parenting style (The Family EnvironmentScale – Moos & Moos, 1986) andparenting practices (Alabama ParentingQuestionnaire – Shelton, Frick, & Wootton,1996; the Parenting Scale – Arnold,O’Leary, Wolff, & Acker, 1993; DyadicParent–Child Interaction Coding System;Eyberg, Nelson, Duke, & Boggs, 2005).Family Environment Scale-Second Edition (FES)The FES (Moos & Moos, 1986) is a90-item true-false questionnaire that iswidely used to assess persons’ perceptionsof their family environment. It is one of themost widely used instruments for assessing

CHAPTER 12 assessing family context281family processes (Piotrowski, 1999). It hasbeen used to assess family functioning ina wide variety of cultures (Bao-Yu & Lin-Yan, 2004; Teufel-Shone et al., 2005) and infamilies of children with a range of adjustmentproblems including Attention DeficitHyperactivity Disorder (Pressman et al.,2006), anxiety disorders (Suveg, Zeman,Flannery-Schroeder, & Cassoano, 2005)and affective disorders (Belardinelli et al.,2008).The FES can be completed by the parentand/or child (over 11 years). There arethree forms of the FES. We focus on theReal Form (Form R) of the FES, whichmeasures the respondent’s actual perceptionsof the family environment. However,there are also two special forms of theFES. The Ideal Form (Form I) allows therespondent to answer items in terms of thetype of family he or she would ideally like.The Expectations Form (Form E) allowsthe respondent to answer items in termsof what he or she expects family environmentsto be like.ContentThe FES is divided into ten subscalesfrom three domains: Relationships, PersonalGrowth, and System Maintenance. Adescription of the ten subscales within thesedomains is provided in Table 12.1. Theitem content was primarily developed onthe basis of family systems theory. This isevident from the emphasis on family structureand organization and the focus on thetransactional patterns between members ofTable 12.1 Subscales of the Family Environment ScaleDimension Subscale Description of Item ContentRelationship Cohesion Commitment, help, and support providedby family membersExpressivenessExtent to which family members areencouraged to express feelingsConflictAmount of anger, aggression, and conflictamong family membersPersonal growth Independence Extent to which self-sufficiency, assertiveness,and independence are encouraged inthe familyAchievement orientation Extent to which activities of familymembers are achievement-oriented andcompetitiveIntellectual-cultural orientation Degree of interest in political, social, andcultural activitiesActive-recreational orientation Emphasis placed on participation in socialand recreational activitiesMoral-religious emphasis Importance placed on ethical and religiousissuesSystem maintenance Organization Importance placed on having a clear familystructure and well-defined rolesControlDegree to which rules and procedure forfamily are explicitSource: Moos & Moos (1986).

282 CHAPTER 12 assessing family contextthe family in the FES item content. Also, asevident from the content, the FES is bestthought of as a measure of parenting styleor family climate, rather than of specificparenting practices.NormsThe FES manual reports information on alarge normative sample that included 1,125families from all regions of the country,single-parent and multi-generational families,families drawn from ethnic minoritygroups, and families of all age groups(Moos & Moos, 1986). It is unclear howrepresentative the sample is on each ofthese variables. However, the authors notethat 294 families from the normative samplewere drawn randomly from specifiedcensus tracts in the San Francisco area, andthe means and standard deviations of FESscales did not differ between this group andthe rest of the normative sample.Although the normative group includesfamilies of all age groups, it is notable thatthe majority of the normative samples wasbased on the reports of adults, with muchless data available on the reports of childrenand adolescents. This is important becausethe authors found small but systematic differencesbetween the scales completed byparents and adolescents (Moos & Moos,1986). Specifically, adolescents perceivedless emphasis on cohesion, expressiveness,independence, and intellectual/religiousorientation and more emphasis on conflictand achievement than did their parents.Reliability and ValidityThe ten subscales of the FES generallyhave been shown to have acceptable levelsof reliability in many samples. Themanual reports internal consistency estimatesin a large community sample (n= 1,067) ranging from a = 61 to a = 78(Moos & Moos, 1986). Two-month testretestreliability in a smaller communitysample (n = 47) ranged from r = 68 to r =86. One note of caution for the reliabilityof the FES is finding that the reliabilityof the scales may be lower in adolescentsamples (Boyd, Gullone, Needleman, &Burt, 1997).Moos and Moos (1986) provide a goodsummary of over 100 research articles usingthe FES, which attests to its correlationwith other measures of family functioning,its ability to differentiate distressed fromnondistressed families, and its sensitivityto treatment effects. Many more studieshave been published since this review(Piotrowski, 1999).As noted above, the FES has beenwidely used with several different clinicalpopulations of children and adolescents.For example, using the FES, Suveg et al.(2005) noted that mothers of children withan anxiety disorder showed less emotionalexpressiveness than non-clinic referredchildren (ages 8–12). As another example,Pressman et al. (2006) found that familiesof children and adolescents with AttentionDeficit Hyperactivity disorder reportedhigher rates of family conflict on the FES.Finally, Lucia and Breslau (2006) reportedthat higher levels of family cohesion, asmeasured by the FES when children wereage 6, were associated with fewer emotionaland attentional problems in thechildren at age 11.Probably the biggest threat to the validityof the FES is the failure to validate thescale structure through factor analyses.The scales were designed primarily on thebasis of content and face validity. Unfortunately,factor analyses have generallyisolated anywhere from two (e.g., Fowler,1982) to six (Sanford, Bingham, & Zucker,1999) or seven (Robertson & Hyde, 1982)factors on the FES. No study has providedconvincing evidence supporting the tenscalestructure that is the basis for mostinterpretations from the FES.

CHAPTER 12 assessing family context283Alabama ParentingQuestionnaire (APQ)The APQ (Shelton et al., 1996) is a measureof parenting behavior that was developedfor use with parents of elementary-schoolagedchildren and adolescents (6 to 17years old). However, it has been used insamples as young as ages 3 and 4 (Clerkin,Marks, Policaro, & Halperin, 2007; Dadds,Maujean, & Fraser, 2003) with some modificationof its content. It consists of 42items that are presented in both globalreport (i.e., questionnaire) and telephoneinterview formats, and there are separateversions of each format for parents andchildren. Also, Elgar, Waschbusch, Dadds,and Sivaldason (2007) have developed a9-item short version of the scale for use asa brief screener.Most of the research using the APQ hasused the questionnaire formats. Items onthis format are rated on a five-point Likerttypefrequency scale and ask the informanthow frequently each of the various parentingpractices typically occurs in the home.On the telephone interview format, fourinterviews are conducted with parents andchildren with at least 3 days between eachinterview. The informant is asked to reportthe frequency with which each parentingpractice has occurred over the previous 3days and responses for each item are averagedacross the four interviews.ContentThe content of the APQ was developedto assess the five dimensions of parentingthat have been most consistently relatedto behavior problems in youth: Involvement,Positive Parenting, Poor Monitoring/Supervision,Inconsistent Discipline,and Corporal Punishment (Shelton et al.,1996). It also includes several other itemsassessing “other discipline practices,” suchas use of time out or taking away privileges.The items on the APQ and its subscales areprovided in Table 12.2. The items used onthe 9-item short version are also designatedin this table (Clerkin et al., 2007). Studiesusing the APQ often have used scores fromthe individual scales (e.g., Frick, Christian,& Wootton, 1999) or they have used compositesof these scales (e.g., Frick, Kimonis,Dandreaux, & Farrell, 2003). ThereTable 12.2 Subscales of the Alabama Parenting QuestionnaireInvolvement1. You have a friendly talk with your child4. You volunteer to help with special activities that your child is involved in (such as sports, boy/girlscouts, church youth groups)7. You play games or do other fun things with your child9. You ask your child about his/her day in school11. You help your child with his/her homework14. You ask your child what his/her plans are for the coming day15. You drive your child to a special activity20. You talk to your child about his/her friends23. Your child helps plan family activities26. You attend PTA meetings, parent/teacher conferences, or other meetings at your child’s school(Continues)

284 CHAPTER 12 assessing family contextTable 12.2 (Continued)Positive parenting2. You let your child know when he/she is doing a good job with something a5. You reward or give something extra to your child for obeying you or behaving well13. You compliment your child when he/she does something well a16. You praise your child if he/she behaves well a18. You hug or kiss your child when he/she has done something well27. You tell your child that you like it when he/she helps around the housePoor monitoring/supervision6. Your child fails to leave a note or to let you know where he/she is going a10. Your child stays out in the evening past the time he/she is supposed to be home a17. Your child is out with friends you do not know a19. Your child goes out without a set time to be home21. Your child is out after dark without an adult with him/her24. You get so busy that you forget where your child is and what he/she is doing28. You don’t check that your child comes home from school when he/she is supposed to29. You don’t tell your child where you are going30. Your child comes home from school more than an hour past the time you expect him/her32. Your child is at home without adult supervisionInconsistent discipline3. You threaten to punish your child and then do not actually punish him/her a8. Your child talks you out of being punished after he/she has done something wrong a12. You feel that getting your child to obey you is more trouble than it’s worth22. You let your child out of a punishment early (like lift restrictions earlier than you originally said) a25. Your child is not punished when he/she has done something wrong31. The punishment you give your child depends on your moodCorporal punishment33. You spank your child with your hand when he/she has done something wrong35. You slap your child when he/she has done something wrong38. You hit your child with a belt, switch, or other object when he/she has done something wrongOther discipline practices34. You ignore your child when he/she is misbehaving36. You take away privileges or money from your child as a punishment37. You send your child to his/her room as a punishment39. You yell or scream at your child when he/she has done something wrong40. You calmly explain to your child why his/her behavior was wrong when he/she misbehaves41. You use time out (make him/her sit or stand in corner) as a punishment42. You give your child extra chores as a punishmentaItems that are included in the 9-item short screening version of the APQ (Clerkin et al., 2007).

CHAPTER 12 assessing family context285are two common ways of forming compositescores. The first way is to standardizethe scores (e.g., create z-scores) and thencombine the Involvement and PositiveParenting scales into a Positive ParentingComposite and the Poor Monitoring/Supervision, Inconsistent Discipline, andCorporal Punishment scales into a NegativeParenting Composite. The secondmethod for forming composites from theAPQ is to create a single DysfunctionalParenting composite by standardizing allfive scales, inversely scoring the positiveparenting scales, and summing all fivedimensions.NormsOne of main limitations in the APQ is thelack of norm-referenced scores that can beused to interpret the scales. However, thereare now two studies that provide scoresfrom fairly large samples of non-referredchildren, from which cut-scores can bedeveloped. The first study was conductedwith 1,402 children ages 4 to 9 (Elgar et al.,2007) in Australia and the second studywas conducted with 1,219 German schoolchildren ages 10 to 12 (Essau, Sasagawa, &Frick, 2006).Reliability and ValidityAs noted previously, most of the publishedresearch using the APQ to date has utilizedthe global report formats of the scale,with the exception of Shelton et al. (1996).Across studies, the reliability and stability ofAPQ scores have generally been acceptablewith several notable exceptions. First, theinternal consistency of the short three-itemCorporal punishment scale has often beenquite low on all formats. Second, the internalconsistency of the Poor Monitoring/Supervision scale has been low in the interviewformat (all alphas below .50 – Sheltonet al., 1996). Third, Frick et al. (1999)showed poor reliability of the child-reportformats in very young children (below age9) (see also Shelton et al., 1996).Several studies have provided factorialsupport for the five dimensions aroundwhich the scale was developed (Elgar et al.,2007; Essau et al., 2006). Also, parent ratingson the APQ are significantly associatedwith observations of parenting behaviorin 4- to 8-year-old boys (Hawes & Dadds,2006). However, the most common use ofthe APQ has been to study parenting infamilies of children with conduct problems.An association between APQ scales and conductproblems has been reported in community(Dadds et al., 2003), clinic-referred(Frick et al., 1999; Hawes & Dadds, 2006),and inpatient samples (Blader, 2004), as wellas in families with deaf children (Brubaker& Szakowski, 2000) and families with substance-abusingparents (Stanger, Dumenci,Kamon, & Burstein, 2004). Also, thesestudies have documented this relationshipin samples as young as age 4 (Dadds et al.,2003; Hawes & Dadds, 2006) and as old asage 17 (Frick et al., 1999).Importantly, Frick et al. (1999) reportedsome differences in which dimensions ofparenting were most strongly associatedwith conduct problems at different ages,with Inconsistent Discipline being moststrongly associated in young children (ages6–8), Corporal Punishment being moststrongly associated in older children (ages9–12), and Involvement and Poor Monitoring/Supervisionbeing most stronglyrelated in adolescents (ages 13–17).Although the most common use of theAPQ has been to assess parenting in familiesof youth with conduct problems, it hasbeen used to assess family correlates toanxiety disorders (Pfiffner & McBurnett,2006) and to assess parenting in families ofdepressed parents (Cummings, Keller, &Davies, 2005). Finally, several studies haveused the APQ scales to test changes in parentingbehaviors following treatment (e.g.,Feinfield & Baker, 2004; Hawes & Dadds,2006).

286 CHAPTER 12 assessing family contextParenting Scale (PS)The PS (Arnold et al., 1993) is anothercommonly used measure of parenting. It isa 30-item parent report scale that focusesspecifically on parents’ attitudes and beliefsabout discipline. The items are all rated ona seven-point scale in which the parent isasked to estimate the probability with whichthey would use a particular discipline strategy(e.g., when my child misbehaves, I spank, slap,grab, or hit my child). It was originally developedfor use with young children (ages 18–48months; Arnold et al., 1993) and has primarilybeen used to assess parenting in preschoolchildren. However, there is evidence for itsutility in samples of children as old as 11 yearsof age (Prinzie, Onghena, & Hellinckx, 2007;Steele, Nesbitt-Daly, Daniel, & Forehand,2005).ContentThe PS items can be grouped into threedimensions: Laxness, Overreactivity, andVerbosity. There has been some debateover the appropriateness of this three scalestructure because it has obtained factoranalytic support in some studies (i.e., Arney,Rogers, Baghurst, Sawyer, & Prior, 2008;Arnold et al., 1993; Rhoades & O’Leary,2007) but not in others (Prinzie et al., 2007;Reitman et al., 2001; Steele et al., 2005).The difference in the factor analyses typicallyinvolves whether the Verbosity factoremerges as a separate dimension. Also,Rhoades and O’Leary (2007) developed aPS-Should scale that was designed to assesshow parents believe they “should disciplinetheir children,” rather than assessing theirreport of actual discipline practices.NormsOne of main limitations in the PS is thelack of norm-referenced scores that canbe used to interpret the scales. However,there are several studies that provide scoresfrom fairly large samples of non-referredchildren from which cut-scores can bedeveloped. Arney et al. (2008) provideddata from 1,656 mothers of children (ages3–5) from South Australia and Prinzieet al. (2007) provided data from 596 mothersand 559 fathers of children ages 5 to11 years in Belgium. Finally, Rhoades andO’Leary (2007) provided data on 453 familiesof children ages 3 to 7 years from thenortheast United States.Reliability and ValidityMost studies of the PS show adequateinternal consistency and test-retest reliabilityfor the PS (e.g., Arnold et al., 1993;Rhoades & O’Leary, 2007) with the exceptionof the Verbosity scale (Arney et al.,2008). Further, the PS has been shownto be correlated with other measures ofparenting practices (Rhoades & O’Leary,2007; Steele et al., 2005) and has beenassociated with measures of adjustmentproblems in children (Arney et al., 2008;Prinzie et al., 2007). Scores on the PS havealso been shown to be sensitive to effectsof interventions designed to improve parentingbehaviors ( Sanders, Markie-Dadds,Tully, & Bor, 2000).Dyadic Parent–Child InteractionCoding System (DPICS)The DPICS (Eyberg et al., 2005) is a highlystructured coding system designed to assessmaternal behaviors and parent–child interactionsin several standard settings. In contrastto the other measures of parenting thathave been reviewed, the DPICS is an observationalsystem. It has typically been usedto code parent–child interactions of preschoolchildren (e.g., Eisenstadt, Eyberg,McNeil, Newcomb, & Funderbunk, 1993;Robinson & Eyberg, 1981). Parents and

CHAPTER 12 assessing family context287children are observed in two 5-min periods,typically in a clinic playroom setting.In the Child Directed Interaction (CDI)the parent is instructed to allow the child tochoose any activity and to play along withthe child. In the Parent Directed Interaction(PDI) the parent is instructed to selectan activity and to keep the child playingaccording to parental rules.ContentThe 5-min interactions are videotaped forlater coding. The DPICS includes a detailedmanual for coding several parent and childbehaviors. The system codes 12 parent and14 child behaviors. A summary of thesebehaviors included in the DPICS is providedin Table 12.3. In addition to discrete behaviors,several additional categories are includedin the DPICS to code sequences of behaviors.Parental responses (i.e., ignores or responds)to child’s defiant behavior and child responses(i.e., compliances, non-compliances, or noopportunity) to parental commands arecoded. The coding system is a continuousfrequency count of all behaviors observedduring the 5-min interaction periods.NormsThe normative information available onthe DPICS is quite limited. Robinson andEyberg (1983) provide data on 22 familieswith children between the ages of 2 and7. The sample was primarily two-parentfamilies (73%) and highly educated (meanof 15.2 years of education for parents). As aresult, the generalizability of this informationto other samples is questionable.Reliability and ValidityNot surprisingly, given the very detailedbehavioral descriptions provided by theDPICS manual, trained observers havebeen able to achieve quite high interraterreliability with the DPICS. In a sampleTable 12.3 Categories from the DyadicParent–Child Interaction Coding SystemMaternal behaviors1. Praise(a) Labeled praise(b) Unlabeled praise2. Command(a) Direct commands(b) Indirect commands3. Other verbalizations(a) Descriptive/reflective questions(b) Descriptive/reflective statements(c) Irrelevant verbalization(d) Verbal acknowledgment4. Responses to child behavior(a) Physical positive(b) Ignore(c) Critical statement(d) Physical negativeChild behaviors1. Deviant(a) Whine(b) Cry(c) Smart talk(d) Yell(e) Destructive(f) Physical negative2. Response to commands(a) Compliance(b) Noncompliance(c) No opportunitySource: Eyberg & Robinson (1983).of 42 families (20 clinic-referred and 22normal control) the mean interrater reliabilityfor parent behaviors was .91 andfor child behaviors was .92 (Robinson& Eyberg, 1981). In addition, DPICSscores have been shown to differentiatefamilies of clinic-referred children withconduct problems from families of normalcontrol children (Eyberg et al., 2005;Robinson & Eyberg, 1981). Scores from

288 CHAPTER 12 assessing family contextthe DPICS have also been shown to besensitive to interventions for families ofchildren with behavior problems (Eisenstadtet al., 1993; Eyberg & Robinson,1982; Hembree-Kigin & McNeil, 1995).Parenting StressThe second dimension of family functioningthat is critical to assess in most clinicalassessment of children and adolescents isparental stress. A high level of stress caninfluence children’s adjustment in a numberof ways, one of which is by making itmore difficult for a parent to use optimalparenting strategies (Whiteside-Mansellet al., 2007). For example, elevated stresscan lead to lower levels of parental warmthand higher rates of harsh parenting (Dopke,Lundahl, Dunsterville, & Lovejoy, 2003;Haskett, Ahern, Ward, & Allaire, 2006).There are two types of measures that canbe used in clinical assessment of childrenand adolescents. The first are measures ofgeneral life stress (e.g., life event scales) andthe second are measures of stress specific toparenting. Examples of general measures ofstress include the Life Experiences Survey(Sarason, Johnson, & Siegel, 1978) and theFamily Events List (Patterson, 1982). Suchmeasures of general stress have proven to beimportant for understanding children withbehavior problems (Johnston, 1996; Snyder,1991) and they have been related to abusivebehavior in parents (Whipple & Webster-Stratton, 1991). However, in the sectionsbelow, we focus on two measures of stressmore specifically related to parenting.Child Abuse PotentialInventory-Second EditionThe CAPI (Milner, 1986) is a 160-itemrating scale completed by a child’s parent.As the name implies, the CAPI was originallydeveloped to assess dimensionsof parental behavior that have provento be risk factors for physical abuse ofchildren. However, the CAPI assessesmultiple areas of family functioning thatare important in many clinical assessments.Further, several of the scales focusdirectly on stressors related to parentinga child or adolescent. The items requirea third-grade reading level to completeand each item is presented in a forcedchoice,agree-disagree format. The fullform takes approximately 15 min to complete.However, a brief 24-item versionof the scale has been developed and hasproven to be highly correlated with thefull version (Ondersma, Chaffin, Mullins,& LeBreton, 2005).ContentThe CAPI contains three validity scales: Lie,Random Response, and Inconsistency. TheLie scale was designed to detect tendenciesto distort responses in a socially desirablemanner. Both the Random Response andInconsistency scales were designed to detecthaphazard or random responses to itemswithout regard to item content. To test theusefulness of these validity scales, Milner andCrouch (1997) had two groups of parents, 106community volunteer parents and 80 parentsattending a program for parents at risk forabuse, complete the CAPI in several differentways: answering honestly, answering in a wayto make themselves “look good,” answeringin a way to make themselves “look bad,” andanswering inconsistently. These differinginstructions did affect how parents answeredthe CAPI questions, suggesting that parentscan intentionally distort their ratings. Withthe exception of detecting the faking-badcondition (58% correct identification), theCAPI validity indexes were good at detectingmost of the other response conditions, rangingfrom 82 to 100% correct identificationacross both samples of parents.There are six primary scales of theCAPI that are combined into a compositeAbuse scale. The items were developedfrom an extensive review of the child

CHAPTER 12 assessing family context289abuse and neglect literature. The Distressscale assesses parental anger, frustration,impulse control, anxiety, and depression.The Rigidity scale assesses parents’ flexibilityand realism in their expectationsof children’s behavior. It includes suchitems as “A child should never disobey,”“A child should always be neat,” and“A child should never talk back.” TheUnhappiness scale assesses a parent’s degreeof personal fulfillment as an individual, asa parent, as a marital/sex partner, and as afriend. Problems with Child and Self is ascale with items tapping parents’ perceptionsof their child’s behavior and theirperceptions of their own self-concept as aparent. The last two scales, Problems withFamily and Problems with Others, assessthe level of family conflict in the extendedfamily and the level of conflict with personsoutside the family or community agencies.NormsNormative information is available inthe CAPI manual (Milner, 1986) from asample of 836 parents, child care workers,and parent aides from Florida, California,North Carolina, Hawaii, Oklahoma,Illinois, New York, and West germany.It is unclear how this normative samplewas selected, and the representativenessof this sample in terms of parental education,socio-economic status, and ethnicityis also unclear. This is crucial informationbecause there is evidence that familyfunctioning can vary as a function of thesevariables (Maccoby & Martin, 1983). Asa result, lack of accessible information onthe normative sample hinders the abilityto make norm-referenced interpretationsfrom the CAPI scales.Reliability and ValidityThe CAPI manual provides evidence thatthe composite Abuse scale and the Distressand Rigidity scales exhibit acceptable internalconsistency and temporal stability. Thereliability of the four other individual scalestends to be more inconsistent across samples.In terms of validity, there is evidence thatCAPI scores are associated with documentedrisk factors for child abuse (Budd, Heilman,& Kane, 2000; Grietens, De Haene, &Uyteeborek, 2007; Haskett, Scott, & Fann,1995). Also there is evidence that the compositeAbuse scale can successfully discriminatebetween proven abusers and controlsubjects (Milner & Wimberley, 1980), andthis extends across cultural groups (Haz &Ramirez, 1998). Finally, the Abuse scalehas proven to be sensitive to the effects ofintervention with high-risk parents (Wolfe,Edwards, Manion, & Koverola, 1988).Therefore, it appears that the CAPI providesa reliable method of assessing dysfunctionalelements of a child’s family environment,including several aspects of parental stressthat are associated with child abuse. In addition,the composite Abuse scale does seemto be an index of risk for abuse, although itis important to recognize that many parentswho score high on the CAPI have no documentedevidence of abuse in the home (i.e.,false positive) (Haz & Ramirez, 1998).Parenting Stress Index-SecondEdition (PSI)The PSI (Abidin, 1986) is unique in itsfocus specifically on stressors related toparenting. It was primarily designed toassess the family context of preschoolchildren between the ages of 1 and 4,although it has been used in older samplesof pre-adolescent children. Completionof the PSI requires at least a fifth-gradeeducation. It contains 151 items andgenerally takes 20–30 min to complete.A short-form of the PSI has been developedwith 36 items (Abidin, 1995).ContentThe items of the full PSI are divided intotwo main categories: Child Domain (47

290 CHAPTER 12 assessing family contextitems) and Parent Domain (54 items). TheChild Domain consists of items that assessqualities of a child that make it difficult forparents to fulfill their parental role. TheParent Domain assesses sources of stressand disability related to parental functioning.Table 12.4 provides a summary of thescales that constitute the Child and Parentdomains. The PSI also allows for the computationof a composite score that providesan overall indicator of the amount of stressin the parent–child system.Table 12.4 Item Content of the Parenting Stress Index-Second EditionScale Items Characteristics of High ScorersChild domain 47 Child displays qualities that make it difficult for the parent to fulfillparenting rolesAdaptability 11 Child shows inability to change from one task to another withoutemotional upset, avoids strangers, is overreactive to changes in routineand difficult to calmAcceptability 7 Child is not as attractive, intelligent, or pleasant as the parent hadhoped or expectedDemandingness 9 Child is very demanding of parents’ time and energy, with patternssuch as frequent crying, frequent requests for help, and frequent minorproblem behaviorsMood 5 Child is frequently unhappy, sad, and cryingDistractibility/HyperactivityReinforcesparent9 Child displays overactivity, restlessness, distractibility, and short attentionspan, fails to finish things, and shifts from one activity to another6 Interactions between child and parent fail to produce good feelings inthe parent; associated with parental feelings of rejection and poor selfconceptas parentParent domain 54 Indicates significant stress on the parent–child system that is related todimensions of parental functioningDepression 9 Parent reports significant feelings of depression and guilt. High scoresmay prevent parent from mobilizing sufficient levels of psychic andphysical energy to fulfill parenting responsibilitiesAttachment 7 Parent does not feel emotional closeness to child and parent perceivesan inability to accurately read and understand child’s feelings and needsRestriction ofroleSense ofcompetence7 Parents feel that parental role restricts their freedom and impairs theirattempts to maintain own identity13 Parents do not feel that they can adequately fulfill their parental roleseither because of a lack of knowledge of child development or a limitedrange of child-management skillsSocial isolation 6 Parents perceive themselves as socially isolated from their peers,relatives, and other social support systemsRelationshipwith spouse7 Parents perceive that they do not receive emotional and physical supportfrom their spouse in area of child managementParent health 5 Parents report a deterioration in physical health that is impacting theirability to fulfill parental responsibilitiesSource: Abidin (1986).

CHAPTER 12 assessing family context291The 36-item short form of the scalehas been tested and items form two relativelydistinct factors of parental distressand dysfunctional parent child interactions(Haskett et al., 2006). Importantly, thecorrelation between the total scores on theshort and long form is generally quite high(e.g., r = 87; Abidin, 1995).NormsThe normative sample for the PSI consistedof 534 parents of children referredto a small group of pediatric clinics in Virginia;the median age of the children was9 months (SD = 23.2 months). The representativenessof the normative sample isone of the major weaknesses of the scale.The sample consisted of primarily White(92%), highly educated (1/3 with collegedegrees) parents from central Virginia.Thus, the use of norm-referenced scoresfor families that do not match these characteristicsis questionable.Reliability and ValidityThe manual of the PSI provides convincingevidence for the internal consistencyand temporal stability of the threecomposite scores: Total Stress, ParentDomain, and Child Domain. The reliabilitycoefficients for the individual scales,however, are much more variable andtypically exhibit relatively low reliabilityestimates. The manual provides one ofthe best summaries of the extensive use ofthe PSI in research on the family contextof preschoolers (see also Abidin, Flens, &Austin, 2006 for an updated review). Ingeneral, the PSI scales have been correlatedwith other measures of family functioning(Pinderhughes, Dodge, Bates,Petit, & Zelli, 2000), including correlationswith observations of parentingbehavior (Bigras, LaFreniere, & Dumas,1996). Also, the PSI has differentiatedfamilies who are experiencing major stressorsfrom nonstressed families (Holden& Banez, 1996; Whiteside-Mansell et al.,2007) and has proven sensitive to treatmenteffects (Nixon, Sweeney, Erickson,& Touyz, 2003).Also, factor analyses generally supportthe broad Parent and Child Domainsfor grouping the various PSI subscales,although some studies have provided supportfor a third Parent–Child InteractionDomain, which includes the subscalesof Child Acceptability, Child ReinforcesParent, and Parent Attachment to Child(Hutcheson & Black, 1996; Solis & Abidin,1991). Further, Bigras et al. (1996)reported that, in a sample of 218 mothersof preschoolers, the Parent and ChildDomains predicted parental, familial,and child outcomes different from thoseobtained from other sources. Specifically,the Parent Domain was more stronglyand independently associated with measuresof marital adjustment and maternaldepression, whereas the Child Domainwas more strongly and independentlyassociated with child difficulties reportedby the mother and children’s problemsobserved during parent–child interactions.These results are important in suggestingthat the two domains are valid inassessing somewhat independent dimensionsof family functioning.Marital ConflictThere is a long history of research showinga link between divorce and childbehavior problems. The most comprehensivesummary of this research comes fromAmato and Keith (1991). These authorsconducted a meta-analysis of 92 publishedstudies of the impact of divorce on a child’spsychological well-being. The combinedsamples from the 92 studies involvedover 13,000 children. This meta-analysisrevealed that divorce consistently had a

292 CHAPTER 12 assessing family contextnegative impact on several types of childwell-being (e.g., conduct problems, schoolachievement, social adjustment, and selfconcept).These studies suggested that therelationship between divorce and psychologicaldifficulties in children was greatestwithin the 2 years immediately followinga divorce.The meta-analysis also provided intriguingdata in support of the theory that it is theconflict that occurs between parents beforeand during the separation that has the mostdetrimental impact on a child’s adjustment(see also Emery, 1982). Whereas childrenof divorced families tended to have pooreradjustment than children in low- conflict,intact families, children in intact, high-conflicthomes tended to have the poorest adjustmentof all three groups. Also consistentwith this perspective, several studies foundthat less conflict and better divorce cooperationbetween parents predicted better postdivorceadjustment for children.The implications of these findings areimportant to clinical assessments of childrenand adolescents. They suggest that it isnot simply enough to determine the maritalstatus of a child’s parents for understandingthe potential impact of the parents’ maritalrelationship on the child. A more importantfocus of assessment is the overt conflictbetween parents that is witnessed by thechild.There are several marital inventoriesthat are frequently used in research andclinical practice, and often included in theclinical assessment of children and adolescents(McMahon & Frick, 2007). TheMarital Adjustment Test (MAT; Locke &Wallace, 1959) and the Dyadic AdjustmentScale (DAS; Spanier, 1976) are two selfreportinstruments that have been shownto produce reliable scores and differentiatepersons in distressed and nondistressedmarriages. However, these inventoriestend to focus on general marital satisfactionrather than on overt conflict per se.The O’Leary-Porter Scale (OPS; Porter &O’Leary, 1980) is a brief rating scale thatfocuses on overt marital conflict and, evenmore specifically, on marital conflict thatis witnessed by the child or adolescent. Asa result, the OPS is uniquely suited for usein clinical assessments of children and adolescents.O’Leary-Porter Scale (OPS)The OPS (Porter & O’Leary, 1980) wasdeveloped specifically for studying the associationbetween marital adjustment and childbehavior problems. The OPS is a 20-itemself-report inventory within which areembedded nine items that assess the degreeof marital conflict witnessed by the child.A parent rates on a five-point frequencyscale (Never to Very Often) how often thechild witnesses arguments between himselfor herself and the spouse over money, discipline,wife’s role in family, and personalhabits of the spouse. Two questions also askfor overall estimates of the amount of verbaland physical hostility between spouses thatis witnessed by the child.There is little normative data on theOPS. However, 2-week test-retest reliabilityin a sample of 14 families was found tobe quite high (r = 92) (Porter & O’Leary,1980). These authors also reported thatthe OPS was correlated with several typesof maladjustment in children. Highlightingthe importance of focusing specificallyon overt conflict, these authors found thatthe OPS was more consistently associatedwith child adjustment difficulties thanwas a measure of general marital satisfaction(i.e., the MAT). This associationbetween scores on the OPS has been replicatedin other studies (Forehand, Long,& Hedrick, 1987; Mann & MacKenzie,1996). Thus, it appears that the OPS capturesthe critical component of maritaldiscord in terms of its detrimental effecton child adjustment.

CHAPTER 12 assessing family context293Parental AdjustmentIn the introduction to this chapter, wediscussed several ways in which assessingparental psychiatric adjustment is criticalto the clinical assessment of children. Wediscussed two areas of research, one onparental depression and another on parentalanxiety, that suggest that informationobtained from a parent must be interpretedin light of the parent’s level of emotionaldistress. We also discussed the importanceof obtaining a family psychiatric history formaking differential diagnoses and treatmentrecommendations. In this section, weprovide a brief overview of basic researchshowing the link between parent and childadjustment difficulties that can aid theclinical assessor in structuring assessmentsand making appropriate interpretationsfrom the assessment information.Parental DepressionOne type of parental adjustment that has awell-documented link to child developmentis parental depression. Studies have foundthat between 40% (Orvaschel, Walsh-Allis,& Ye, 1988) and 74% (Hammen et al.,1987) of the children of depressed parentsexhibit significant adjustment problems.Depression in parents places children atrisk for a number of problems spanningacademic, social, emotional, and behavioraldomains (Downey & Coyne, 1990).Therefore, it seems that parental depressionis a nonspecific risk factor for problemsin children. That is, it is not specificallyrelated to the development of a single typeof child behavior problem.There are two possible exceptions tothis non-specific relationship, both relatedto subtypes within affective disorders.First, Weissman, Warner, Wickramaratne,& Prusoff (1988) found some preliminaryevidence that early-onset recurrent depression(depression that has its initial onsetbefore adulthood) might have a more specificlink with childhood depression. Second,family histories of bipolar disordersin parents predict a risk for subsequentbipolar disorder in adolescents with achildhood-onset of depressive symptoms(Geller et al., 1993).Goodman and Gotlib (1999) reviewedthe literature on the risk for problems inadjustment in children of depressed mothers.They outline four possible mechanismsto explain this risk: (a) an inherited predispositiontransmitted from parent to child;(b) failure of the child to develop appropriateemotional regulation strategies;(c) exposure to negative maternal moods,thoughts, and behaviors; and (d) exposureto a high rate of stressors associated withmother’s depression (e.g., higher rates ofmarital conflict). These links clearly illustratethe need to assess parental depressionto understand several potential causal factorsthat could help to explain a child’s oradolescent’s problems in adjustment.Parental Substance AbuseA comprehensive review of the literaturefound that, like parental depression, parentalalcoholism is associated with a numberof child adjustment problems. West andPrinz (1987) reported studies finding anassociation between parental alcohol abuseand the following problems in their children:hyperactivity, conduct problems,delinquency, substance abuse, intellectualimpairment, somatic problems, anxiety,depression, and social deficits. Like depression,some of the lack of specificity in itseffect on child adjustment may be due to afailure to define subgroups within parentswho abuse substances (Frick, 1993). Alternatively,West and Prinz reviewed severalstudies suggesting that the effects of havinga substance-abusing parent on a child’sadjustment may be mediated through the

294 CHAPTER 12 assessing family contextimpact on the home environment andthe impact on parent and child interactions.Consistent with this view, parentalsubstance use has been linked to a host ofproblematic parent practices, includinghigher rates of abuse (Ondersma, 2007;Walsh, MacMillan, & Jamieson, 2003).Parental Antisocial BehaviorThe intergenerational link to antisocialbehavior is a consistent finding in researchand one that has long intrigued social scientistsand policy-makers alike (see Frick& Loney, 2002, for a review). Early studiestended to focus on the intergenerationallink to criminality. This research foundthat the link was independent of socioeconomicstatus, neighborhood, and intelligence(Glueck & Glueck, 1968). Morerecent studies have focused on psychiatricdefinitions of antisocial disorders. As instudies of criminality, children diagnosedwith antisocial disorders are significantlymore likely to have parents with antisocialdisorders than are children without conductproblems (Frick et al., 1992; Monuteaux,Faraone, Gross, & Biederman, 2007).An important methodological pointin the more recent family history studieswas the fact that each used clinic controlgroups and found that histories of antisocialdisorders in parents were specificto conduct problems in children. That is,children with conduct problems not onlyhad higher rates of parental antisocial disorder(APD) than normal controls, butthey also had higher rates of parental APDthan clinic-referred children with otherproblems in adjustment (Frick et al., 1992).Therefore, unlike parental depression andsubstance abuse, parent antisocial behaviorappears to have a more specific relationshipto a particular child problem (i.e., conductproblems).Frick and Loney (2002) reviewed datasupporting several potential mechanismsto explain this link including an inheriteddisposition passed from parent to child,parental modeling of antisocial behaviors,and disruptions in the family caused bythe parent’s antisocial behavior. In supportof at least some inherited predisposition,Tapscott, Frick, Wootton, and Kruh (1996)showed that a paternal history of antisocialpersonality disorder was associated with ahigher rate of Conduct Disorder in theirbiological offspring, even if the father hadno contact with the child since the firstyear of life.Parental ADHDThere is evidence that parents and otherbiological relatives of children with ADHDshow more attentional problems (Albert-Corush, Firestone, & Goodman, 1986) anda higher rate of ADHD (Faraone, Biederman,Keenan, & Tsuang, 1991). However,these studies may have underestimatedthe link between parent and child ADHDby studying the parents’ current adjustment.Given that 30–50% of children withADHD may not be diagnosed with thisdisorder as adults (Barkley, Fischer, Smallish,& Fletcher, 2002), it may be that manyof the parents of ADHD children exhibitedADHD as a child but are not currentlyshowing symptoms.To test this possibility, Frick et al. (1991)studied the childhood histories of parents ofclinic-referred children. A child’s biologicalparent reported on whether or not he or shehad problems associated with ADHD beforethe age of 18 and then completed a similarfamily history questionnaire for all firstdegreerelatives. Children with ADHD weremore likely to have mothers, fathers, andother biological relatives who also exhibitedADHD as children than were other clinicreferredchildren. In fact, approximately75% of the 103 children with ADHD hadone biological relative with a significant historyof ADHD (27% of mothers and 44% offathers) and 46% had two biological relativeswith a significant history of ADHD. This

CHAPTER 12 assessing family context295study suggests that an assessment of parents’childhood histories of behavior problemscould aid in the assessment of ADHD inchildren.Parental AnxietyAnother type of problem that appears tohave a familial link is anxiety. Last, Hersen,Kazdin, Francis, and Grubb (1987) reportedthat in their sample of children with an anxietydisorder (n = 58), 83% of the childrenhad a mother with a lifetime history ofanxiety disorders. Furthermore, 57% hada mother experiencing significant levels ofanxiety concurrently with the child. Bothof these proportions were significantlygreater than what was found in parents ofclinically referred children without anxietydisorders.Importantly, Frick et al. (1994) foundsimilar results but also found that the linkbetween mother and child anxiety couldnot solely be attributed to anxious mothersreporting more anxiety in their children.All of the children in the Frick et al.study who self-reported an anxiety disorderhad a mother with a history of an anxietydisorder. Further, there are a number ofstudies suggesting that parental anxietycan influence the attachment between theparent and child (Costa & Weems, 2005)and can lead to parenting behaviors (e.g.,overprotectiveness, failure to encourageindependence) (Dadds, Barrett, Rapee, &Ryan, 1996) that can place a child at riskfor anxiety and other problems in adjustment.Thus, it is clear from these findingsthat parental anxiety is an important area tobe assessed in clinical assessments of anxiouschildren.Parental SchizophreniaAnother type of maladjustment with aclear familial link is schizophrenia. Childrenof one schizophrenic parent appear tohave a 10 to 15% likelihood of developingschizophrenia; the children of two schizophrenicparents have about a 25 to 46%risk (Gottesman, McGuffin, & Farmer,1987). These rates of disorder in offspringsof schizophrenic parents are striking giventhat the prevalence of schizophrenia in thegeneral population is between 1 and 10per 1,000 individuals (Helzer & Pryzbeck,1988). However, parents who have anotherrelative with schizophrenia are often quiteconcerned over the risk for their children,on the basis of the evidence for a familialtransmission. Therefore, it is often importantfor clinical assessors to also view theserisks from the point of view that the vastmajority of children with a schizophrenicrelative, even if that relative is a parent, donot develop schizophrenia.Assessment of Family HistoryFrom this very brief overview of the familiallink to childhood disorders it is evidentthat obtaining a family history is a criticalcomponent of most clinical assessments ofchildren and adolescents. However, likeall aspects of the assessment process, whatareas to be assessed and the depth at whichthey will be assessed depend on the individualcase. In some cases a screening forpsychiatric disorders in a child’s relativescan be conducted as part of an unstructuredinterview followed by a more indepthfamily history assessment only if thisis judged to be warranted from the initialscreening. In other cases a more detailedand structured assessment may be neededfrom the outset.It is beyond the scope of this book tocover assessment of adult psychopathologyin great detail. However, it is importantto note that one can assess a widerange of problems in parents or otherrelatives through omnibus rating scalesor structured interviews. For example, theNEO-Personality Inventory (NEO-PI;Costa & McCrae, 1985) and the Minnesota

296 CHAPTER 12 assessing family contextMultiphasic Personality Inventory-SecondEdition (MMPI-2; Hathaway & McKinley,1989) are two widely used and readily availableobjective personality inventories thatcover a number of areas of psychologicalfunctioning. There are also numerousstructured diagnostic interviews that areavailable, like the NIMH Diagnostic InterviewSchedule-Third Edition (DIS-IIIA:Helzer & Robins, 1988) and the StructuredInterview for DSM-IV (SCID; First,Spitzer, Gibbon, & Williams, 1995).When considering what type of assessmentmay be needed, it is important tonote that structured interviews tend tofocus on more severe pathology, assessingfor diagnosable disorders, than the objectivepersonality inventories do. In addition,the structured interviews tend to be moreamenable to the family history methodof assessment, in which a family memberreports on him- or herself and other relativeswho cannot be assessed directly.There may be some assessments whena more focused family history is deemedappropriate, such as when one wants tofocus on some specific domain of parentaladjustment. For example, the Beck DepressionInventory (BDI; Beck, Steer, & Garbin,1988) and the State-Trait Anxiety Inventory(Spielberger, Gorsuch, & Lushene, 1970)are brief screening measures for depressionand anxiety, respectively that are oftenused to assess parental adjustment in clinicreferredchildren. These are just a few ofa host of domain-specific rating scalesthat can be used to assess a specific areaof adjustment in a child’s parent or otherrelatives (see McMahon & Frick, 2007 forothers).ConclusionsIn this chapter we discussed several reasonswhy the assessment of the family environmentis critical to most clinical assessmentsof children and adolescents. Family factorsoften play a critical causal role in child maladjustmentand familial factors can aid inmaking differential diagnoses and treatmentrecommendations, two important goals ofmany clinical assessments. Further, factorswithin the family are often important forinterpreting information provided by membersof the family on a child’s or adolescent’sadjustment.There are a large number of dimensionsof family functioning that can influencechild adjustment. We focused on fourdimensions that we think are especiallyimportant to assess in most clinical assessmentsof children and adolescents: parentingstyles and practices, parenting stress,marital conflict, and parental adjustment.For each of these dimensions, we reviewedthe research linking them to child adjustmentand then provided several methodsfor assessing them in clinical assessments.Chapter Summary1. There is no context more important tounderstanding a child’s emotional andbehavioral functioning than understandingthe child’s family context. This isbecause of the following factors:(a) Familial influences often play majorcausal roles in a child’s or adolescent’spsychological difficulties.(b) A family psychiatric history can beinstrumental in making differentialdiagnoses.(c) Understanding a child’s or adolescent’sfamily context can help tointerpret information provided bymembers of the family.(d) Understanding a child’s family contextcan help to determine the mostimportant targets for intervention.2. Many behavior rating scales reviewedin previous chapters, such as the PIC-2,the PIY, the MMPI-A, and the BASC-

CHAPTER 12 assessing family context2972-SRP, have scales that assess variousaspects of a child’s family context.3. Research suggests that two criticaldimensions of family functioningrelated to child adjustment are parentingstyle, which is the emotional climateprovided by the parents, and parentingpractices, which are techniques used byparents to socialize their children andenforce rules.4. The Family Environment Scale-Second Edition (FES) is a commonlyused measure of parenting style andthe emotional climate of the family.The FES is divided into subscales fromthree domains: Relationships, PersonalGrowth, and Systems Maintenance.5. The Alabama Parenting Questionnaire(APQ) and the Parenting Scale (PS)are measures that focus more specificallyon parenting behaviors, such asparents’ discipline strategies.6. The Dyadic Parent Child InteractionCoding System (DPICS) is an observationalcoding system designed toassess parent and child behaviors in twostandardized parent–child interactiontasks.7. The amount of stress experienced byparents is also important to child adjustment.Scales can assess general stressorsor stressors specific to parenting.8. The Child Abuse Potential Inventory-Second Edition (CAPI) was developedto assess parents’ risk for abusing theirchildren and includes several scalesrelated to family stress.9. The Parenting Stress Index-SecondEdition (PSI) focuses specifically onstressors related to parenting.10. The level of marital discord andovert marital conflict in the home hasproven to be important in understandingchildren’s functioning.11. Assessing parental adjustment canprovide critical information in clinicalassessments, especially family historiesof depression, anxiety, antisocial behavior,attention deficit disorder, substanceabuse, and schizophrenia.

C h a p t e r 1 3History TakingChapter Questionsl What are the unique contributions of historicalinformation to the child assessmentprocess?l What are the typical domains, variables,or behaviors, assessed by such strategies?l What structured and unstructured historytaking methods are available?l How should clinicians go about collectingcomprehensive historical informationefficiently?The Role of History Taking inChild AssessmentHistory taking, often through a clinicalinterview is central to the purpose ofchild assessment. Indeed, it is perhaps theessential component of child psychologicalassessment, as a good history enables theclinician to conceptualize a case by providinginformation about the developmentalcourse of the child’s difficulties, the specificpresentation of the individual child’sdifficulties, risk and protective factors, andthe important contextual influences on thechild’s functioning. Such factors are notroutinely assessed by rating scales, selfreportinventories, or other widely usedmeasures. Indeed, it is impossible to conceiveof a competent assessment that wouldnot include history taking in some form.Although it is at least as crucial as othertypes of information for child assessment,historical information is often underutilizedby clinical assessors. This oversight isespecially regrettable given the widespreadavailability of history-taking measures orguides and the ease with which they can beused. It appears that history taking is simplynot as well entrenched in child behavioral299P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_13, © Springer Science+Business Media, LLC 2010

300 chapter 13 history takingassessment, as it is in medical assessment,where history taking (interviewing) is centralto making initial diagnostic decisions.Instead, many clinicians may focus on thecurrent presentation of symptoms or problems,as well as behavioral observations(which is important information), to theexclusion of gathering information aboutthe child’s developmental history.On the other hand, clinicians may centeran entire assessment around history takingthrough an unstructured clinical interviewwhich is necessarily idiosyncratic by childand clinician. Such approaches to assessmentfail to gather specific information onthe symptomatology and the child’s functioningrelative to developmental normsthrough norm-referenced assessment.Therefore, as with every other methoddiscussed in this text, history taking should beconsidered part, but not all, of a comprehensiveassessment. In short, we consider historytaking necessary but not sufficient. There arenumerous reasons for taking histories as partof any child assessment. A few of the variablesthat are uniquely assessed by history takinginclude age of onset, course/prognosis, etiology,previous treatment history, and otherrelevant background information.Age of OnsetMost rating scales and self-report inventoriesdo not directly assess the age ofonset of problems. Age of onset is particularlyimportant in child assessmentas it affects diagnosis directly. In the caseof mental retardation, for example, ageof onset must occur during the developmentalperiod (i.e., prior to age 18). Furthermore,the age of onset is crucial forthe diagnosis of ADHD, autistic disorder,learning disabilities, and many otherchildhood disorders (APA, 2000).Most psychologists have seen casesof suspected mental retardation with areported age of onset in the early 20s.Similarly, children are often referred forADHD with an age of onset after age 7.Sometimes, however, the age of onset isunclear or disputed, and the typically lessstructured format of history taking is idealfor clarifying the onset issue for a particularchild.As one example, a mother may bringin a college freshman because he is suddenlyhaving problems in school. On arating scale or other more structuredassessment format, the clinician may betempted to accept this current time as theage of onset. In the less structured history-takinginterview, one can pursue theissue of onset more thoroughly by askingthe mother if her son had ever receivedspecial education services, had a history ofdifficulty concentrating or finishing tasks,had been prescribed medication for attentionproblems, or had received tutoring inan early grade. The clinician may also askif there were any teacher complaints inearly elementary school concerning attentionproblems.It is not unusual for a clinician to find amuch earlier age of onset than that initiallystated by a child, teacher, parent, or otherinformant. If, in this sample case, the collegestudent did experience some difficultiesin the first grade that were associatedwith inattention and hyperactivity, andhe also struggled to pass both the ninthand twelfth grades, then the clinician mayinvestigate more carefully the diagnosisof ADHD with its associated early age ofonset. A more extended case study is providedin Box 13.1.Course/PrognosisThe course of a child’s difficulties refers tothe assessment of the developmental trajectoryof symptomatology. Differencesin the stability of symptoms can have substantialimplications for differential diagnosis.For example, if a parent describes achild as having had severe hyperactivityin first grade that was not apparent in second

chapter 13 history taking301Box 13.1Case Example of the Use of History TakingJames is a 15-year-old male who was referredby his father for suspected ADHD. Thereferral for evaluation was also supported byhis high school teachers.According to James’s father, his school difficultiesbecame pronounced about 3 years ago.He rarely does homework and is described byhis teachers as exhibiting hyperactivity, impulsivity,and inattention. He is also describedas the class clown by his teachers. A recentclassroom observation found that James wasgrossly inattentive in comparison to the otherchildren. He, for example, was the only onewho briefly rolled around on the floor in therear of the room during part of a teacher’s lecture.Previously, norm-referenced behaviorratings by four of his teachers identified himas highly hyperactive and inattentive.Being mindful of the late onset of James’symptoms, the clinician asked James’s fatherfor more detail about the onset of his problems.His father could not identify significantsymptomatology prior to age 12. It wasalso discovered that James’s parents wentthrough a contentious divorce concurrentlywith the onset of symptoms. It was furtherdisclosed that James recently has threatenedto attempt suicide on several occasions. Aninterview with James subsequently revealedsignificant evidence of depression with suicidalideation.In this case, history taking, by clarifyingthe age of onset, made a significant contributionto the conceptualization of James’s difficultiesand intervention design. Failure touse history taking to rule out ADHD in thiscase could have been disastrous, as poor historytaking could have resulted in perhaps noassessment of suicidality and in a failure totreat his depression.grade but then emerged again in the fourthgrade, the diagnosis of ADHD may be calledinto question. Such a pattern would signalthat contextual variables or other difficultiesmight be related to the emergence of symptoms.Similarly, if a child displays symptomsof autism that seem to spontaneously remitand reappear, then the diagnosis of autismbecomes questionable. Furthermore, a previoushistory of episodes of depression placean individual at significantly greater risk forfuture such episodes (Klein, Dougherty, &Olino, 2005). Thus, history taking is criticalfor understanding an individual’s risk fordepression and to plan for the most appropriateinterventions.Evaluation of course also provides evidenceof the effectiveness of previous interventionstrategies. A child’s, or family’s,failure to respond to previous interventionssuggests that different and/or moreintensive interventions may be warranted.The trajectory of behavioral excesses ordeficits is uniquely accessed via history taking.Subtleties of this nature are difficult toassess even with many of the structureddiagnostic interviews. However, clinicianswho are aware of the influence of courseon case conceptualization, treatment planning,and prognosis are more likely to beadept at gathering and integrating suchinformation into their assessments.EtiologyEtiology refers to the likely presumed causeof a child’s difficulties. The assessment ofetiology is crucial in that it has importantimplications for treatment. The discoverythat a child’s learning difficulties inmathematics did not begin until the ninthgrade may very well rule out a learningdisability, depending on other aspects of

302 chapter 13 history takingthe client’s history. If the clinician thendiscovers that the child was for the firsttime placed in the advanced section of amathematics class, then curriculum placementbecomes a potential etiologicalagent. A simple change in classes may bean effective intervention for this suspectedmathematics learning disability. Such aparsimonious intervention may not havebeen tried, however, without importanthistorical information that led the clinicianto a potential etiology.Knowledge of etiology, in general, isof potential importance for at least fourreasons (AAMR, 1992).1. The etiology may be associated withother problems (e.g., Down Syndrome)that may impair other aspects of functioning- physical, social, etc.2. The etiology may be amenable to treatment/intervention.3. Knowledge of etiology may lead to thedesign of prevention programs for certainetiologies.4. Etiologies may be useful for forminghomogeneous groups for research ortreatment.It is apparent that part of this assessmentshould include a history of other symptomatologyto assess for comorbidity, as wellas issues that could have actually precipitatedcurrent difficulties (e.g., a history ofdepression that now manifests partly asdifficulty concentrating).In addition, attempting to determine etiologyfor a specific client helps clarify both caseconceptualization to answer the referralquestion and the recommended interventionstargeting the presumed etiologies.After all, these activities are the primarypurpose of child assessment, and historytaking is valuable to these ends.Unfortunately, for most childhoodproblems the range of potential singleand multiple etiologies is often extensive.The potential etiologies associatedwith depression, for instance, are multitudinous.A non-exhaustive list of thesefactors is shown in Table 13.1. Moreover,multiple etiologies may be interactingto produce symptoms of a problemsuch as depression. It would not be unusual,for example, for a depressed child to beaffected by parental depression, poverty,and the death of a friend, all of which mayrequire simultaneous and/or coordinatedintervention.Commonly used history-taking formsmay not fully address the various etiologiesassociated with a problem or set of problems.Clinicians often have to use theirknowledge of child development, psychopathology,and other areas of psychologicalresearch to go beyond the standardquestions included on a history form (i.e.,branching) to rule out important high-frequencyetiologies.This process of branching from a historyform is difficult for even the mostsavvy clinician. For the majority of examiners,the most realistic option will be toassess history over the course of two ormore assessment sessions. A second sessioncould be as simple as a telephone callthat allows the clinician to rule out an etiologythat has been hypothesized basedon previous history, assessment data,or other information. As Green (1992)notes: “Time between sessions is oftenan important diagnostic and therapeutically” (p. 460), and taking such time wouldallow the clinician to be more accurate inhis/her case conceptualization. However,it is acknowledged that many practitionersare working in settings or undercircumstances in which assessments areexpected to be completed very quickly.Therefore, it is important that the clinicianbe highly knowledgeable aboutetiology at the outset of the assessment,approach the history taking interviewaccordingly, and still take the necessarytime to not hastily attempt to answer areferral question.

chapter 13 history taking303Table 13.1 Selected Etiologies of DepressionMedical disorders associated with depressionAutoimmune disorders: systemic lupus, rheumatoid disease, sarcoidosisCancers: head, pancreas, gastrointestinal, lung, renalCentral nervous system diseases: Parkinson’s disease, degenerative dementias, normal-pressurehydrocephalus, subarachnoid hemorrhages. Huntington’s disease, reversible dementias, focal lesions(nondominant), stroke, head traumaEndocrinopathies/metabolic disorders: hypothyroidism, Addison’s disease, Cushing’s disease, pituitarytremors, diabetes, hyperparathyroidism, porphyriaIntoxications: lead, mercury, thalliumOccult infections: genitourinary tract, liverViral infections: influenza, viral pneumonia, mononucleosis, hepatitisDrugs associated with major depression or dysthymiaAnticancer: vincristine, vinglastineAntihypertensives: reserpine, methyldopa, propanolol, guanethidine, hydralazine, clonidineAnti-infectives: cycloserineAnti-Parkinson agents: levodopa, arnantadine, carbidopaCorticosteroidsHormones: estrogens, progesteronePsychotropicsOther psychiatric disorders that may confound the diagnosis of depressionAnxiety disorders: panic disorders, phobic disordersObsessive-Compulsive DisorderSleep disorders: narcolepsy, sleep apneaPsychotic disordersEating Disorders: Anorexia Nervosa, bulimiaVague somatic complaintsDementiaSexual dysfunctionsFamily Psychiatric HistoryAs noted in the previous chapter, a specificetiological agent of increasingimportance is familial psychiatric history.Research has underscored the importanceof family psychiatric history as a risk factorfor similar problems in offspring andas a variable that may contraindicatesome treatment options (see Klein et al.,2005). School screening is an example ofthe application of family psychiatric historytaking. Prekindergarten screeningprograms often ask questions aboutchild behavior, and the inclusion of familyhistory information may trigger preventionor intervention efforts. If, for example,a father reports a history of separationanxiety as a young child, then follow-upcould be planned to monitor this father’soffspring upon entry into kindergarten.While some minor separation anxietysymptoms are common at this age, if achild with family resemblance for suchproblems displays some difficulties, thenearlier and more aggressive interventionmay be warranted. The importanceof family history is illustrated in the casestudy provided in Box 13.2.

304 chapter 13 history takingBox 13.2A Case Study Illustrating the Importance of Family HistoryBradford is a 10-year-old male who wasreferred for school difficulties. He has reportedlyalways had school problems, but theyhave worsened in the fifth grade. His teacherscomplain that he is unmotivated, moody,irritable, and oppositional.His father stated that he has had difficultygetting Bradford to complete homework. Thesehomework sessions usually turn into powerstruggles, and Bradford ends up crying. He alsoreportedly seems unusually emotional in comparisonto his younger brother. According to hismother, he angers quickly and cries easily.Bradford’s developmental history is unremarkable,with the exception of the continuingproblems with work completion. Recently,however, he has had problems interacting withhis peers. He cannot identify a best friend, andhe is less frequently invited over to other children’shomes, according to his mother.Assessment results did not reveal any problemsof note. His intelligence and academicachievement scores ranged from average toabove average, with his scores in math in theLow Average range. From his teachers’ report,it appears that Bradford’s difficulty with completingwork has contributed to his fallingbehind somewhat in math.Parent and teacher rating scales did not produceany significant T-scores. His self-reportscores on internalizing scales such as anxiety anddepression were slightly elevated (in the 60s).Bradford’s developmental history and testresults make his case conceptualization difficult.The results suggest that he has internalizingproblems, but he does not display symptomatologythat approaches the severity necessary tomeet diagnostic criteria for depression, dysthymia,or anxiety disorders. The clinician beganto question the parents’ reason for the referral.In particular, it was difficult to discern why theparents were so concerned about what appearsto be a circumscribed problem with homeworkcompletion and achievement motivation. Theclinician discovered the real reason for the parents’referral at the feedback session.After the clinician presented the findings,recommended behavioral intervention forhomework incompletion, and did not makea diagnosis, Bradford’s father (who had notattended the previous assessment session)disclosed his concern about Bradford’s internalizingsymptoms, saying that Bradford hadinherited his father’s susceptibility to depression.Bradford’s father then disclosed that hehad not wanted the clinician to know about hisprevious and ongoing treatment, so as to notbias the clinician. Bradford’s dad then revealedthat he has suffered from depression since childhoodand has found it difficult to control. Hehas reportedly been involved in psychotherapyat various times, including currently. He hasalso received pharmacological treatment fordepression and is currently taking Prozac. Hereported that he has been hospitalized previouslybecause of suicidal ideation. He also saidthat although he is a successful businessman, hehas difficulty functioning without ongoing psychotherapyand medical management. He alsorevealed that his father suffered from depressionand eventually committed suicide.These belated but important historicalfindings certainly affected the clinician’s conceptualizationof the case and interventionplanning. The clinician changed his recommendationsto include psychotherapy as afurther assessment tool and as a means fordesigning preventive interventions.One of the methodological problemsassociated with family health and psychiatrichistory taking is that of underreportingof illness. More reliable information may beobtained by directly interviewing each familymember, a method known as family study(Rende & Weissman, 1999). Numerous interviewmethods exist for conducting a familystudy of adults including the DiagnosticInterview for Genetic Studies (DIGS; Nurnbergeret al., 1994) and the Family HistoryScreen (Weissman et al., 2000). Apart froma specific focus on family psychiatric history,some methods exist that may help a clinicianorganize information about the child’s familystructure and history. A brief discussion ofgenograms as one such method is presentedin Box 13.3.

Box 13.3Genograms, not unlike profile plots on measuressuch as the MMPI-A or intelligencetests, present information graphically in amanner that can be quickly interpreted by thetrained professional. In its most widely usedform, the genogram is essentially a family treethat allows the professional to document theclient’s family structure (i.e., parents, siblings,grandparents, etc.), the relationships amongfamily members, critical events, as well as anyparticular variables of interest (e.g., psychiatrichistory, genetic history).Several volumes on genograms exist (e.g.,McGoldrick, Gerson, & Shellenberger, 2008).The uses of genograms vary widely includingas a means for obtaining medical history (seePapadopoulos, Bor, & Stanion, 1997), forunderstanding an individual’s cultural context(Shellenberger et al., 2007), and to aid inclinical supervision (Aten, Madson, & Kruse,2008). For clinical assessment, Papadopouloset al. (1997) emphasize that genograms arenot a quantitative technique with an objectiveinterpretative strategy. Instead, genogramsprovide a visual representation of family historythat can be used, in conjunction with305other information, to allow the clinician todevelop hypotheses about the case.In addition to gathering information toaid in case conceptualization and answeringa referral question, genograms may also beuseful to convey etiological theories to theclient/parent about the presenting problems(Papadopoulos et al., 1997). For example,visually depicting an extensive family historyof depression (or depicting protective factorsin the family context) may help parents reconceptualizetheir child’s problems.Of course, to take advantage of the presumedbenefits of genograms, including efficiencyof information gathering, one must bewell-trained at constructing and interpretinggenograms and must still conduct a clinicalinterview in a manner that will obtain thedesired information. Genograms are not amenableto the same tests of reliability, validity, andutility of many of the other tools described inthis text. Therefore, the decision of whether ornot to routinely employ such a strategy comesdown to the clinician’s training, comfort, andperception of a genogram’s usefulness.A common set of symbols, and a very basicgenogram example is given below:

306 chapter 13 history takingPrevious Assessment/Treatment/InterventionA child’s history of previous intervention/treatment and assessment of these factors isprimarily available through history taking.Similarly, previous assessment results mayaffect interpretation of current findings.A clinician, for example, may encountera child who has received psychotherapyfor depression for 2 years, pharmacologicaltreatment for 3 months, and partialhospitalization for 6 months prior tothe current evaluation. If the child isdemonstrating symptomatology duringthe current evaluation that ismore severe than previously noted,the child’s clinician is more likely toadvise aggressive treatment, perhaps evenhospitalization. The need for aggressivetreatment, however, may not appear compellingwithout the knowledge of the previousintervention failures.Information about previous assessmentalso allows clinicians to gauge the accuracyof their current findings. A clinician canvalidate their current findings by comparingthem to the previous assessmentresults. A child, who was diagnosed withan anxiety disorder 2 months earlier, maydemonstrate significant anxiety as indicatedby current MMPI-A and RCMASresults. This congruence between currentand recent findings may lend some validityto the psychologist’s current results.A lack of congruence between the recentand current test findings, however, can beequally insightful. In this situation, an advisedfirst step would be to check the scoring of thecurrent measures. Previous diagnoses mayalso help guide further assessment. If a childhas been diagnosed previously with ADHD,then the examiner may wish to expand theassessment to other forms of disruptivebehavior, given the high comorbidity amongsuch problems (American Psychiatric Association,2000). Knowledge of the ADHDdiagnosis and its associated features andcomorbidities gives the clinician a specifichypothesis totest with further assessmentresults and a potentially valuable set of targetsfor intervention planning.Contextual FactorsAs noted above, clinicians should be preparedto branch off of standard historyquestions or forms to assess issues that areclient-specific and highly important to acomprehensive case conceptualization.For even a novice clinician, it is readilyapparent that contextual factors in a child’shistory may exert great influence on his/her present functioning. Such factorsinclude family relationships, divorce, moves,socioeconomic status, ethnicity, culture,neighborhood environment, and traumaticevents (see Dodge & Pettit, 2003). Forexample, it is one thing to determine that achild’s present academic difficulties are thecause of a parent’s concerns and the reasonfor referral. It is quite another to learn thatthere were no such difficulties for the childuntil after his/her parents divorced.A more complex contextual factor that mayplay a direct role in the design of interventionsis that of the goodness-of-fit betweenthe child’s characteristics and the contexts inwhich he/she is expected to function. In particular,contexts in which the child spendsmore time or that are more proximal likelywill exert a stronger positive or negativeinfluence (Cicchetti & Toth, 1998). If a childis prone to problems with anxiety and hisor her parents likewise openly worry aboutmany things and interact with the child inan emotional manner, the child’s anxiety islikely to be exacerbated. Being in the sameclass as another anxious student with whomthe child does not interact will not have aninfluence on the child’s anxiety. Treatmentrecommendations will likely include suggestionsfor the parents in managing their ownanxiety and/or in altering their communicationstrategies with the child.

chapter 13 history taking307These contextual factors should be routinelyincluded as part of history takingeven if there is no indication of adversecontextual factors at the outset of theassessment. The influence of context canalso be protective, and thus, these variablesmay also serve as strengths around whichinterventions can be built.ContentThere is considerable overlapping contentamong history measures althoughvarying degrees of emphasis exist. Someof the most common areas of inquiryinclude (adapted from the Departmentof Psychiatry and Child Psychiatry, TheInstitute of Psychiatry, and the MaudesleyHospital of London, 1987):ComplaintsPresent illnessFamily historyMotherFatherSiblingsOther relativesFamily atmospherePersonal historyEarly developmentBehavior and temperamentSchoolOccupationAdolescenceSexual historyMedical historyPrevious psychiatric problemsAlcohol, tobacco, and drug useAntisocial behaviorCurrent life situationDevelopmental delays or disabilitiesPersonalityAttitudes toward others in social, family,and sexual relationshipsAttitudes toward selfMoral and religious attitudes and standardsMoodLeisure activities and interestsFantasy lifeReaction pattern to stressA history that is of interest to a certainspecialty practice may have a differingemphasis. The content of a history thatmay be taken by a neurologist or neuropsychologistis described by Teodori (1993)as including the following topics:Birth historyDevelopmental historySchool historyIQ test resultsSocial historyNutritional historyHistory of exposure to chemicals or toxinsHistory of other family illnessesOther medical illnessesMedicationsHospitalizationsSurgeryReview of systems inquiring about allother aspects of the child’s physical functioningPrevious recordsFormatsHistory taking is typically conducted viatwo formats: an interview and a writtenform. While history taking research hasfocused on the interview, there are oftenoccasions where a written form may beused, particularly with the parents of children.Parents are already accustomed tocompleting such forms while waiting inphysicians’ offices. Increasingly, parentsare also accustomed to completing ratingscales in psychologists’ offices.Most history-taking forms, however,require considerable English language fluencyon the part of parents. Technical topics,such as the type of special education programthat their child attends or a previous medicalcondition, are difficult for even the mostsophisticated parents. Regardless, parents

308 chapter 13 history takingmay be able to complete part of the historyand thereby save the examiner interviewingtime. Assessment systems such as the BASC(Reynolds & Kamphaus, 2004) which was discussedin previous chapters on rating scales,also include forms for history taking (e.g.,BASC-2 Structured Developmental History;SDH). An example of a history form isshown in Fig. 13.1. Of course, history formswill vary greatly in their depth and breadth.

chapter 13 history taking309

310 chapter 13 history taking

chapter 13 history taking311

312 chapter 13 history takingThe important issue is to cover essential historyinformation as part of the assessmentand to ideally do so in an efficient manner.Forms, such as the one shown in Fig. 13.1,allow the clinician to gain such informationbut also require the clinician to follow upbased on the parent’s responses.Many clinicians, however, prefer a completelyinterview format because it allowsthem to use a branching procedure to pursuetopics that are of particular relevance toissues of onset, course, etiology, and previoustreatment history among other factorswithout a potentially cumbersome formthat includes some irrelevant informationfor a particular child. A sample scenario isa case where an 8-year-old child is referredbecause of a decline in academic achievement.The examiner may discover that thechild’s performance in school was exemplaryuntil the current academic year, whenthe child missed 35 days of school becauseof an automobile accident. The examiner, inthis case, will likely branch to ask questionsabout the child’s length of absence from theschool and the extent of documented headinjury. New and detailed information aboutthese problems may profoundly influenceintervention planning. Special educationmay be placed in abeyance in favor of remedialcoursework and monitoring to see if thechild recovers his premorbid level of functioningwith little intervention, as wouldoften be predicted. Hence, the interview orpartial interview format allows the clinicianthe flexibility necessary for thorough explorationof variables that may impact interventionrather than have the parent spendtime on a number of history questions thatare not central to the referral concern. Itmust be emphasized that such interviewsare, by nature, clinician- and client-specific,so psychometric properties, particularlyreliability, are not meaningful in evaluatinga clinician’s interview as part of theassessment process. Despite their lack ofamenability to empirical evaluations, asnoted above, interviews are indispensable

chapter 13 history taking313in psychological assessment. History formsmay be quite useful for efficiency andresearch purposes, but in clinical assessmentthey should be augmented by a directclinical interview.The face-to-face approach of the interviewformat may offer additional advantagesbeyond the possibility for branching.See, for example, Green’s (1992) advice onobserving nonverbal cues during historytaking given in Box 13.4.ConclusionsThis chapter puts forth the argument thathistory taking is a central, but often overlookedcomponent, of child assessment.Various reasons for making history takingmore central to the child assessmentprocess are offered with the focus on theunique contributions that history takingmakes to the process. Specifically, historytaking is ideal for assessing factors such asage of onset, etiology, course, and previousassessment/intervention outcomes.Psychologists are advised to selectappropriate history forms and practice thebranching procedures necessary for integratingpsychological science with childhistory. In fact, history taking provides anideal venue for the integration of psychopathology,child development, and cognitivescience literatures, among others,with child characteristics. Empirical evidenceshould shape the domains coveredin standard history-taking procedures;however, the process will, by nature, beclient-specific.Box 13.4Observing Nonverbal Cues During History TakingGreen (1992) provides some advice to pediatriciansregarding behaviors to observeduring the interviewing process. This compendiumis equally useful to the psychologistwho, while interviewing, is developinghypotheses about family dynamics and otherfactors that impact child development. Thefollowing is an adapted list of behaviors forwhich psychologists should observe duringhistory taking with a caregiver informant orset of informants.l Perspiration, blushing. or paling; controlled,uneven, or blocked speech; plaintivevoice; talking in a whisper; gait;posture; tics; affirmative nodding; negativeshaking of the headl Frequent swallowing tenseness, fidgeting,preoccupied air, avoidance of eye contact,social distancel A sudden glance at the interviewer orsomeone else in the family precipitated bya statement or questionl Clenching, rubbing, wringing hands,searching or nail bitingl Dress and personal groomingl Reddening of eyes or cryingl Frowns, smilesl Affect inappropriate to ideation (e.g., aparent smiles when talking about a child’ssevere behavior problems)l Interactions among parents, child, andexaminerl Developmentally inappropriate behavior(e.g., older child on parent’s lap)l The way in which the infant or young childis held or helped during the interviewl The parent’s ability to have child respondto a parent’s requestl Adopted from Green (1992), p. 455.

314 chapter 13 history takingChapter Summary1. Most rating scales and self-report inventoriesdo not directly assess the age ofonset of problems as well as do historytaking methods.2. The use of history taking to clarify ageof onset is crucial for the diagnosis andconceptualization of a number of childhooddisorders.3. Etiology refers to the likely presumedcause of a child’s difficulties.4. Knowledge of etiology is of potentialimportance for at least four reasons:(a) The etiology may be associatedwith other problems (e.g., DownSyndrome) that may impair otheraspects of functioning - physical,social, or other domains.(b) The etiology may be amenable totreatment/intervention.(c) Knowledge of etiology may lead tothe design of prevention programs.(d) Etiologies may be useful for forminghomogeneous groups for research oradministrative purposes.5. Research indicates a number of importantareas to cover in child and adolescenthistory taking. Clinicians shouldbe aware of these areas and use the“branching” technique to appropriatelyfollow up on any areas of concern for theclient.

C h a p t e r 1 4Adaptive Behavior ScalesChapter Questionsl What is meant by the term adaptivebehavior?l What are the typical domains of adaptivebehavior scales?l How is adaptive behavior different fromother constructs such as personality orintelligence?l What are some of the most popularadaptive behavior scales?History of the ConstructThe adaptive behavior construct traces itsroots to early work in mental retardation,which, in turn, is linked to the roots ofintellectual assessment (Kamphaus, 2001).Although intelligence tests contributedmightily to the recognition of the mentalretardation syndrome, Doll (1940) notedthat intelligence measures lacked sufficientbreadth for assessing all of the relevantdomains of behavior that needed tobe considered in treatment of individualswith mental retardation.Doll drew on his experience as a psychologistat the Vineland State TrainingSchool in New Jersey, where he wascharged with the assessment and rehabilitationof individuals with mental retardation.He noted that such individuals notonly lacked intellectual abilities necessaryfor academic attainment, but they alsooften appeared to lack day-to-day livingskills needed for independent functioning.In Doll’s terminology, they lacked socialmaturity.In order to intervene and improvea child’s social maturity, Doll created ascale to assess specific behaviors that were315P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_14, © Springer Science+Business Media, LLC 2010

316 chapter 14 Adaptive Behavior Scalesdeemed necessary for successful living.The Vineland Social Maturity Scale (Doll,1953) was the first of its genre. It includedseveral sets of items that assessed variousaspects of social maturity including locomotion,social skills, and grooming. TheVineland yielded a score that was roughlyparallel to a composite score offered bymany intelligence tests of the day. His totalscore was dubbed a “Social Quotient.” TheVineland became the premier measure ofadaptive behavior up to the present day,and Doll’s pioneering work became thebasis for all subsequent measures.The AAIDD CriteriaThe American Association on Intellectualand Developmental Disabilities (AAIDD),formerly the AAMR, bases its criteria for anintellectual disability, or mental retardation,on deficits in both intellectual functioningand adaptive behavior (see Luckassonet al., 2002), similar to previous criteria(e.g., AAMR, 1992). The current criteriathat indicate that three broad areas of adaptivebehavior should be assessed list tendomains of adaptive behavior that should beassessed for the purposes of mental retardationdiagnosis and intervention planning.These domains are: conceptual skills (e.g.,language, reading, writing, money skills),social skills (e.g., interpersonal, responsibility,avoiding victimization, obeying rules),and practical skills (e.g., maintaining a safeenvironment, self-care such as dressing, eatingmeals, hygiene).It should be noted that competence inthese areas of function may provide a basisof intervention for children with otherproblems such as autism, ADHD, learningproblems, or anxiety. In fact, the assessmentof adaptive behavior has been referred to asan essential part of assessments for autism(Ozonoff, Goodlin-Jones, & Solomon,2005). Adaptive functioning has also beendescribed as one factor that can help differentiatemental retardation from learningdisabilities, in that the former would bemarked by broad deficits in adaptive functioning,and the latter would be marked bymore specific deficits (Fletcher, Francis,Morris, & Lyon, 2005). The acquisitionand demonstration of adaptive skills havealso been identified as an important bufferagainst the development of psychopathology.For example, social competencies mayallow a child to resist being overwhelmedby negative life stressors (Tanaka & Westerman,1988). In this context, adaptive behaviorscales may serve an important functionin the assessment of all children referred forevaluation, not just those who are suspectedof having an intellectual disabilityDefining Adaptive BehaviorAdaptive behavior has been defined as “theperformance of the daily activities that arerequired for social and personal sufficiency”(Sparrow, Cicchetti, & Balla, 2005, p. 6).Hence, adaptive behavior is the antithesis ofmost of the behavioral constructs discussedin this book in that it deals with behavioralcompetencies and their absence as opposedto assessing behavioral problems.Most definitions of adaptive behaviorhave some core similarities including thepremise that it is an age-related construct.Specifically, adaptive behavior increaseswith age in the absence of interfering circumstances,much as academic achievementsaccrue over time. Adaptive behaviorsare identified by the standards of othersand the social context in which the childfunctions (DeStefano & Thompson, 1990).Finally, adaptive behavior assessmentfocuses on typical behavior as opposed toability. This assessment emphasis is alsoconsistent with the assessment of otherdevelopmental accomplishments such asacademic achievements.

chapter 14 Adaptive Behavior Scales317Relationship to IntelligenceThe choice of informant is apparently amoderator variable affecting the correlationof measures of adaptive behavior andintelligence (Kamphaus, 2001).Correlations between intelligence andadaptive behavior measures are modest,.40 to .60 (DeStefano & Thompson,1990) indicating some overlap but independence.This modest relation, however,is an oversimplification. The correspondencebetween the two constructs hasbeen found to be higher in individualswith pervasive developmental disorderswho also have deficits in intelligence thanin individuals with pervasive developmentaldisorders who do not have intellectualdeficits (Bolte & Poustka, 2002). In bothgroups, adaptive behavior across domainswas below average. In addition, moderatecorrelations between adaptive behaviorand intelligence have been found for juvenileoffenders (Hayes, 2005). The informantand domain of adaptive behaviorassessed also seem to affect the relation.Specifically, when teachers are used asinformants, the correlation between intelligenceand adaptive behavior increases,and domains that assess communicationand functional academic skills tend to correlatehigher with intelligence test results(Kamphaus, 1987).Hence, if one sees a stronger relationshipbetween intelligence and adaptive behavioras desirable, an emphasis on the assessmentof functional academics rated by teacherswill provide adaptive behavior results ofdesired value. If, however, one views theadaptive behavior construct as separate orcomplementary to intelligence assessment,parent ratings of less academically relatedskills (e.g., socialization) should be sought.In essence, where possible, an evaluatorshould seek information from both parentsand teachers in a variety of adaptive behaviordomains.Otherwise, Harrison (1990), who hasmade numerous contributions to the adaptivebehavior assessment literature, favors the useof parents as informants. She observes:“The third-party method of administration isparticularly appropriate for the assessment ofadaptive behavior. Because adaptive behavioris generally conceptualized as the daily activitiesin which a per-son engages to take careof himself or herself and get along with otherpeople, the information sup-plied by a thirdparty will be more valid than the direct administrationof tasks. The third-party methodalso allows for the assessment of individualswho cannot participate in the administration ofmany tests, such as the severely handicappedand young children” (pp. 472–473).The modest to small relation of adaptivebehavior to intelligence is punctuated in astudy by Szatmari et al. (1993). This studyfollowed 129 children of extremely lowbirthweight (501–1,000 g) to the ages of 7or 8 years and compared their performanceto that of a control group. They found significantdecrements in intelligence for thelow birthweight group but no significantdeficits for adaptive behavior. Such resultsindicate that different mechanisms affectthe development of intelligence and adaptivebehavior skills.While the low to moderate relationbetween intelligence and adaptive behaviorscales is well documented (Kamphaus,2001), the relation between adaptive behaviorand academic achievement is of considerableimportance given that academicachievement is often the criterion variableof interest to child clinicians. di Sibio(1993) examined the relation of adaptivebehavior to achievement after the varianceattributable to intelligence was removed.While intelligence scores were morehighly correlated with achievement, themeasure of adaptive behavior in that studyadded 11% of the variance to the prediction

318 chapter 14 Adaptive Behavior Scalesof achievement beyond that predicted byintelligence. The results of this study supportthe importance of assessing adaptive behaviorand intervening in order to enhance achild’s behavioral competencies which, inturn, provide another approach to enhancingacademic achievement.Uses of Adaptive BehaviorScalesTraditionally, adaptive behavior scaleshave been considered central to the diagnosisof mental retardation (DeStefano& Thompson, 1990). Although this useis circumscribed, it is an important onegiven the epidemiology of mental retardationand the importance of accuratediagnosis that includes a comprehensiveassessment of adaptive functioning.It has also been argued that assessmentof adaptive behavior has been includedin order to reduce the number of falsepositive diagnoses of mental retardationoccurring based solely on IQ scores (seeGreenspan, 2006).Perhaps as importantly, as the counterpointof behavior problem scales, adaptivebehavior scales hold unique potential forintervention design based on assessmentresults. Adaptive behavior scales measurekey skills that contribute to a child’s successfulfunctioning in a variety of environments.That is, adaptive behaviorscales serve a valuable function for theclinician in that they pinpoint specificskills that a child has not acquired, whichthen may serve as the focus of treatmentefforts.Adaptive behavior scales are particularlyuseful in educational settings where theirresults can be integrated with the objectivesof individualized educational plans(IEPs). For instance, adaptive behaviorscales can be used to identify social skillsand other target behaviors for classroomintervention planning.Characteristics of AdaptiveBehavior ScalesDomains AssessedAdaptive behavior scales are analogous tomeasures of behavior problems in that thedomains assessed vary somewhat from testto test. In Table 14.1 we illustrate the contentfor three measures of adaptive behavior.Domains assessing various aspectsof independent functioning/daily living,social skills, and communication skills arecommon to many tests. These domainsare reflective of those in the intellectualdisability criteria described above. Thedomains of behavior that are assessed bya particular test are influenced by its agerange. Children’s tests, for example, mayplace less emphasis on occupational skills,independent living, and interpersonal relationshipsthan measures that are moreconcerned with assessing adult adaptation.Measures of adult functioning requiredomains aimed at assessing occupationalskills, whereas the child’s analog is schoolfunctioning. The domains of an adaptivebehavior scale, thus, become an importantconsideration in test selection. A client’sage, the institution’s treatment program,and other factors may also influence testselection.In a 1993 study, Widaman, Stacy, andBorthwick-Duffy applied multitrait/multimethodmatrix procedures to the identificationof major domains of adaptive behavior.The participants for this study were 157persons with moderate, severe, and profoundmental retardation. The authorsfound clear evidence for the existence offour major domains: cognitive competence,social competence, social maladaptation,and personal maladaptation. These resultssuggest that, when assessing individualswith mental retardation, measures of atleast these four domains would be desirable.Fortunately, the majority of adaptive

chapter 14 Adaptive Behavior Scales319Table 14.1 Overview of Three Measures of Adaptive BehaviorTest Name Age Range Scales/Domains Administration TimeVineland AdaptiveBehavior Scales,2nd editionBirth through 90 Communication (receptive, expressive, written); dailyParent/caretaker and classroom living skills (personal, domestic, community); socialization(interpersonal relationships, play and leisureRating forms; survey andexpanded interview formstime, coping skills); motor (gross, fine); maladaptivebehavior (internalizing, externalizing, other)20–60 minScales of IndependentBehavior: RevisedAdaptive BehaviorAssessment System-IIInfancy through adulthood Motor skills (gross motor, fine motor); social interactionand communication skills (social interaction, languagecomprehension, language personal expression);personal living skills (eating and meal preparation,toileting, dressing, personal self-care, domestic skills);community living skills (time and punctuality, moneyand value, work skills, home/community orientation)Birth to 89 years Communication, community use, functional academics,Separate forms for 0–5; 5–21;home living, health and safety, leisure, self-care,and 16–89self-direction, social, workBroad independence scale,45–60 min; short form scale,10–15 min; early developmentscale, 10–15 min15–20 min to administer

320 chapter 14 Adaptive Behavior Scalesbehavior scales measure these domains and,in many cases, additional areas as well.Norm vs. Criterion ReferencingPrior to the publication of the revisedVineland in 1984, there were no nationallynormed adaptive behavior scales. Severalscales possessed local or regional norms,and many were created locally and interpretedinformally. Unfortunately, many ofthese scales were used for making normreferenceddecisions such as determiningwhether or not a child had adaptive behaviordeficits that were significant enough to warrantthe diagnosis of mental retardation.Among the adaptive behavior assessmentquestions most frequently posed bypsychologists are the following:1. Does the child have adaptive behaviordeficits that are significant enough towarrant the diagnosis of mental retardation?2. What are the adaptive behavior deficitsthat most influence the child’s adjustmentand therefore, require intervention?3. What are the adaptive behavior strengthsdisplayed by the child?Questions 2 and 3 are less likely torequire norm referencing, althoughit may still be of some benefit. Inorder to answer these questions, theclinician could make intraindividual comparisonsand/or gauge deficits on the basisof how the deficits impair adaptation toparticular environments (e.g., failure tofollow rules in games).Choice of InformantThe Vineland Social Maturity Scale representsone of the first scales in psychologyto place a premium on parents as informants.This approach stood in stark contrastto the popularity at the time of usingdirect measures of child behavior, such asintelligence tests.Modern adaptive behavior scales suchas the Vineland-2 (Sparrow, Cicchetti, &Balla, 2005) and Scales of IndependentBehavior-Revised (Bruininks, Woodcock,Weatherman, & Hill, 1996) still emphasizethe use of parents as informants. Formost scales in which parents serve as informants,maternal reports are used nearlyexclusively for item development, scaling,and norming. Although fathers are usedless frequently, there are no systematicdata currently available that clarify the differencesbetween mothers and fathers asinformants regarding adaptive behavior.Secondarily, teachers often serve as ratersof adaptive behavior. However, teachershave different views of a child’s adaptivebehavior because of the varied demands ofschool and home settings. Domains thatare commonly included on parent scalesof adaptive behavior, such as toileting,bathing, dressing, budgeting, and healthcare are impractical for teachers. Likewise,parents have difficulty reporting onfunctional academics in reading, writing,calculation, and some vocational skills.The differing demands of school andhome environments virtually ensure thata clinician will have an incomplete understandingof a child’s adaptive behavior ifeither teacher or parent adaptive behaviorratings are not used.Other caregivers also serve as importantinformants regarding a child’s adaptivebehavior. Caregivers may include psychiatricaides in hospitals, nurses, mental healthassistants, nannies, grandparents, teacheraides, work supervisors, or others whohave sustained nearly daily contact with achild or adolescent. These individuals may,in some circumstances, fulfill parent orteacher roles and therefore may be competentprimary or secondary informants forsuch scales. Even when parents or teachers

chapter 14 Adaptive Behavior Scales321complete an adaptive behavior scale, othercaregiver information may be of value. Anadolescent’s work supervisor is one exampleof an informant who may contributeunique and important information to thepool of assessment data gathered on adaptivefunctioning.Finally, the child as an informant shouldnot be overlooked. In other words, thechild may be tested directly in order toassess adaptive behavior, although this is aless popular option with many clinicians.Perhaps one reason for its lack of popularityis a dearth of available instruments. TheChildren’s Adaptive Behavior Scale (CABS;Richmond & Kicklighter, 1979) is one ofthe few direct measures of adaptive behavioravailable. The CABS uses an individualtesting format to assess Language Development,Independent Functioning, FamilyRole Performance, Economic/VocationalActivity, and Socialization. Obviously,these domains are common to many typesof adaptive behavior measures.As is the case with behavior rating scales,there is often a high level of disagreementbetween the informants who give informationor adaptive behavior scales. Parentsand teachers have been found to generallyagree on overall adaptive behavior estimates(e.g., Harrison & Oakland, 2003).However, they may disagree considerablyat the scale or domain levels (Shaw, Hammer,& Leland, 1991). Self- and caregiverreports have been found to sometimesagree on adaptive behavior ratings but todisagree considerably on ratings of maladaptivebehavior (Voelker et al., 1990).Factors such as child age and length ofinvolvement in a treatment program mayalso affect agreement between informants(see Shaw et al., 1991).Administration FormatAdaptive behavior scales are commonlyadministered using a checklist or semistructuredinterview technique, except, ofcourse, in the case of direct assessment ofthe child. The checklist format is equivalentto the approach used for parent orteacher ratings of behavior problems discussedin previous chapters. The semistructuredinterview method espoused byDoll (1953) and popularized by Sparrow,Balla, and Cicchetti (1984) differs substantiallyfrom rating scale methods.The semi-structured interview technique,however, requires a high level ofclinical skill. The clinician has to make theinterview conversation-like, topical, andempathic, while at the same time collectingthe necessary information to allow foraccurate rating (scoring) of individual items.In addition, younger children could havea difficult time articulating their skills invarious adaptive behavior domains. Therefore,when using self-report procedures foradaptive behavior, a semi-structured interviewformat is not recommended. Therating scale method is more time-efficientand practical in that the clinician does noteven need to be present for the administrationof the scale.However, the semi-structured interviewtechnique has many virtues, including thefollowing:l Allowing the examiner to clarify questionsfor the informant by providing examples,and so forthl Contributing to the establishment ofrapport between clinician and parentbecause of the conversation-like natureof the interactionl Mitigating against response sets such asfake good or fake badl Permitting the assessment of adaptivebehavior, despite poor English-languagereading skillsThe semi-structured interview techniquecan be easily mastered with practice. Sometechniques for mastering the techniquefollow.

322 chapter 14 Adaptive Behavior Scales1. Begin by asking general questions andthen proceed to the specific informationneeded to score items. If one startedwith the Communication domain of theVineland-2, for example, a good startingquestion for a parent of a toddler mightbe something like, “Tell me about someof the things that Tom is saying thesedays.”2. Ask for examples of day-to-day behaviorbecause adaptive behavior scales aredesigned to assess typical behavior ratherthan ability (see Sparrow et al., 2005). Afollow-up expressive language questionmight be, “Tell me the words that youcan remember Tom saying today.”3. Become very familiar with the interviewitems and scoring criteria for specificitems in order to ensure that adequateclarification is sought.4. Conduct the interview topically. Forexample, ask all of the items regardingtelephone skills (answering appropriately,states telephone number, uses payphone, etc.) before proceeding to thenext topic.5. Pursue questioning until you have aclear picture of the child’s day-to-daybehavior. Once you have achieved thisportrait, you can confidently rate thechild’s behavior on individual items.Omnibus Adaptive BehaviorScalesVineland Adaptive BehaviorScales, 2nd edition (TheVineland-2; Sparrow, Cicchetti,& Balla, 2005)The Vineland-2 (Sparrow et al., 2005) isthe latest version of an assessment toolfor adaptive behavior that traces its rootsto the originator of the adaptive behaviorconstruct, Edgar Doll. Doll created thefirst widely used scale of adaptive behavior,the Vineland Social Maturity Scale. TheVineland-2 and its predecessor (Sparrow,Balla, & Cicchetti, 1984) represent a substantialrevision, adaptation, and extensionof Doll’s original scale.The Vineland-2 consists of a family ofscales each of which possesses characteristicsthat make it well-suited for particularpurposes. The Vineland componentsinclude the following: The Teacher RatingForm assesses adaptive functioning in theclassroom for children ages 3½ through 18.The Survey Interview Form is administeredto parent/caretakers of individual frombirth to age 90 in semi-structured interviewformat (an Expanded Interview Form isalso available). New to the Vineland-2 is aParent/Caregiver Rating Form that allowsparents to rate adaptive behavior items in arating scale format.The Vineland-2 has many uses in additionto its popularity as a tool in assessmentsand diagnoses of mental retardation.The Vineland-2 provides a rather comprehensiveassessment of an individual’sbehavioral competencies and can be usedto assess treatment progress, as well as todetermine treatment goals.ContentAside from re-standardization, the mostsubstantial change in the Vineland-2 fromthe previous Vineland is in the expansionof the number of items in each domainand the inclusion of the Parent/CaretakerRating Form (Sparrow et al., 2005). TheVineland-2 includes the same domains asits predecessor: Communication, DailyLiving Skills, Socialization, Motor Skills,and Maladaptive Behavior. The MotorSkills domain is designed for ages frombirth through 6 years and for older individualswith motor handicaps. The MaladaptiveBehavior domain is essentially abehavior problems checklist that assesses

chapter 14 Adaptive Behavior Scales323very severe difficulties such as problems inpublic, sexual misbehavior, self-injuriousbehavior, bedwetting, and truancy. Each ofthe domains consists of following content:CommunicationReceptiveExpressiveWrittenDaily Living SkillsPersonalDomesticCommunitySocializationInterpersonal RelationshipsPlay and Leisure TimeCoping SkillsMotor SkillsGrossFineMaladaptive BehaviorInternalizingExternalizingOther (e.g., wets bed)Administration and ScoringThe Vineland Survey form uses Doll’s wellknownsemi-structured interview technique,a method that is often not followedloyally by Vineland users. The techniquehas many advantages; unfortunately, its disadvantagesare more salient. Said simply,mastering the semi- structured interviewtechnique is not easy.A central problem with the method isthe necessity of organizing the interviewtopically, while the items are placed onthe response form by difficulty order. Forexample, several items of the Daily LivingSkills domain of the Survey Form haveto do with telephone skills – answering,dialing, and so on. The semi-structuredinterview technique involves obtainingadequate information to score these items,even though they are scattered throughoutthe record form. This central contradictionof simultaneously interviewing topicallyas well as having items ordered byincreasing difficulty undoubtedly leads toroutine usage of the Vineland as a ratingscale despite the fact that this is a violationof standardized procedure. Two aspectsof the Vineland-2 speak to the tendencyfor clinicians to have difficulty with thesemi-structured interview approach: (a)icons are now shown on the record formsuch that interviewers can more easily spotitems that go together (e.g., a book iconsignals items that have to do with readingskills); and (b) the Parent/Caretaker RatingForm and the Teacher Rating Formnow make it possible for the clinician toobtain Vineland ratings without the use ofan interview. Nevertheless, given the manyadvantages of the semi-structured interviewtechnique cited earlier, it behoovesclinicians to acquire this unique skill.Vineland interpretation features standardscores based on a mean of 100 andstandard deviation of 15 for the AdaptiveBehavior Composite and for each of theDomain scores. A new feature is the useof “v-scale” scores for subdomains. Thesescores have a mean of 15 and standarddeviation of 3.NormingThe Survey Form norming sample of theVineland was collected on the basis of2001 US Census Bureau statistics (CurrentPopulation Survey, 2001). The samplewas made up of 3,687 subjects from birththrough 90 years. The sample appears tobe representative of the larger populationin terms of ethnicity, SES, geographicregion, and disability status. Clinical samplesof individuals with ADHD, autism(both with and without language), hearingimpairment, and “emotional or behavioraldisturbance” were included in the standardizationand subsequent studies (Sparrowet al., 2005, p. 91).

324 chapter 14 Adaptive Behavior ScalesReliabilityReliability analyses indicate good internalconsistency for the Adaptive BehaviorComposite and the domain scores, withcoefficients generally .90 and higher forthe former and .80 and higher for thelatter. Subdomain internal consistencycoefficients likewise tended to be .80 andhigher, with the exception of the Receptive(communication), Personal (dailyliving skills), and Play and Leisure time(socialization) which were somewhatlower (see Sparrow et al., 2005).Test-retest reliability (approximately2–5-week interval) coefficients werealso good, and the parent forms of theVineland-2 demonstrated adequate intervieweragreement on the Survey InterviewForm and adequate interrater reliabilityacross parents on the Parent/CaregiverForm (Sparrow et al.).ValidityMany aspects of validity are addressedby the authors of the Vineland-2 (Sparrowet al., 2005). The content of theVineland-2 was designed to be reflectiveof AAIDD and DSM-IV-TR criteriafor adaptive functioning, as outlined fordiagnoses of intellectual disabilities ormental retardation. In addition, the rawscores increase lawfully from age to agelending credence to the argument thatthe Vineland measures adaptive behavioras a developmental phenomenon.In other words, with development, anindividual should acquire more adaptiveskills, and this phenomenon is reflectedin the Vineland-2 raw score data.Differential validity studies describedby the authors note that the Vineland-2Adaptive Behavior Composite and domainscores differentiated among individualswith mild, moderate, or severe mentalretardation. Similarly, the scores for individualswith autism who also had verbalskills were higher than those for individualswith autism without verbal skills (Sparrowet al.).Convergent validity was examinedby the associations between Vineland-2domains and analogous scales on theBASC-2. In general, parent reportedVineland-2 domains were moderatelycorrelated with similar parent reportBASC-2 scales (e.g., Communication withFunctional Communication). These coefficientswere generally higher for individualsages 12–18 than for younger subjects.Similar results were found for analyses ofthe Vineland-2 and the Adaptive BehaviorAssessment System-II (ABAS-II; seebelow), with moderate correlations foundacross age groups for analogous domainsand subdomains.Strengths and WeaknessesThe Vineland-2 and its predecessors benefitfrom a long history of successful useand numerous research investigations. TheVineland’s noteworthy strengths includethe following:1. Multiple components that are useful fora variety of diagnostic and interventionplanning purposes2. A supportive research base that suggeststhat the Vineland-2 possesses expectedcorrelations with measures of similar(convergent validity) constructs3. An exhaustive item pool which allowsfor the ready identification of treatmentgoals and objectives4. A large national normative sample andseveral local norm samples, which makeit particularly well suited for diagnosticdecision making5. Modifications of the new version whicheases administration in either semistructuredinterview or rating scaleformats

chapter 14 Adaptive Behavior Scales325Among the Vineland’s weaknesses are thefollowing:1. The considerable training that isrequired to properly use the semi-structuredinterview technique2. Relatively little research on theVineland-2 which is offset to someextent by research on its predecessorsand its familiarity to many clinicians.3. More limited research on the classroomversion, particularly regarding its utilityfor educational interventions.In Box 14.1 we provide a case exampleillustrating the use of the Vineland-2.Reason for ReferralJoan, a 12-year-old girl, was referred forevaluation by her father to determine ifshe is receiving the education services thatshe needs in school. Joan is a sixth-gradespecial education student who is currentlyin a fifth-grade inclusion classroom whereBox 14.1A Case Example Use in theVineland-2Name: JoanAge: 12 yearsGrade: 6Evaluation ProceduresDifferential Abilities Scale (DAS)Woodcock Johnson Tests of Achievement,3rd edition (WJ-III)Behavior Assessment System for Children(BASC-2): Parent Rating Scales(PRS), Teacher Rating Scales (TRS), andSelf-Report of Personality (SRP)Vineland Adaptive Behavior Scales, 2ndedition (Vineland-2): Classroom RatingScale, Parent/Caretaker Rating Scaleshe reportedly is receiving extra help fromher teacher and an aide throughout theday. Her reading skills are reportedly wellbelow what would be expected for hergrade level. According to Joan’s mother,her school has recommended retainingJoan in the sixth grade next year, and Joan’sfather would like a second opinion.Background InformationAccording to her mother, Joan has lived withher biological parents and paternal grandparentsfor the last 10 years. Joan’s motheris a legal secretary, and her father works ata lumberyard. Joan reportedly has no othersiblings. At birth, Joan weighed 7 lb, 11 ozand was described as a healthy baby. Joan’sdevelopment during infancy was reportedlynormal, and she did not have any seriousdiseases or other difficulties.Mrs. Jordan indicated that Joan isinvolved with her church youth group andthe local 4H Club. She enjoys caring forher pets which include a rabbit, goat, andtwo dogs. Joan also reportedly enjoys makingcrafts with her grandmother.Joan’s teacher and aide were interviewed.Both stated that they are impressedand pleased with Joan’s progress, sincethe beginning of the school year. Shewas described as a nice kid who workshard. Reading reportedly continues to beJoan’s weakest area, and math is a relativestrength. According to her teacher, Joan iseasily frustrated with reading assignments.She is described as a sight reader. Joan’ssocialization skills in the classroom andthe community have reportedly improved.Her teachers and mother noted that shegets along well with her peers and adults inher environment.According to her mother, the schoolteam has recommended placing Joan ina sixth-grade inclusion class next year inorder to give her another year to developbefore starting middle school. Althoughher socialization skills have improved, they

326 chapter 14 Adaptive Behavior Scalesreportedly believe that she needs the additionalyear to mature. Joan’s father reportsthat Joan would prefer to go to the middleschool next year with her friends.Previous EvaluationJoan was reportedly in a speech programuntil third grade for fluency and articulationdifficulties. She continues to haveproblems with expressive speech, but herteachers report that this is not a significantproblem. Joan is described as speakingrelatively clearly in class. Joan was recentlyprescribed glasses, but reportedly she failsto wear them in class, and she did not wearthem during this evaluation.Joan was initially evaluated when shewas in the first grade. From that evaluation,she was diagnosed as mildly intellectuallydelayed and placed in a self-containedclassroom. Joan’s Mental Processing Compositeon the Kaufman Assessment Batteryfor Children, 2nd edition (KABC-2)at that time was in the Borderline range(MPC = 70). She was re-evaluated threeyears later, and her Full Scale score on theWechsler Intelligence Scale for Children-Fourth Edition (WISC-IV) was also in theBorderline range (FSIQ = 74). Her AdaptiveBehavior Composite on the VinelandAdaptive Behavior Scale was also in theBelow Average range (ABC = 72).Behavioral Observations DuringTestingJoan was appropriately dressed and wellgroomedduring the evaluation. No auditoryor motor abnormalities were noted.She also did not appear to have any visualdifficulties, but she was not wearing herprescribed corrective lenses. Rapport waseasily established; however, Joan was nottalkative during the evaluation, and she didnot initiate conversation. At times, she wasdifficult to understand because she mumbled.However, she willingly answered theexaminer’s questions and smiled frequently.She appeared to be nervous in the beginning,but she became more relaxed, as suggestedby more frequent smiling and relaxedbody posture, as testing progressed. Joanwas cooperative and polite. Although Joanworked hard during testing, she becamefrustrated easily, especially during readingexercises. When frustrated, Joan tended torespond more impulsively, shift often in herseat, rub her face, avoid eye contact, andadopt a flat affect.Observations during testing suggest thatJoan’s self-report personality test resultsare somewhat suspect. Her limited readingabilities and verbal expression skills, aswell as her impulsivity on tasks requiringreading limits the potential value of thosefindings.Test Results and InterpretationPsychoeducational Assessment: Cognitivefunctioning was assessed using the DAS.She obtained a General Conceptual Abilityscore on the DAS of 59. There is a 90%probability that Joan’s DAS General ConceptualAbility score falls between 54 and64, indicating that her performance meetsor exceeds less than 1% of her peers. Bothher Verbal and Nonverbal scores were alsoin the Significantly Below Average range,but her Spatial score was somewhat higher,falling in the Significantly Below Averageto Moderately Below Average range. Theseresults indicate that her intellectual functioningis below that of her same-aged peers.These results are also highly consistentwith Joan’s performance on academicachievement measures. Standard scores onthe Kaufman Tests of Educational Achievement(KTEA) ranged from 48 (ReadingComprehension) to 64 (Math Applications),which correspond to percentile ranks ofless than 1. All of the academic skill areasassessed were significantly below average.In general, these results are supported by

chapter 14 Adaptive Behavior Scales327Joan’s prior evaluations, curriculum-basedmeasures taken at her school, and teacherreports.Adaptive Behavior: The Vineland-2 wascompleted by Joan’s mother, who indicatedthat Joan’s overall adaptive behavioris significantly below average for her age.Her Socialization and Daily Living Skillsdomain scores were in the Below Averagerange, suggesting that Joan has some difficultycommunicating her needs and ideasand caring for herself relative to her sameagedpeers. Her rating for the CommunicationDomain, which was well below average,suggests that Joan has substantial difficultyexpressing her ideas verbally or in writtenform and in understanding what she readsor hears from others. Her teacher’s ratingson the Vineland-2 were higher in eachdomain, indicating that her overall adaptivebehavior in the classroom is slightly belowthat of her same-aged peers. Joan’s teacherrated her in the Average range in Daily LivingSkills and Socialization. Her rating inthe Communication Domain was in theBelow Average range.The results of the Vineland-2 suggestthat the following behaviors should beconsidered as objectives for intervention:Communication: Reading simple stories,printing three- and four-word sentences,reading vocabulary of at least tenwords, and writing vocabulary of at leastten wordsDaily Living Skills: Dressing appropriatelyfor the weather, avoiding individualswith contagious diseases, telling time by5-min segments, caring for her hair andfingernails, and using household cleaningproductsSocialization: Controlling anger whendenied own way and attending after schoolor evening activities with same age/gradepeersThe results suggest that Joan’s adaptivefunctioning may be better at schoolthan at home. However, her overall adaptivefunctioning is still below what wouldbe expected for her age, particularly in thearea of Communication. Her intellectualfunctioning and academic achievement aresubstantially below what is expected for herage. The results indicate that Joan meetscriteria for a diagnosis of Mild MentalRetardation.Reports by Joan’s mother and teacheron the BASC-2 of poor functional communication,poor adaptability to change, anddifficulties with self-care (parent report)are consistent with their reports of Joan’sadaptive functioning on the Vineland-2and with her diagnosis of Mild MentalRetardation. Joan’s teacher reported on amoderate level of concerns on the LearningProblems scale of the TRS, which isconsistent with Joan’s history of academicdifficulties.General Recommendations1. From the results of the current evaluation,it is apparent that Joan shouldreceive resource educational servicesthrough her school. Joan should not beexpected to make one year of academicprogress for each year of attendance inschool. Therefore, slow academic attainmentby itself may not warrant retentionin grade. Joan’s parents and school officialsshould communicate regularly as tothe appropriate placement and accommodationsfor Joan.2. Joan’s parents are advised to read someliterature on other children functioningat Joan’s intellectual level to understandhow she learns, what expectationsare reasonable for her, what her needsare, and how to meet those needs.3. The focus of academics for Joan shouldbe to help her develop minimal competencyin independent living skills, includingvocational skills.4. Joan should receive vocational counselingto guide her educational program.

328 chapter 14 Adaptive Behavior Scales5. Joan may need to be reevaluated forthe speech program at school given hercontinued difficulties with communicationand the difficulty of understandingher at times during the presentevaluation.6. If a computer is available, Joan coulduse software designed to increase herreading, math, and spelling skills. Shecould work on the computer alone orwith a teacher. Computer usage wouldincrease her independence, give her theattention that she needs, and provideindividualized instruction.7. Joan should be allowed as much as possibleto be around her regular classroompeers (i.e., education in the least restrictiveenvironment possible), so that shecan learn socialization, coping, andbehavioral skills by observing and interactingwith them.Scales of Independent Behavior:Revised (SIB-R; Bruininks, Woodcock,Weatherman, & Hill, 1996)The SIB-R in many ways resembles theVineland. It is a broad-based assessment ofadaptive behavior spanning an age range ofinfancy through adulthood.ContentThe SIB-R includes 14 subscales that aresubsumed under four clusters (i.e., MotorSkills, Social Interaction and CommunicationSkills, Personal Living Skills, andCommunity Living Skills). These clustersare summarized by a Broad Independencescale composite score. The clusters andtheir subscales are the following:Motor SkillsGross MotorFine MotorSocial Interaction and CommunicationSkillsSocial InteractionLanguage ComprehensionLanguage ExpressionPersonal Living SkillsEating and Meal PreparationToiletingDressingPersonal Self-CareDomestic SkillsCommunity Living SkillsTime and PunctualityMoney and ValueWork SkillsHome/Community OrientationAn Early Development scale is providedon the SIB-R for assessing the adaptive skillsof infants and young children. This scalemay also be of benefit for the assessmentof significantly impaired individuals withdevelopmental ages of 8 or lower.Flexible use of the SIB-R is furtherenhanced by the availability of an IndividualPlan Recommendation form. Thisform aids the process of intervention planningthrough needs identification andprogress monitoring.Another valuable feature of the SIB-Ris the provision of a short form that can beadministered in 10–15 min. The short formconsists of items from the longer scale thatgive a quick indication of the overall presenceor absence of adaptive behavior deficits.A low score on this scale would triggerthe administration of the complete form oranother adaptive behavior scale.A Problem Behaviors scale is also included.This scale measures eight areas of problembehavior:Hurtful to SelfHurtful to OthersDestructive to PropertyDisruptive BehaviorUnusual or Repetitive HabitsSocially Offensive Behavior

chapter 14 Adaptive Behavior Scales329Withdrawal or Inattentive Behavior;Uncooperative BehaviorAn additional advantage of this scale isthat problem behaviors are not only ratedaccording to frequency but also by severity.The severity ratings provide the clinicianwith more guidance to follow in a problem-solvingassessment paradigm (Bruininkset al., 1996).ReliabilitySplit-half coefficients for the clusters arereported in the test manual to be in the.80s to low .90s (Bruininks et al., 1996).The Broad Independence scale reliabilityestimates were commonly in the mid tohigh .90s. Test–retest reliabilities for theclusters and the Broad Independence scalewere similarly high.Administration and ScoringThe SIB-R is administered via structuredinterview or checklist in approximately45–60 min. The questions are merely posedto a parent, teacher, or other caregiver asthey are written. Respondents are givenfour response options for each item:1. Never or rarely, even if asked2. Does, but not well, or about 1/4 of thetime; may need to be asked3. Does fairly well, or about 3/4 of thetime; may need to be asked4. Does very well, or always or almostalways; without being askedThe items are worded as precise behavioralobjectives that require little comprehensionon the part of the informant.The SIB-R offers a large and usefulvariety of derived scores includingstandard scores, percentile ranks, andage equivalents. The more novel scoresinclude training implication ranges, a relativemastery index, and a support score,although more traditional scores of differentadaptive behavior domains are alsoavailable.NormingA sample of 2,182 subjects was used fornorming the SIB-R (Bruininks et al., 1996).The age range of the sample extended frominfancy through adulthood.ValidityDeStefano and Thompson (1990) laudedthe original SIB for its evidence of contentvalidity by the observation: “The SIBshows good content validity, in that itsstructure and content cover a broad rangeof skills and traits included in current modelsof adaptive and maladaptive behavior”(p. 461).Roberts et al. (1993) conducted a correlationalstudy of the previous SIB andVineland Survey Form that demonstratesconsiderable criterion related validity forboth instruments. This study involved 1284-year olds to 7-year olds with developmentaldisabilities. Both tests were foundto produce one large factor. Similarly,McGrew, Bruininks, and Thurlow (1992)found the SIB to correlate significantly withcommunity adjustment for 239 adults withmild to severe levels of mental retardation.While the SIB-R is not as well-researchedas its predecessor or the Vineland, the fewstudies available show good evidence ofcriterion-related validity.InterpretationAs with the other adaptive behavior scalesreviewed in this chapter, we recommendinterpretation of the SIB-R first at the compositelevel and then at the cluster level.In most cases, the composite will give anaccurate overall picture of the individual’sadaptive functioning. However, there may

330 chapter 14 Adaptive Behavior Scalesbe relative strengths or weaknesses acrossclusters (and perhaps on specific items/skills) that would inform intervention ina meaningful way. It is good practice forclinicians to consider what specific sets ofskills led to particular strengths or weaknessesin areas of adaptive functioningto illustrate the skills that the client hasacquired or struggled to acquire.Strengths and WeaknessesThe SIB-R possesses many admirabletraits that make it a very practical tool. Itsstrengths include the following:1. Flexibility of administration with theavailability of both short and long forms2. An objective item-scoring scheme thateases scoring3. Its link to other Woodcock-Johnsontests, which foster transfer of training tothe SIB-R.4. Broad item coverage (i.e., content validity)The weaknesses of the SIB-R includethe following:1. Relatively complex scoring algorithms(which are eased considerably by theuse of its accompanying Compuscoresoftware program)2. A lack of factor-analytic and criterionrelatedvalidity studies conducted bythose other than the test developersAdaptive Behavior AssessmentSystem, 2nd edition (ABAS-II;Harrison & Oakland, 2003)The ABAS-II, a newer adaptive behaviorscale, is organized in a manner that isclosely related to DSM and AAIDD criteria.It is a comprehensive system that isdesigned to assess individuals from 5 to 89years of age (Harrison & Oakland).ContentThe ABAS assesses ten constructs including:CommunicationCommunity UseFunctional AcademicsHome/School Living (denoted as“Home” on the Parent Form and“School” on the Teacher Form)Health and SafetyLeisureSelf-CareSelf-DirectionSocial Work (only completed if the individualhas a full or part-time job)The ABAS-II includes various forms thatassess the ten skill areas. Specifically theABAS-II consists of a parent and a teacher/daycare provider form for ages 0–5, separateparent and teacher forms for ages 5–21, andan adult form, completed by a caregiver, forages 16 through 89 years.Administration and ScoringThe ABAS-II is a rating scale that appearsto be straightforward to use. According tothe manual (Harrison & Oakland, 2003),an informant can rate the child in 15–20min. Items are organized by skill area inorder of difficulty, which probably reducesadministration time. The work area isskipped for most children, for example,and many skills listed are so far beyond thecompetencies of most children that theycan quickly be answered with zero.The item response scale is 0 = is not able;1 = never when needed; 2 = sometimes whenneeded; 3 = always or “almost always” whenneeded; and a last category is “check if youguessed.” This latter category may helpexaminers determine the amount of confidenceto place in a rater’s responses.Item scores merely need to be summedto produce raw scores by skill area. The rawscores are then transferred to the back ofthe record form to conduct table conver-

chapter 14 Adaptive Behavior Scales331sions to standard scores for each scale (M =10, SD = 3) and a “General Adaptive Composite”(M = 100, SD = 15). Percentile ranksand confidence intervals are also offered.NormingThe ABAS was normed on a total sampleof 7,370 individuals with demographic samplingbased on 1999 US Census estimates.General national norms are offered as wellas numerous validity studies of individualswith disabilities. Stratification variablesfor the sample included geographic region,parental educational attainment, sex, andrace/ethnicity. The norming sample mayhave included a slight under-representationof children with disabilities (see Richardson& Burns, 2005).ReliabilityReliability of the ABAS-II was evaluatedvia internal consistency, test–retest, andinter-rater reliability. Reliability coefficientswere generally quite high with theinternal consistency coefficients for theGeneral Adaptive Composite and the skillareas being all around .90 and the interraterreliability coefficients were generallyin the .80s. In addition, cross-informantcoefficients across forms and skill areaswere good (i.e., approximately .70; Harrison& Oakland, 2003).ValidityThe ABAS-II has been described as havinggood content validity on the basis ofits theoretical foundation in diagnosticcriteria for mental retardation (Richardson& Burns, 2005). Confirmatory factoranalyses support the factor structure of theABAS-II, and the manual presents goodevidence of both convergent and divergentvalidity (Harrision and Oakland, 2003).In addition, the differential validity of theABAS-II was demonstrated for childrenin a variety of diagnostic categories comparedto matched controls. Richardson andBurns (2005) suggest that more research isneeded concerning the applicability of variousitems of the ABAS-II cross-culturally.The same could be said for other measuresof adaptive functioning.InterpretationABAS-II interpretive guidelines are offeredin the manual along with a few case studies.The interpretive suggestions are basic,clear, and sensible. We again recommend atop-down approach to interpretation (i.e.,composite to items) with skill examplesbeing provided to illustrate the client’srelative strengths and weaknesses. In addition,corroborating evidence is needed onadaptive functioning such that clinicians donot hastily use a profile of scores to make adiagnostic decision (Harrison & Oakland,2003).Strengths and WeaknessesThe ABAS-II is a viable alternative to theVineland and SIB-R given its ease of useand linkage to the AAIDD and DSM-IVcriteria. Noteworthy strengths include thefollowing:1. A close link to diagnostic criteria2. Ease of administration and hand scoring3. Inclusion of numerous validity studiesin the manualSome potential weaknesses of the ABAS-IImay include the following:1. Inability to gather unsolicited clinicalinformation as it is often available in asemi-structured interview2. Lack of validation research by othersthan the test developers3. Slight underrepresentation of childrenwith disabilities in the norming sample.

332 chapter 14 Adaptive Behavior ScalesMeasuring Social SkillsA universally accepted definition of adaptivebehavior remains elusive resulting inscales that may have a variety of domains(Merrell & Popinga, 1994). The developmentand revision of adaptive behaviorinstruments of late have resulted in measuresthat reflect AAIDD criteria for mentalretardation. However, any one of thosedomains (e.g., daily living skills, socialskills) can be assessed in a more detailedfashion than is typically the case with adaptivebehavior scales. One aspect of adaptivebehavior, social competence, or socialskills is of such importance for a variety ofchild outcomes that scales which measurethis construct exclusively are available.Social Skills Rating SystemThe SSRS (Gresham & Elliott, 1990) is acomprehensive measure of social skills thatincorporates multiple domains and raters.Teacher, parent, and student forms areprovided for measuring a variety of socialskills across settings. Although the SSRSalso measures externalizing, internalizing,and hyperactivity problem behaviors,social skills are the focus of the system. Theteacher form of the SSRS also includes arating of academic competence.ContentThe SSRS differs from other crude assessmentsof social skills or omnibus rating scales(e.g., BASC-2) by assessing multiple domainsof social skills. The domains correspondingto each form are shown in Table 14.2. A totalscore is also available for each form.Administration and ScoringEach of the SSRS forms is straightforwardto use. The parent and teacher forms aresomewhat unique rating scales in comparisonto many of the measures reviewed thusfar. In addition to requiring the rater toassess the frequency of a behavior, the rateris also asked to indicate the importance ofthe behavior for a child’s development.The availability of these importance ratingsallows the clinician to better prioritizebehaviors for intervention.A full range of scores are offered by theSSRS, including standard scores, percentileranks, and a behavior level. Behavior levelsare somewhat like stanines in that they divideup portions of the distribution of scores intothree levels where > + l sd = more,

chapter 14 Adaptive Behavior Scales333average. Such indices are available for eachdomain and the total score.NormingThe SSRS norming samples consisted of4,170 children, 1,027 parents, and 259 teacherswho completed forms in 1988 (Gresham& Elliott, 1990). The sample used for theStudent Form is well-described in themanual. On the other hand, the Parent andTeacher Form samples are described in onlya few paragraphs. The characteristics of thechildren rated on the Teacher Form, such astheir SES and ethnicity, are not given. TheSES of the parent sample is heavily skewedtoward high levels of SES. The lack ofdetailed information given for norming ofthe Teacher and Parent Forms makes it difficultto evaluate the quality of the norms andto make recommendations for their use.ReliabilityThe reliability estimates for the Teacherand Parent forms reported in the SSRSmanual are generally adequate. Mean coefficientalpha reliability estimates for theTeacher Form subscales were in the high.80s and .90s. Parent Form coefficients areslightly lower with more coefficients in the.70s. Student Form reliabilities were generallylow. Coefficients for the Cooperation,Assertion, and Self-Control scales did notexceed .70. The test-retest coefficients forthe Student Form were even more disappointing,with none of the values exceeding.70, including the Total Scale value,which was only .68 (Gresham & Elliott,1990). Moderate correlations between parentand teacher ratings on the SSRS havebeen found (Ruffalo & Elliott, 1997).ValidityConducting criterion-related validity studieswith the SSRS may be difficult becauseof debate about the appropriate criterion.There is not a clear criterion in theassessment of social skills. Some sense ofcriterion-related validity may be gainedfrom the studies reported in the manual(Gresham & Elliott, 1990).A study of the SSRS Teacher Form correlationswith the Social Behavior Assessment(SBA) lends some criterion-related validationto the Teacher Form. The majority ofcorrelations were significant for a sample of79 cases, suggesting that the two measuresshare considerable overlap. The Cooperationsubscale of the Teacher Form was mosthighly correlated with scores from the SBA.This scale correlated -.70 with the Interpersonaldomain of the SBA and -.73 with theTask-Related domain of the SBA.A study of the relation of the TeacherForm to the Achenbach Child BehaviorChecklist-Teacher Report Form (CBCL-TRF) provided mixed support for theProblem Behavior scales. The ExternalizingScale correlated .69 with its counterparton the CBCL, but the Teacher Formcorrelated only .33 with its CBCL counterpart(Gresham & Elliott, 1990).Similar results were obtained for theSSRS Parent Form. In a study of 45 parentratings of children, the Externalizing Scalecorrelated .70 with the Externalizing scoreof the CBCL-PRF. Again, however, theInternalizing Scale correlated only .50 withthe Internalizing Scale of the CBCL-PRF(Gresham & Elliott, 1990). These studies ofthe Problem Behaviors scales of the SSRSsuggest that the Internalizing Scale measuressomething different from the correspondingscale of the widely used Achenbach.The factor analyses provided in theSSRS manual provide limited insights intothe underlying traits assessed by the SSRSprimarily because of the methods used.Principal components were extracted, butfactor analysis was not conducted. Furthermore,the components were apparentlyderived solely on the basis of empiricalmethods. Subsequent research has founda poor fit of the factor structure describedin the manual for parent ratings of elementaryschool-aged children (Van Horn et al.,

334 chapter 14 Adaptive Behavior Scales2007). In particular, Van Horn et al. (2007)found better support for a shorter versionof the SSRS with the same domains as traditionallyused in the SSRS (Gresham &Elliott, 1990). Importantly, that longitudinalstudy also suggested that the structureof the SSRS may vary across ethnic groupsand that the SSRS may evaluate a somewhatdifferent construct for 3rd grade childrenthan for children in kindergarten. In otherwords, SSRS scores may not be a good wayto track changes in social skills over timebecause the scores may change in meaningover time (Van Horn et al., 2007). Similarly,Whiteside, McCarthy, and Miller (2007)found a lack of support for the proposedfactor structure of the SSRS for elementaryschool-aged children. However, in supportof the SSRS’s overall construct validity, theyfound that scores on the SSRS were significantlyrelated to a history of peer problems.Strengths and WeaknessesThe SSRS provides a unique assessment toolfor child clinicians in that it is a thoroughmethod for assessing behaviors that are oftenlabeled as social skills. Furthermore, althoughthe norming of the SSRS is not ideal, it is farsuperior to other measures that do not possessadequate norms and yet are used to makenorm-referenced decisions. Other strengths ofthe SSRS include the following:1. An attempt at a multi-domain assessmentof social skills2. The use of multiple informants3. An integrated method of interpretationand intervention planningSome of the weaknesses of the SSRSinclude the following:1. Inadequately described norm samplesfor the Teacher and Parent Forms2. Poor reliabilities for some of the subscales,especially those on the Student Form3. Potential differences in meanings ofscores across ethnic groups and developmentalstages (Van Horn et al., 2007)4. Emerging research that does not supportthe SSRS factor structure described inthe manual5. Less than adequate criterion-relatedvalidity for the Internalizing scale of theProblem Behaviors domainConclusionsAdaptive functioning is increasingly beingrecognized as an important, if not essential,aspect of child assessments. Although Dollintroduced this intuitive concept in the1930s, adaptive behavior was not formallyincluded as central to the mental retardationdiagnostic process until the 1950s.In addition to its inclusion in diagnosticcriteria, adaptive functioning also hasan important role in intervention planning.To this end, components of adaptivebehavior scales are now included onmany behavior rating scales, such as theAchenbach and BASC-2, and these componentsare summarized in Box 14.2.Anotherinteresting trend is for adaptive behaviorscales to be single-domain measures. TheSSRS is an excellent example of the trendtoward developing assessment measures ofwhat may be called sub-constructs of adaptivebehavior.This chapter highlighted several of themost popular adaptive behavior scales, butthe reader should be aware that there are asubstantial number of such scales available.Chapter Summary1. Edgar Doll first discussed the constructof adaptive behavior by drawingon his experience as a psychologist at

chapter 14 Adaptive Behavior Scales335Box 14.2Adaptive Behavior as a Component of Omnibus Rating ScalesWith recent research and development inthe assessment of adaptive behavior, it is sufficientto say that this construct is no longerconsidered a mere afterthought in childassessment or as a secondary component ofevaluations for mental retardation. Instead,adaptive behavior is viewed as an importanttarget of assessment for describing anindividual’s strengths and weaknesses andfor designed skill-based interventions. Thisincreased recognition of adaptive behavior asa central aspect of child assessments is evidentin recent changes to omnibus rating scales.For example, although the original BASC(Reynolds & Kamphaus, 1992) was uniqueat the time in its multi-domain assessmentof adaptive behavior, the BASC-2 (Reynolds& Kamphaus, 2004) has an expanded set ofadaptive behavior domains. The BASC-2 sawthe addition of a Functional Communication(parent and teacher) and Activities of DailyLiving (parent) scale. The reader will notethat these scales in particular seem consistentwith adaptive behavior domains included inthe measures reviewed in this chapter. Theother adaptive scales (i.e., Adaptability – parentand teacher; Study Skills – teacher; SocialSkills – parent and teacher; Leadership – parentand teacher; Self-reliance – self; Selfesteem– self; Relations with Parents – self;Interpersonal Relations – self) may overlap tosome extent with Socialization, but they mayalso provide some unique information (e.g.,one’s ability to adjust to change in the case ofthe Adaptability scale). Research is quite scanton these scales to date. Initial validation of theBASC-2 found that the adaptive scales weregenerally negatively correlated with clinicalscales of the BASC-2, as well as other measures(Reynolds & Kamphaus, 2004). Such resultsdo not provide information as to the potentialclinical utility of each individual scale. Thatis, although the psychometric properties ofthese scales are adequate, interpretation andclinical use might be best at the item levelin terms of pinpointing specific behavioralconcerns in the areas of adaptability, communication,self-care, and social skills. Onthe basis of state of the relevant research, aclinician cannot say with much confidence,for example, what a low (or high) score onthe Leadership scale of the BASC-2-PRSreally means. However, some items (e.g., “Iscreative;” “Is good at getting people to worktogether.”) may highlight specific strengthsof the child or areas that may need someimprovement.The BASC-2 is not the only omnibus scalethat directly assesses adaptive behavior. TheAchenbach and PIC/SBS/PIY (see Chaps. 6and 7) systems also have strategies for assessingstrengths and socialization. In the case ofthe former, these are limited to a few screenertypeitems regarding overall competencies. Inthe case of the latter, these are framed in termsof social problems, the absence of which wouldimply adaptive social functioning. Therefore,the assessment of adaptive behavior offered onscales such as the BASC-2 or Achenbach cannotbe considered substitutes for more detailedevaluations such as through the Vineland-2or ABAS-II. If clinicians are routinely usingomnibus rating scales as part of comprehensiveassessments, then at the very least theywill now have the ability to efficiently screenfor deficits in adaptive functioning that canbe more thoroughly evaluated through meanssuch as those described in this chapter.the Vineland State Training School,where he was charged with the assessmentand rehabilitation of individualswith mental retardation. He notedthat such individuals not only lackedintellectual abilities that were necessaryfor academic attainment, but theyalso often appeared to lack day-to-dayliving skills needed for independentfunctioning.

336 chapter 14 Adaptive Behavior Scales2. The American Association on Intellectualand Developmental Disabilities(AAIDD), formerly the AAMR, liststhree broad domains (i.e., conceptualskills, social skills, and practical skills) ofadaptive behavior that should be evaluatedin assessments of mental retardation(see Luckasson et al., 2002). In addition,adaptive functioning is an importantconstruct for case conceptualizationand treatment planning for a number ofproblems (e.g., autism, ADHD).3. Adaptive behavior has been defined asthe performance of the daily activitiesthat are required for social and personalsufficiency.4. Adaptive behavior scales serve a valuablefunction for child clinicians in that theypin-point specific deficits that a child hasnot acquired, which in turn may serve asthe focus of treatment efforts.5. Widaman, Stacy, and Borthwick-Duffy(1993) found clear evidence for the existenceof four major adaptive behaviordomains: cognitive competence, socialcompetence, social maladaptation, andpersonal maladaptation. Most present daymeasures of adaptive behavior includedomains consistent with these findings.6. Parents, teachers, and other caregiversare most often used as informants foradaptive behavior scales. Children mayalso be used as self-informants for someadaptive behavior scales, although sucha strategy has some obvious limitations.7. Correlations between intelligence andadaptive behavior measures are modest(i.e., .40 to .60), indicating some overlapbut also substantial independence.8. Adaptive behavior scales are commonlyadministered using a checklist or semistructuredinterview technique.9. The semi-structured interview techniquerequires a high level of clinicalskill. The clinician has to make theinterview conversation-like, topical,and empathic, while at the same timecollecting the necessary informationto allow for accurate rating (scoring)of individual items.10. The Vineland-2 is probably the mostwell-known and widely used tool forassessing adaptive behavior for individualsfrom birth through adulthood.There is an extensive body of researchon its predecessor, and a number ofchanges were made to the present versionto facilitate administration, scoring,and interpretation.11. The Scale of Independent Behavior(SIB-R) in many ways resembles theVineland. It is a broad-based assessmentof adaptive behavior spanningan age range of infancy throughadulthood.12. The ABAS-II is closely aligned withDSM and AAIDD criteria. So far, ithas demonstrated good reliability andvalidity.13. One aspect of adaptive behavior,social competence or social skills, isof such importance that scales thatmeasure this construct exclusively areavailable.14. The Social Skills Rating System(SSRS) is a comprehensive measure ofsocial skills which incorporates multipledomains and raters. Teacher, parent,and student forms are providedfor measuring a variety of social skillsacross settings.15. The assessment of adaptive behavioris increasingly being viewed as a centralcomponent of child assessment.Expanded components of adaptivebehavior scales are now being includedon omnibus rating scales such as theAchenbach and BASC-2.

P a r t I I IAdvanced Topics

C h a p t e r 1 5Integrating and InterpretingAssessment InformationChapter QuestionsIntroduction• What are some reasons why assessmentinformation may differ depending onwho is providing it?• What are the issues involved in usingsimple (equal weighing) or complex(unequal weighing) methods for combininginformation from different sources?• How might age of the child and type ofbehavior affect the agreement betweendifferent sources on a child’s or adolescent’sadjustment?• What steps should one go through tointegrate information from differentsources, from different methods, andacross different areas of functioning?Throughout this book we have emphasizedthe need for a comprehensive evaluationwhen testing children and adolescentsin most circumstances. Comprehensivemeans that (1) many areas of functioningshould be assessed, (2) assessments shouldbe conducted using multiple techniques,and (3) assessments should obtain informationfrom many sources (e.g., child, parent,teachers, peers). If one follows thisadvice, one is confronted with a dizzyingarray of information gathered during theassessment. In the previous chapters wediscussed each individual component ofthe assessment in isolation. However, the339P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_15, © Springer Science+Business Media, LLC 2010

340chapter 15 integrating and interpreting assessment informationmost important and most difficult part ofthe evaluation is the integration of thisinformation into a clear case formulationthat answers the referral questions andpoints the way to the most appropriateintervention.Although we discuss integrating informationacross multiple domains of functioningand across multiple assessment techniques,a main focus of this chapter is on integratinginformation from different informants.This is a particularly difficult endeavorbecause one is theoretically obtaining thesame information on a child’s adjustmentfrom different sources. One might expectthat there would be a high degree of correlationbetween information provided bydifferent informants. Unfortunately, thisis not the case. A large body of researchhas indicated that there are generally quitesmall correlations between different informantsproviding the same type of information(Achenbach, McConaughy, & Howell,1987; De Los Reyes & Kazdin, 2005).One explanation for this low rate ofagreement between informants is that differentpeople see a child in different settings.That is, the low correlations betweeninformants may reflect real differences ina child’s behavior across different settings.The meta-analysis conducted by Achenbachet al. (1987) provides some support forthis possibility. Across the 119 studies thatthey reviewed, the average correlation inratings between informants who see a childin different settings (e.g., parent/teacher)was .28. In contrast, the average correlationbetween informants who see a childin the same setting (e.g., pairs of parents orpairs of teachers) was .60. This substantialincrease in agreement between informantswho see a child in the same setting providesevidence that the lack of informant agreementreflects, at least in part, real situationalvariability in children’s behavior.Konold, Walthall, and Pianta (2004)also provide evidence to support the contentionthat much of the cross-informantdisagreement is due to situational specificbehaviors in children across settings. Specifically,these authors obtained parent andteacher ratings of young children’s behavior.They reported that the general formand factor loadings for the ratings werevery similar across mother, father, andteacher ratings, even though the correlationsamong the ratings from each informantwere fairly low. Thus, the low levelof agreement appeared to be more likelydue to situation-specific behavior, than dueto the manner in which the behaviors weremeasured.This situational variability has importantimplications for clinical assessments.It suggests that to make sense of informationon a child’s emotional and behavioralfunctioning, a clinical assessor must alsoassess the contextual demands that mightinfluence a child’s behavior and, thereby,account for the discrepant information oneis obtaining. For example, a clinical assessormight be quite puzzled over a teacher’sreport of significant problems of inattention,disorganization, and hyperactivityat school, when it appears to be in starkcontrast to only minimal problems beingreported by a child’s parent. However, furtherassessment of the home context mayindicate that the parents do not put manydemands for sustained attention, organization,or sitting still on the child at home. Asa result, the behaviors are not problematicin this setting.Another important aspect of the lack ofagreement between informants is determiningthe level of analysis at which agreement isbeing measured. Individual behaviors tendto show less consistency across informantsthan broader dimensions or composites ofmultiple behaviors. A good example of thisphenomenon comes from a study by Biederman,Keenan, and Faraone (1990). In aclinic sample of children and adolescents,the average correlation between parent andteacher reports of individual symptoms ofAttention-Deficit Hyperactivity Disorder

chapter 15 integrating and interpreting assessment information341(ADHD) on a structured interview wasquite low (r = .21). However, these authorsfound that, if one looked at the diagnosticlevel, the level of agreement between parentsand teachers was quite high. If a parentreported that a child had ADHD onthe structured interview, there was a 90%probability that the child’s teacher wouldreport the presence of the disorder.There are two ways one can conceptualizethis issue. On a psychometric level,classic measurement theory states that anaggregate measure shows higher reliabilitythan its individual components (Peterson,Kolen, & Hoover, 1989). Alternatively, thesame issue can be framed as another aspect ofthe situational specificity of behavior. Thatis, the same underlying construct or psychologicaldimension may be present acrosssituations, but the behavioral manifestationmay change, depending on the situationaldemands. If one frames it in this way, theimplication for clinical assessment is thesame as that discussed previously. To understandthe different behavioral manifestations,one must understand the demands ofthe different contexts under consideration.Karver (2006) provided data to suggestthat some of the disagreement betweeninformants is related to the types of behaviorsbeing assessed. Using 20 judges forratings of 59 child behaviors on 11 dimensions,three primary dimensions of behavioraccounted for 43% of the level agreementbetween parent and child ratings. The firsttwo dimensions relate to the saliency of thebehavior to the parent or child (e.g., ease ofrecall of the behavior, objective nature ofthe behavior, perceived seriousness of thebehavior). The third dimension involvedhow observable the behavior was and howwilling the parent or child would be toreport the behavior accurately because ofsocial desirability.One final issue on informant disagreementis the fact that some apparent disagreementmay be an artifact of differentmeasurement techniques used across informants.An example is a case in which achild is showing indications of depressionon a projective technique but this is notreported by parent and teacher on ratingscales. A clinical assessor may consider thisa disagreement across informants. In fact,this is more a function of the lack of agreementacross methods, because studies haveshown that the agreement between projectiveand objective self-report measures isgenerally low (Ball, et al., 1991). However,even within the same method, what isassessed may differ across informants. Forexample, some rating scales include verydifferent items for parents and teachers(Richters, 1992). As a result, it becomesunclear whether disagreements are due toactual differences in perceptions betweenparents and teachers or whether they aredue to different questions being asked ofthe two informants.Considering these issues, clinical assessorsshould first try to determine if seeminglydiscrepant information can beattributed solely to varying situationaldemands, the level of analysis, the types ofbehaviors being assessed, or to differencesin the assessment strategy. If the discrepantinformation can be explained by these factors,one has gone a long way in explainingthe reasons for a child’s problems and oneusually has obtained important informationfor designing treatment. For example, if achild is displaying highly anxious behaviorin the school setting only, one should lookfor what aspects of the school environmentare leading to the anxiety (e.g., social-evaluativesituations). In doing this, one haspointed the way to one route of intervention.That is, to systematically desensitizethe child to the anxiety-provoking stimuli.Unfortunately, in many situations one cannotbe confident that any of these factorsare contributing to the discrepant information.In such cases, one is confronted withthe difficult decision of what to do withconflicting information in developing a caseformulation.

342chapter 15 integrating and interpreting assessment informationIntegrating Informationacross InformantsSimpler May be BetterTo deal with this issue of conflicting information,clinical assessors have often developedtheir own algorithm or methods ofdeciding how to weigh the reports of differentinformants. In Box 15.1 we summarizethe results of a survey of child mentalhealth professionals in which respondentsrated the importance of different informantsfor assessing various types of childhoodpsychopathology. This provides agood illustration of common practice inweighing different informants. In Box 15.2we summarize an explicit diagnostic decisiontree that was developed to standardizethe diagnostic decision-making process.This decision tree was designed for usewith structured diagnostic interviews. Wefeel that these two pieces of research illustratethe problems one faces when attemptingto systematically integrate discrepantBox 15.1A Survey of Mental Health Professionals’ Use of Different Informants to AssessChildhood PsychopathologyLoeber, Green, and Lahey (1990) conducteda survey of 105 members of the Society forResearch in Child and Adolescent Psychopathologyassessing their perceptions of therelative usefulness of pre-pubertal children,their mothers, and their teachers as informantson emotional and behavioral problemsin children. Two-thirds of the respondents hadPh.D’s and one-third had M.D.’s. Fifty-ninepercent described themselves as child researchersand clinicians, 34% described themselves asresearchers only, 1.9% described themselves asclinicians only, and 5% described themselves inan “other” category. Respondents rated the utilityof information provided by mothers, teachers,and 7–12-year-old children on a 0 (notuseful) to 3 (very useful) scale. There were 44behaviors rated covering the domains of inattention/hyperactivity,oppositional problems,conduct problems, anxiety, and depression.These authors found that the perceivedrelative usefulness of children, mothers, andteachers varied as a function of the domainbeing assessed. Specifically, teachers were ratedas more useful than both parents and childrenin the assessment of inattention/hyperactivity.In contrast, parents were rated as more usefulthan teachers and children in assessing severeantisocial behavior and aggression. Both parentsand teachers were rated as more useful inassessing children’s oppositional behavior thanchildren themselves, although there was no differencein the perceived usefulness of parentsand teachers within this domain. In general,parents and children were judged more usefulin assessing internalizing problems (anxietyand depression) than teachers. Somewhatsurprisingly, parents were judged significantlymore useful than children in assessing internalizingproblems.In sum, these authors found that there weresystematic preferences for certain informantsover others in assessing childhood psychopathology,and these preferences varied as a functionof the behavioral domain being assessed.These authors correctly point out that theresults were limited to mental health professionals’perceptions of pre-pubertal (ages 7–12)children. The results may have been different,especially in the relative importance placed onchild self-report, if the assessment of psychopathologyin adolescents had been the focus ofthe survey (see for example Cantwell, Lewinsohn,Rhode, & Seeley, 1997).Source: Loeber, R., Green, S. M., & Lahey, B. B. (1990). Mental health professionals’ perception of the utility ofchildren, mothers, and teachers as informants on childhood psychopathology. Journal of Clinical Child Psychology,19, 136–143.

chapter 15 integrating and interpreting assessment information343Box 15.2An Attempt to Develop Rules for Combining Multiple Sources of Information onChildhood PsychopathologyThe struggle to develop standardized methodsof integrating information from multipleinformants is exemplified in an article by Reichand Earls (1987). These researchers set out todevelop “a replicable strategy to make psychiatricdiagnoses when reports are obtainableindependently from parents and children” (p.601). The authors clearly state that the goalwas not to achieve perfect parent-child agreementbut to learn how to evaluate these differentsources of information. The followingis a summary (see appendix of research articlefor full criteria) of the decision tree used bythese authors to make diagnoses according tothe DSM-III (American Psychiatric Association,1980). We present this approach not somuch as a recommended decision tree, but asan example of strategies that assessors havedeveloped to resolve the problem of discrepantinformation.1. Attention-Deficit Disorder(a) Minimum of eight symptoms fromparent report, at least six symptomsfrom child report, and an age of onsetbefore 7.(b) Evidence of impairment in schoolfrom teacher’s report either of significantinattention, disruptive behavior,academic underachievement, or peerdifficulties.(c) If parent reports six or seven symptomsand child report shows four or five,diagnosis can be made with compellingevidence from teacher.2. Oppositional Disorder(a) Minimum of two symptoms from bothparent and child and a report of at least6 months duration.(b) Teacher indicates a pattern of negativeand defiant behavior.3. Conduct Disorder(a) Two or more symptoms from eitherparent or child.(b) Confirmation of antisocial behaviorfrom at least one other source.4. Major Depression(a) For children over 13, a diagnosis canbe made from child report alone if (1)child reports dysphoric mood or anhedoniafor 2 weeks or longer; (2) at leastfour vegetative symptoms from DSM-III criteria are present for 2 weeks orlonger; (3) there is evidence of impairment,such as grades dropping, irritability,social withdrawal, etc.(b) A diagnosis of depression would normallynot be made by parents reportalone unless child report was justunder threshold.5. Separation Anxiety and Overanxious Disorder(a) Diagnosis can be made from eitherparent or child report alone only if theother informant provides some evidenceof anxiety and/or depressive symptoms.Source: Reich, W., & Earls, E. (1987). Rules for making psychiatric diagnoses in children on the basis ofmultiple sources of information: Preliminary strategies. Journal of Abnormal Child Psychology, 15, 601–616.assessment information either in clinicalpractice or in clinical research (see alsoSmith, 2007).Piacentini, Cohen, and Cohen (1992)provide an interesting discussion on developingideal weighing systems for combininginformation across multiple informants.These authors label systems in which onesource of information is weighed moreheavily than others (e.g., teachers’ reportof inattention given more weight than parents’report) as complex schemes. In contrast,simple schemes are those where informationfrom all sources is weighed equally. These

344chapter 15 integrating and interpreting assessment informationsimple schemes can also be called the either/or rule, since a finding is considered significantif it is reported by any informant (e.g.,parent, child, or teacher).These authors provide both a theoreticaland an empirical rationale for whysimple schemes are either as good as orbetter than complex schemes. First, theseauthors argue that most weighing systemsrely on clinical judgment, in the absence ofany clear research base to guide decisions.They cite numerous findings that “expertjudgments” are “almost always inferior toequal weighing in predicting outcomesbased on multiple variables” (p. 54). Thisrationale, however, would only hold forweighing systems based on clinical judgment.What about empirically derivedweighing systems?The authors cite a study in which a complexmethod of combining information wasdeveloped on the basis of logistic regressionanalyses. When Bird, Gould, and Staghezza(1992) compared their empirically derivedweighing procedure with a simple either/orscheme, there were no differences in howwell either system predicted clinicians’ diagnoses.Moreover, weighing systems basedon regression analyses from one sampleare likely to underperform simple schemeswhen applied in a different sample, as theweighing system is at least partially dependenton the idiosyncratic characteristics ofthe sample. In summary, Piacentini, Cohen,and Cohen (1992) provide compellingarguments for the use of simple schemes incombining discrepant reports.Although in principle this approachseems sound and we advocate that itshould be the primary or default optionfor most clinicians, there are several issuesthat one needs to consider when using thisapproach. First, this approach assumesthat false positives (i.e., a clinically significantfinding from an informant that is nottrue) are rarer and less harmful than falsenegatives. For many clinical situations, thisemphasis on ensuring that children in needof treatment are not missed is appropriate.However, we also feel that such a decisionshould not be made blindly. Cliniciansshould always be careful in diagnosing andclassifying childhood problems and carefullyconsider both positive and negativeconsequences of any decision (see Chap. 3for issues in diagnosis and classification).Therefore, one must carefully consider therisks and benefits of both false positivesand false negatives when developing caseformulations.A second issue was raised by Piacentini,Cohen, and Cohen (1992), who noted thatsimple schemes are better only if informantsare asked to provide “only that informationthey would ordinarily be expectedto know” (p. 59). In essence, this impliesthat there will be certain informants whohave better knowledge of a child’s adjustmentthan others. For example, teachersmay not have knowledge about certainbehaviors that occur outside of the schoolsetting, such as firesetting, cruelty to animals,or needing to sleep with parents. Inaddition to the type of adjustment beingassessed, we feel that there are other factorsthat may make some informants betterthan others in a given case. In the nextseveral sections we discuss these factorsthat may affect the quality of informationobtained from specific sources. From theoutset we should note that most of thesefactors largely will affect the planning ofthe evaluation rather than the interpretation.Specifically, many of the factors willaffect the choice of informant and of themethod used to obtain information.Informant Discrepanciesand the Type of ProblemThe survey by Loeber, Green, and Lahey(1990), cited in Box 15.1, indicated thatclinicians tend to weigh adult informants(e.g., parents, teachers) more heavily forobservable behaviors and tend to weigh

chapter 15 integrating and interpreting assessment information345child self-report more heavily for emotionalproblems. The logical basis to thisargument is that emotional distress islargely an internal subjective event that ismore difficult to tie to behavioral referentsthat can be observed by others. However,there is a growing body of research thathas largely supported this method of clinicalpractice.For example, mothers and fathers showa higher level of agreement in rating children’sobjective, observable, behaviorproblems than in rating less observableemotional problems (Christensen, Mergolin,& Sullaway, 1992). Also, researchhas found that adults tend to report moreconduct problems and overactivity thanis reported by child self-report (Kashani,et al., 1985; Reich & Earls, 1987), andchildren tend to report more emotionalsymptoms than do parents or teachers(Bird, Gould, & Stagheeza, 1992; Cantwellet al., 1997; Edelbrock et al., 1986). Thisdifferential prevalence in reporting wouldsupport the common clinical practiceof emphasizing adult informants in theassessment of externalizing problems andemphasizing children in the assessment ofinternalizing problems, if one assumes thatfalse positives are rare. This view is consistentwith the either/or approach discussedin the previous section.However, there are other studies thathave attempted to go beyond simply documentingdifferences in prevalence to testthe differential validity of various informantson child adjustment. These studiestest the differential validity of various informantsacross behavioral domains usingsome external and clinically importantcriteria. A good example of this approachis a study by Loeber et al. (1991). In a clinicreferred sample of boys between the agesof 6 and 13, these authors tested the correlationsamong parent, child, and teacherreports of disruptive behaviors and severalimportant impairment criteria (e.g.,school suspensions, police contacts, graderetention, special education placement)assessed one year later. Supporting commonclinical practice, teachers’ report ofinattentive/hyperactive behaviors werethe best predictors of impairment oneyear later. For conduct problem behaviors,parental report tended to show themost consistent prediction of impairment,although teachers’ and children’s informationalso had utility in predicting schoolsuspensions. These findings should beinterpreted in light of the age of the samplestudied. Specifically, this study focused onelementary school-age children. As we willdiscuss in more detail in the next section,there is some evidence that the age of thechild may influence the importance of differentsources of information. For example,the importance of child report for assessingantisocial behavior, especially covert antisocialbehavior (e.g., lying, stealing), mayincrease in adolescents, while the importanceof parental report may decrease asparents’ knowledge of their child’s behaviorsdecreases (e.g., Cantwell, et al., 1997;Elliott, Huizinga, & Ageton, 1985).The validation of different informantsrequires the measurement of some independentand clinically important criteriaagainst which to judge the usefulness ofinformation provided. This has been veryproblematic for testing the differentialvalidity of informants’ reports of internalizingchild behavior because it is difficultto determine what should be the validatingcriteria. One notable exception was a studyby Frick, Silverthorn, and Evans (1994)that used a maternal history of anxiety asa way of validating informants’ reportsof childhood anxiety. Specifically, theseauthors studied the correlations amongparent, child, and teacher reports of anxietyin children and a history of an anxietydisorder in mother. In the sample of childrenages 9–13, these authors found thatteacher report of anxiety was not relatedto a family history of anxiety, but bothparent and child reports of anxiety were.

346chapter 15 integrating and interpreting assessment informationHowever, there was additional support forthe relative importance of children’s selfreportof anxiety. Children’s self-reportsof anxiety disorders were always accompaniedby confirmation by their parents.However, parental report was not alwaysconfirmed by the child, and parent–childdisagreements on the presence of an anxietydisorder were systematically related tothe parents’ own level of anxiety. Theseauthors concluded that there was evidencefor parental projection of their own anxietyonto their reports of anxiety in their children.This provides some preliminary supportfor the relative emphasis on children’sself-report of anxiety, at least in this agegroup (9–13 years) of children.Informant Discrepanciesand Age of ChildAnother factor that might affect the qualityof information provided by different informantsis the age of the child or adolescentbeing assessed. First, one would expectthat, as a child grows older, parents wouldhave less knowledge of the child’s emotionsand behaviors, especially as parent–child relationships change in adolescence(Paikoff & Brooks-Gunn, 1991). Similarly,as a child leaves early elementary school,the likelihood of a single teacher spendinga great deal of time with a child decreases.Third, as a child develops cognitively, heor she may become better able to reporton such abstract concepts as emotionsand thoughts. As a result of these factors,one would expect that the importance ofparents and teachers as informants woulddecrease with age and the importance ofchildren’s self-report would increase.Edelbrock et al. (1985) tested thesehypothesized age effects on parent andchild reports using structured diagnosticinterviews. Using two-week test-retestcorrelations as the index of reliability, theseauthors found that the reliability of parentreport decreased with age as predicted,although the reliability generally remainedat an acceptable level into adolescence.Children’s self-report showed a more dramaticage trend. Children’s self-report onthe structured interviews showed a clearincrease in reliability with age. Importantly,the reliability of child self-report onthe structured interviews was quite poorbefore age 9. Evidence for the predicteddecrease in reliability of teacher’s reportalso was obtained from another source.Specifically, in the initial developmentof the Behavioral Assessment System forChildren, it was found that the reliabilityfor the Teacher Report Form decreasedwith age (Reynolds & Kamphaus, 1992).One question in interpreting theseresults is whether the changes in reliabilityacross age are confined to structuredtechniques, like structured interviewsand behavior rating scales. At least thefindings for children’s self-report are notconfined to structured techniques. Thereliability of children’s responses to projectivetechniques has also been foundto increase with age (Exner, Thomas, &Mason, 1985). Therefore, the hypothesizedage-related changes in the reliabilityof various informants have generallybeen supported in research across a numberof assessment domains. These findingson the limited reliability of childself-report for very young children maynot be surprising for many clinicianswho work with young children. However,clinical assessors who work largelywith older children, adolescents, oradults tend to use a traditional approachto assessment that relies heavily on selfreportand may be uncomfortable with areduced role of self-report in the assessmentof young children.A final issue related to informant discrepanciesand age of the child relates tothe level of agreement between informantsother than the child (e.g. parent and teachers).In their meta-analysis of 119 studies,Achenbach et al. (1987) reported that parentand teacher ratings were in greater

chapter 15 integrating and interpreting assessment information347agreement for younger children than foradolescents. As noted above, part of thisage trend may be due to adolescents spendingless time with teachers and parentsand, as a result, these informants may haveless complete information on the adolescent’sadjustment. However, Achenbachet al. also suggest that younger children’sbehavior may be more cross-situationallyconsistent. Importantly, however, theseage related findings have not always beenconsistently reported across samples (DeLos Reyes & Kazdin, 2005).Informant Discrepancies: OtherFactorsIn addition to type of behaviors and age ofthe child, there are other factors that mightinfluence the report of various informantsand which therefore should be consideredwhen interpreting discrepant information.In a previous chapter we discussed factorswithin the family, such as parental adjustmentand marital conflict, which may affectthe information provided by the parent on achild’s adjustment. For example, Foley et al.(2005) reported on a sample of 2,798 twinsages 8–17. They found that maternal alcoholismand marital difficulties were characteristicof cases in which an anxiety disorderwas reported by mothers but not children.Also, informants might have differingmotivations, both conscious and unconscious,that can affect the informationthey provide (Karver, 2006). For example,a child may not want to admit to problembehavior or a teacher may be interested ingetting a child placed outside of his class.Several rating scales discussed in previouschapters include validity scales that attemptto detect such response sets and aid in theinterpretation of information provided.It is also important to note testing conditionswhen interpreting the report of differentinformants. For example, a childmay have been administered a self-reportquestionnaire after a long testing sessionand it is obvious in watching him completethe questionnaire that he is not reading theitems carefully.SummaryFrom this discussion it is clear that,although the simple scheme of equallyweighing the report of different informantsusing an either/or approach is a good startingpoint, a clinical assessor cannot use thisapproach blindly. There are numerous factorsthat must be considered in trying toexplain seemingly discrepant informationfrom different sources.The previous discussion outlines someof the more important issues that havebeen uncovered in research that can helpguide clinical decision making. However,the final case formulation that results fromthe integration of multiple types of assessmentinformation involves a number ofcomplex clinical decisions. To aid in thisprocess, in Box 15.3 we provide a summaryof an article by Nezu and Nezu (1993) thatoutlines (1) some of the common cognitivestrategies that are used by people inmaking decisions but which could lead toerrors in clinical reasoning and (2) a generalorientation to clinical reasoning thatcan minimize the effects of these errorsin clinical judgment. Using this generalproblem orientation as a basis, the followingsection outlines a step-by-step strategythat can help to guide the clinician in theintegration of assessment information.A Multistep Strategy forIntegrating InformationThe following strategy assumes one hasconducted a comprehensive clinical evaluationof the child or adolescent. The prerequisitesare (1) having information ona child’s adjustment from various sources

348chapter 15 integrating and interpreting assessment informationBox 15.3Research Note: Common Errors in Clinical Reasoning and a Problem-SolvingModel for Developing a Case FormulationNezu and Nezu (1993) conceptualize a clinicalassessment as a method of translating a client’scomplaints of distress into a meaningful set oftarget problems and treatment goals. This iswhat we have referred to throughout this bookas developing a case formulation. Nezu andNezu conceptualize this process as delineatinginstrumental outcomes (target problems)that are believed to directly or indirectly relateto a desired outcome for the client (treatmentgoals). For clinicians to use assessment informationin delineating instrumental outcomes,they must go through a complex problemsolvingprocess. Nezu and Nezu first presentseveral common cognitive strategies that canlead to errors in the problem-solving process.Next, these authors outline an orientation toproblem-solving that should limit these errorsin clinical reasoning.Errors in Clinical Reasoning1. The availability heuristic occurs when cliniciansattempt to estimate the probabilityof an event on the basis of the ease withwhich examples of that event come tomind. Nezu and Nezu use the example of aclinician who may overestimate the risk ofsuicide in a new client, if the clinician wasrecently involved in the case of a client whocommitted suicide but was judged to be oflow risk.2. The representative heuristic occurs when aschema is accessed by a given characteristicto the exclusion of other schemas.For example, the symptom of sadness mayautomatically lead the clinician to considermajor depression (i.e., diagnostic schema)to the exclusion of other possible diagnosticschemas (e.g., medically related mooddisorder, personality disorders).3. The anchoring heuristic occurs when predictionsor decisions are overly dependenton initial impressions and later informationis discounted, even if the new informationis in disagreement with the initialimpressions. For example, if a depressedclient discusses interpersonal problemsin the initial meeting with the clinician,the anchoring heuristic would lead theclinician to select poor social skills as theprimary target of intervention to the exclusionof other information (e.g., presenceof negative cognitions) obtained in subsequentassessment.4. Confirmatory search strategies involve usingprocedures that seek only to confirm initialimpressions and failing to seek disconfirmingevidence. For example, a child referredfor problems of inattention and motor restlessnessis assessed only for other symptomsof ADHD, without considering other possiblereasons for his or her symptoms (e.g.,learning disability, depression, anxiety).Problem OrientationNezu and Nezu propose that a systematicproblem-solving approach to clinical decisionmakinghelps to limit the use of these cognitiveheuristics. The key to this approach is theclinician’s problem orientation. This refers tothe clinician’s overall theoretical orien tationfor viewing problem behavior, which definesthe proper content and methods of assessment.These authors recommend an orientation ofplanned critical multiplism, which assumes thatclinical outcomes are brought about by a multiplicityof interacting factors. The basis ofthis problem orientation is that:“A variety of biological (e.g., genetic, neurochemical,physical), psychological (e.g.,affective, cognitive, behavioral) and social(e.g., social and physical environment) variablescan serve to act and interact as casualagents or maintaining factors. Such variablescan influence the pathogenesis of a symptom,disorder or both in either a proximal (e.g.,(Continues)

chapter 15 integrating and interpreting assessment information349Box 15.3 (Continued)immediate antecedent stimulus) or distal (e.g.,develop-mental history) manner” (p. 256).Nezu and Nezu state that even if this modelis incorrect in a given case (i.e., there is a unitarycause), by starting with the planned criticalmultiplism orientation, the clinician is lesslikely to selectively focus on any single typeor class of instrumental outcomes. As a result,this problem orientation limits the potentialfor cognitive biases.Source: Nezu, A. M. & Nezu, C. M. (1993). Identifying and selecting target problems for clinical interventions:A problem-solving model. Psychological Assessment, 5, 254–263.and using multiple methods; (2) havinginformation on the child’s developmental,medical, and psychiatric history; (3) havinginformation on multiple areas of functioning,such as information on academiccapabilities and peer relations, in additionto information on multiple areas of behavioraland emotional functioning; and (4)having information on the important contexts(e.g., home, school, work) in which achild functions. Once this information isobtained, the following multistep procedurecan be used to develop a case formulation.In Boxes 15.4 and 15.5 we providetwo case examples that illustrate this interpretiveprocess.Step 1: Document AllClinically SignificantFindings Regarding theChild’sAdjustmentThe first step of the procedure follows theeither/or rule. The assessor sifts throughall of the information on a child’s or adolescent’sadjustment from all sources andmethods and determines if there are anysignificant findings, such as norm-referencedelevations on rating scales, diagnosesbased on structured interviews, or clinicallysignificant material from projective tests.As discussed previously, this is a very sensitiveprocedure that can result in high ratesof significant findings. We feel that it is agood starting point in the interpretive process.However, we also feel that stopping atthis point would result in an unacceptablyhigh rate of false positives. The subsequentsteps help to increase the specificity of theprocess.Step 2: Look for ConvergentFindings acrossSources and MethodsAfter noting all of the significant findingsacross methods and sources, the nextstep is to isolate areas that are consistentacross the various pieces of assessmentinformation. In looking at cross-methodand cross-informant convergence, it is recommendedthat one takes a closer look atthe assessment information and does notsimply determine whether or not a scorehas crossed some threshold or elevation.We have discussed at several points in thisbook the arbitrary nature of many clinicalcut-offs, such as elevations on a scaleof a behavior rating scale or a diagnosis ona structured interview. How different isa child with a T-score of 69 from a childwith a T-score of 70 on a given scale? Howdifferent is a child with seven symptoms ofADHD from a child with eight?

350chapter 15 integrating and interpreting assessment informationBox 15.4Case Study to Illustrate the Multistep Interpretive Procedure: A 7-Year-Old Boywith Behavior ProblemsJake was a 7-year, 1-month-old boy who wasreferred to an outpatient psychological clinicfor a comprehensive psychological evaluationby his mother. His mother reported that Jakewas having great difficulty completing schoolwork and that he was extremely oppositionaland hard to manage at home. Jake was administereda comprehensive battery of tests thatincluded (1) an unstructured clinical interview;(2) an extensive family background assessment;(3) a psychoeducational assessment; (4)an assessment of his emotional and behavioralfunctioning through structured interviewsconducted with his mother and teacher andrating scales completed by his mother, histeacher, and by Jake himself; (5) a sociometricexercise conducted with his class; and (6) anobjective assessment of his classroom performance(e.g., amount of work completed, accuracyof work).Step 1: Document All ClinicallySignificant Findings on theChild’s AdjustmentOn a structured interview, the DiagnosticInterview Schedule for Children-Version4 (DISC-IV; Shaffer et al., 2000), Jake metcriteria for ADHD according to both hismother and his teacher. On the Child BehaviorChecklist (CBCL: Achenbach, 2001), hehad T-scores above 70 on both the ThoughtProblem scale and the Attention Problemscale according to the father and T-Scoresof 70 and 67 on the Attention problem andThought Problem scales, respectively, accordingto his mother’s report. On the teachercompletedComprehensive Behavior RatingScale for Children (CBRSC; Neeper, Lahey,& Frick, 1990), he obtained a T-score of 71on the Daydreams scale and a T-score of 65on the Motor Hyperactivity scale.Also on the DISC-IV, Jake met criteriafor oppositional Defiant Disorder (ODD)according to his mother’s report. Similarly, heobtained T-scores above 70 on the AggressiveBehavior scale of the CBCL by both motherand father reports.A final significant finding was the report ofa sufficient number of depressive symptomsto meet criteria for current Major Depressionaccording to maternal report on the DISC-IV.Step 2: Look for ConvergentFindings across Sources andMethodsIt was clear that there were convergent findingsfor significant problems of inattention–disorganization and impulsivity–hyperactivityassociated with ADHD, according to parentsand teachers and across structured interviewsand rating scales. However, for the oppositionalbehaviors, the teacher report gave no indicationsof these behaviors in the school environmenteither on rating scales or on structuredinterviews. For the parental report of depression,although Jake’s teacher did not reportsymptoms of depression of sufficient severityto warrant a diagnosis solely on her report, shedid endorse several symptoms, such as feelingsad and irritable, feeling bad about himself, andoften seeming as if he was about to cry. However,on the Children’s Depression Inventory(CDI; Kovacs, 1991), Jake did not report significantfeelings of depression.Step 3: Try to ExplainDiscrepanciesIn talking to Jake’s teacher, it appeared thatshe had a very structured classroom with ateacher’s aide who implemented behavioralprograms for students. Jake’s parents weredivorced, and his mother seemed quite distressedover Jake’s behavior. As a result, sheseemed to be very unsure of how to discipline(Continues)

chapter 15 integrating and interpreting assessment information351Box 15.4 (Continued)Jake, leading to a great deal of inconsistencyand sometimes harshness in her disciplineattempts. These differences in teacher andparent handling of Jake’s behavior seemed toexplain the fact that he was not showing significantoppositional behaviors at school. Whilethere was some support for the depressivebehaviors by Jake’s teacher, Jake did not reportsuch behaviors. However, given Jake’s youngage, there was some concern over his ability toaccurately report on internal feelings using thestructured CDI.Step 4: Developing a Profileand Hierarchy of ProblemAreasJake exhibited a common profile of showingboth ADHD and ODD. However, there wereseveral reasons why ADHD was consideredprimary. First, the ADHD behaviors had beenpresent throughout Jake’s life, and they werethe only behaviors shown at both home andschool. Second, Jake’s teacher reported thatJake’s grades were primarily affected by hisfailure to complete work or his hasty completionof work that led to careless errors. Thisteacher report was consistent with a psychoeducationalassessment that revealed an IQ inthe average range and academic achievementalso in an age-appropriate range. A sociometricexercise indicated that Jake was dislikedby a significant number of his classmates(5 out of a class of 17 nominated him as “likedleast”). Therefore, the ADHD was consideredto be causing a high level of impairment forJake, educationally and socially. Third, Jake’sbiological father reported that he had had significantproblems with inattention and hyperactivityas a child.Jake’s depressive features were also consideredto be secondary to the ADHD. Hismother reported that they started after thepast semester in school, when he had receivedhis worst grades ever which led to high ratesof conflict at home.Step 5: Determine CriticalInformation to Place in theReportJake’s mother had seen a counselor for 6months following her divorce from Jake’sfather, 3 years prior to the evaluation. Thispiece of information did not seem to be crucialfor understanding our case formulation ofJake. Also, because the psychoeducational testingdid not reveal many significant strengthsor weaknesses in cognitive processing or invarious areas of academic achievement, thisinformation was only briefly summarized,indicating that it was not suggestive of thepresence of a cognitive deficit or a learningdisability.Box 15.5Case Study to Illustrate Multistep Interpretive Procedure: A 14-Year-Old Boywith a Language Disorder and Social PhobiaJarrod was 14 years, 2 months old when hisschool referred him for a comprehensive psychologicalevaluation. They were concernedabout Jarrod’s inability to complete schoolworkand his extreme shyness. His teachers reportedthat Jarrod rarely initiated conversations witheither peers or teachers and responded to otherpeople with only one-word answers. Jarrod hadbeen receiving resource help at school in a learningdisabilities classroom for the past 2 years.(Continues)

352chapter 15 integrating and interpreting assessment informationBox 15.5 (Continued)Jarrod was administered a comprehensivebattery of tests that included an unstructuredclinical interview, an extensive familybackground assessment, a psychoeducationalassessment, assessment of his emotional andbehavioral functioning through structuredinterviews conducted with his mother andJarrod and rating scales completed by Jarrodand his mother.Step 1: Document All ClinicallySignificant Findings on theChild’s AdjustmentOn structured interviews, Jarrod met criteriafor Social Phobia on the DISC-IV by both hisown self-report and the report of his mother.In the psychoeducational assessment, Jarrodscored 25 points lower in his Verbal IQ thanon his Performance IQ. This verbal deficit wasconsistent with previous testing conducted byhis school. In addition to the verbal weakness,Jarrod scored poorly on tests of processingspeed.Step 2: Look for ConvergentFindings across Sources andMethodsAlthough not in the clinically significantrange, there was support for Jarrod’s socialanxiety and social withdrawal on parent- andchild-completed rating scales. On the CBCLcompleted by Jarrod’s mother, the Withdrawalscale had a T-score of 68. On the self-reportPersonality Inventory for Youth (PIY; Lachar& Gruber, 1994), Jarrod obtained a T-scoreof 65 on the Social Withdrawal scale. Thiswas consistent with reports from his teachersand from behavioral observations during testingthat indicated that Jarrod was very quietand withdrawn. The other main areas of convergencewere that Jarrod did not show significantbehavioral problems and his parentsseemed to provide a warm and stable homeenvironment.Step 3: Try to ExplainDiscrepanciesIn this case the only discrepancy was the somewhatmilder level of social withdrawal beingreported at home than was being reported byteachers at school, although clearly the shynesswas still evident at home. This was likelydue to Jarrod’s higher level of comfort withfamily members, which made him more willingto converse with family members.Step 4: Developing a Profileand Hierarchy of ProblemAreasIt seemed as though Jarrod had an expressivelanguage disorder and a deficit in processingspeed. As a result, Jarrod was likely to be uncomfortableconversing in social situations, leadingto the high rate of social withdrawal. Thelanguage disorder and processing speed deficitwere judged to be primary in this case becausethey had been chronic throughout his life andJarrod did not show a high rate of anxiety inother situations that did not require a great dealof verbal expression.Step 5: Determine CriticalInformation to Place in theReportA great deal of the information on Jarrod’sfamily background, current family functioning,and functioning in other behavioral, andemotional areas (e.g., conduct and depression)were only summarized briefly in the report.They were felt to be important to documentJarrod’s many psychological strengths, but adetailed discussion of these areas was not feltto be needed in the written report.

chapter 15 integrating and interpreting assessment information353Therefore, when looking for convergentpieces of evidence, one should lookat the full range of scores, problems, orsymptoms. For example, suppose a childreports a high level of anxiety on a selfreportquestionnaire, passing a clinical cutoffpoint as required in Step 1. However,on a parent-structured interview there areno diagnoses of anxiety disorders reported,but the parent has reported three symptomsof overanxious disorder and two symptomsof separation anxiety disorder. Further, ateacher rated the child with a T-score of64 on the anxiety scale of a behavior ratingscale. If one just viewed scores thatexceeded a clinical cut-off, this patternwould look like the child’s self-report ofanxiety was unsupported by other informants,when, in fact, we would argue thatthere is fairly consistent support for thepresence of anxious behaviors, althoughthe anxiety is not being perceived as severeby parents and teachers.Step 3: Try to ExplainDiscrepanciesThe complexity of the assessment processis dramatically reduced when assessmentinformation is consistent across methodsand sources. This is the ideal casethat is the dream of every clinical assessor.Unfortunately, existing research andclinical experience suggest that this is notlikely to happen often. In most cases, therewill be numerous discrepancies between theinformation provided by the differentsources after steps 1 and 2. At this stage,one should take the information discussedin previous sections of this chapter and tryto develop explanations for the discrepancies.Can the discrepancies be explainedby different demands in the various settingsin which a child is observed? Can thediscrepancies be explained by differencesin the measurement techniques used orby certain characteristics/motivations ofthe informants? Can the discrepanciesbe explained by differing knowledge ofthe child’s behavior across informants? Asmentioned previously, if one can answerthese questions and account for discrepantinformation, one has gone a long wayin developing a good case formulation anddeveloping goals for treatment.Step 4: Develop a Profileand Hierarchy ofStrengths and WeaknessesBoth research and clinical practice indicatethat children rarely have problems thatare specific to one area. As a result, thenext step in the interpretive process is todevelop the profile of a child’s or adolescent’sstrengths and weaknesses across thedifferent domains of psychological functioningthat have been assessed. In addition,we feel that this process needs to gobeyond simply documenting the differentareas of strengths and weaknesses, to alsoprioritizing the different areas of concern.This prioritization should be both a conceptualand practical endeavor.Conceptually, one should consider whatproblematic area may be primary andwhich areas may be secondary, with secondarybeing defined as areas that seem to belargely a result of some other primary factor.For example, for a child who becomesdepressed because of his/her frequentschool suspensions and police contacts forantisocial behavior, the depression maybest be considered secondary to the antisocialbehavior.Practically, one needs to consider whatarea should be the primary focus of intervention.This may follow closely with thedetermination of primary and secondaryproblems, where intervention targetingof primary areas (e.g., antisocial behavior)

354chapter 15 integrating and interpreting assessment informationmay also alleviate problems considered tobe secondary (e.g., depression). However,these two levels of analysis do not necessarilyhave to be congruent. For example,a child with ADHD may have developedconduct problems secondary to the ADHDbecause of the effects of the child’s behavioron parent–child interactions. However,given the current problems causedby the conduct disturbance and the relationof these problems to future problemsin adjustment, intervention targeting theconduct problems may be the primarytreatment recommendation, rather thantreatment for ADHD.Most assessors would agree with theneed to prioritize problem areas, but mostwould also agree that this is difficult toaccomplish with any degree of reliabilityand validity. We must admit that prioritizingareas of need and determining primaryand secondary areas of dysfunction areoften largely made on the basis of clinicaljudgment in the absence of a clear researchbase to guide this process. However, wefeel that there are three primary factorsto consider in prioritizing areas within achild’s profile.First, one should look at the degree ofimpairment associated with different areasof dysfunction. As a standard part of anyassessment, one should measure the degreeto which various behaviors are affecting achild’s functioning in at least three majorlife areas: in school, with peers, and athome. Then several questions can be askedin an effort to determine if some problemareas are contributing more to impairmentthan others. Are the problematic areas differentiallyaffecting a child’s ability to learnor to perform well academically? Has oneor more of the areas significantly affecteda child’s ability to form or maintain meaningfulrelationships with same-age peers? Isany problem area causing a higher level ofconflict with parents and/or siblings thanothers? This method of prioritization isone of the most important factors clinically,because areas that cause the greatest degreeof impairment typically are the most importanttargets of interventions.Second, one can look at the temporalsequencing of when problem behaviorsdeveloped to help to determine what mightbe primary or secondary. Did symptoms ofdepression develop after a chronic patternof antisocial behavior? A good example ofthe therapeutic implications of temporalsequencing comes from a study in whichchildren with major depression who onlyshowed conduct problems after the onsetof the depression showed a significantreduction of conduct problems after thedepression was successfully treated (Puig-Antich et al., 1978).Third, viewing family history data canhelp in making a determination of primaryand secondary areas of disturbance. As wediscussed in Chap. 12, in the section onassessing a child’s family history, informationon parental psychiatric history canoften aid in making differential diagnoses.This is largely due to the fact that (1)adult psychological disturbance tends to bemore easily defined and (2) many forms ofchildhood disturbance show a familial link.As a result, assessing parental adjustmentpatterns can provide clues as to what is theprimary area of dysfunction in their childor adolescent.Step 5: DetermineCritical Information toPlace in the ReportIn the next chapter we provide a detaileddiscussion of report writing. However, itis important to note that the last stage ofintegrating information is a filtering process.After one has developed a profile ofthe child that explains the referral problemsand suggests treatment goals, it is likely thatmany pieces of information gathered will

chapter 15 integrating and interpreting assessment information355have either minimal or no bearing on thefinal case conceptualization. A good assessorwill limit the amount of unessentialinformation that is collected in an evaluationby carefully tailoring the evaluation tothe needs of an individual child. However,much information cannot be determined tobe irrelevant until after it is collected.Many clinicians feel that, if informationwas collected, it should be included in thewritten and/or oral report of the assessmentresults. Our feeling is that too muchirrelevant information detracts from thecase formulation, making it unduly confusing.Therefore, we feel that the final stepin the interpretive process is determiningwhat information is essential for understandingthe case formulation and subsequentrecommendations for treatmentand what information should be omittedor discussed only minimally in the report.This final step sets the stage for the writingof a clear, concise, and readable summaryof the assessment results.ConclusionsIn this chapter we discuss some of the issuesinvolved in integrating assessment informationacross informants, across methods,and across psychological domains. One ofthe more difficult aspects of this endeavoris integrating discrepant information fromdifferent informants on the same dimensionof functioning. We discussed evidencethat simple decision rules that equallyweigh information given by various informantsmay be as good as or better thanmore complex methods of differentiallyweighing information provided by variousinformants. However, this simple ruleneeds to be qualified by whether the informantis knowledgeable about the child’sbehavior and whether the informant canreport the information effectively. Thesetwo factors can be influenced by manyfactors, such as age of the child and thespecific behavioral domain being assessed.We conclude the chapter by providing amulti-step procedure for integrating thevarious pieces information from a comprehensiveevaluation into a clear caseformulation that answers the referral questionand points the way to a recommendedcourse of intervention.Chapter Summary1. Because most clinical assessments of childrenand adolescents must be comprehensive,one of the most difficult partsof the assessment process is integratingassessment information into a clear caseformulation that points the way to themost appropriate interventions.2. The most difficult task is integratinginformation on the same behavior providedby different informants. Researchindicates that poor agreement amonginformants is the rule, rather than theexception.(a) These discrepancies may reflect realdifferences in a child’s functioningin different contexts (e.g., home,school, clinic). Therefore, assessingthe characteristics of the differentcontexts can help to explain somediscrepancies.(b) Some discrepancies may be an artifactof the level of analysis. Forexample, individual symptoms maybe very discrepant across informants,whereas the presence of a disordermay show greater concordance.(c) The level of agreement can berelated to certain characteristics ofthe behaviors being assessed, suchas the salience of the behavior tothe informant, how observable thebehavior is to others, and how willingthe informant is to report on thebehavior.

356chapter 15 integrating and interpreting assessment information(d) Discrepancies can also be an artifactof different measurement techniquesused to obtain information from differentinformants.3. Unfortunately, in many cases it is difficultto attribute discrepant informationsolely to varying situational demands, tothe level of analysis, to the type of behaviorbeing assessed, or to differences inassessment strategy. In these cases thereis evidence that a simple either/or systemfor combining information worksbest. That is, a finding is considered significantif reported by any informant.(a) This approach assumes that falsepositives are rarer and less harmfulthan false negatives.(b) This approach also assumes thatinformants are only asked to provideinformation that they should beexpected to know.4. The quality of information providedby different informants seems to varydepending on the domain of functioningbeing assessed and the age of thechild being assessed.(a) Research suggests that teacherreport of inattentive-hyperactivebehaviors, parent report of conductproblems, and child report of emotionalproblems seem to be relativelymost important in preadolescentchildren.(b) The usefulness of child self-reportincreases with age and the usefulnessof teacher report decreases withage. Age-related changes in parentalreport are less consistently found,although there are some indicationsthat the usefulness of parental reportalso decreases somewhat as the childenters adolescence.5. Other factors, such as a parent’s emotionaladjustment, the degree of maritalconflict, idiosyncratic motivations of aninformant, and testing conditions, caninfluence informant agreement.6. A multistep procedure for integratingdiverse assessment information is recommended.Step 1: Document all clinically significantfindings on the child’sadjustment.Step 2: Look for convergence acrosssources and methods.Step 3: Attempt to explain discrepancies.Step 4: Develop a profile and hierarchyof findings.Step 5: Determine what critical informationto place in the formalreport.

C h a p t e r 1 6Report WritingChapter Questionsl How does the call for evidence-basedassessment lend itself to report writing?l How does a clinician clarify the referralquestion?l What are the common mistakes madein report writing?l How should assessment results bereported to parents in conferences?Reporting Problemsand ChallengesPresenting assessment results orally or inwriting can be a foreboding task. However,this process is central to assessment. Themost sophisticated, accurate, and comprehensivecase conceptualization is uselessif the key figures in a child’s life (e.g., parents,teachers) are unaware of the results orrecommendations from an assessment ordo not fully understand them. Aside fromappropriately and accurately conveyingresults and recommendations, clinicians arealso often faced with fears of litigation andinsecurities about their interpretive skill.Thus, a chapter on report writing is crucialfor an assessment text. As clinicians knowwell, their written products can carry a greatdeal of importance, and if done well, theycan facilitate positive outcomes for a child.On the other hand, if a report is faulty (i.e.,inaccurate, unclear, full of errors), it willmake no impact aside from being a negativereflection on the clinician who wrote it.Effective psychological report writing istaking on increased importance for practicingpsychologists. Psychological reports aremade available to parents, judges, lawyers,357P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_16, © Springer Science+Business Media, LLC 2010

358 chapter 16 report writingand other non-psychologists, creating theopportunity for improper interpretation ofthe results by untrained individuals. Morepositively, psychological reports remainparticularly useful to other clinicians whoevaluate a child who has previously beenseen by a psychologist. A previous psychologicalreport can provide a valuable baselineagainst which a clinician can gaugeresponse to treatment, the emergence ofa comorbid problem, and other factors. Aprevious diagnosis of Conduct Disorder,for example, may encourage the evaluatingpsychologist to screen for depressionbecause of known comorbidity (DeBaryshe,Patterson, & Capaldi, 1993). A cliniciancan significantly enhance the quality ofwork conducted by a successor through theproduction of an articulate written report.Despite its importance, the topic ofreport writing is relatively neglected in theresearch literature (Ownby & Wallbrown,1986). While a number of works are availableon this topic (e.g., Braaten, 2007; Lichtenberger,Mather, Kaufman, & Kaufman,2004; Tallent, 1993), little research has beenconducted to assess the effects of reportwriting on important outcomes such as thelikelihood that a recommendation will befollowed (Ownby & Wallbrown, 1986).Ownby and Wallbrown (1986) draw severaldiscouraging conclusions. They concludeas follows on psychological reports:l Considered useful to some extent byconsumers such as psychiatrists andsocial workersl Frequently criticized by these professionalgroups on both content and stylisticgroundsl May (or may not) make substantial contributionsto patient managementIn addition to the opinions of psychiatristsand social workers, a number of studies haveassessed teachers’ satisfaction with psychologicalreports and have found that they arefrequently dissatisfied with them (Ownby& Wallbrown, 1986). One can get a senseof why teachers and other professionals aredissatisfied with psychological reports byreading the following excerpt that was takenverbatim from a report. All of the conclusionsdrawn by the evaluator in this case areon the basis of one test requiring the child tosimply reproduce nine designs with penciland paper.We quote:The Bender-Visual Motor Gestalt test suggestsdelinquency and an acting out potential. He isanxious, confused, insecure and has a low selfesteem.He may have difficulties in interpersonalrelationships and tends to isolate himself whenproblems arise…. [He] also seems to have a lotof anxiety and tension over phallic sexuality andmay be in somewhat of a homosexual panic.This clinician was apparently using a cookbookapproach to interpretation, conveyingno clear evidence to support his/hercase conceptualization. A report like this isof no help to anyone, especially not to thechild being evaluated.One of the difficulties with report writingis that different audiences require differentreports. For example, a psychometricsummary (i.e., a portion of the report thatpresents only test scores and is usually givenat the beginning or as an appendix at the endof a report) given without context is likely tobe of little use to parents but of great potentialuse to colleagues and perhaps teachers.An important decision that each psychologistmust make prior to report writing isto determine the primary audience for thereport. For example, a psychometric summarymay be of minimal use to parentswho have contracted with the psychologistin private practice for an evaluation. Inthis case, it is more sensible to present testresults in context in order to communicateeffectively with the parents. A psychometricsummary is more in order in a treatmentteam situation, where it is imperative that apsychologist communicate effectively withknowledgeable colleagues.

chapter 16 report writing359In most situations, we recommend thatthe clinician attempt to make his or herreports accessible and useful to all pertinentaudiences so that interested parties do nothave different reports for the same assessmentof the same child. Our discussion willfocus mainly on the expectation that onereport will be made available to parents,teachers, physicians, etc., with the understandingthat certain presentations of resultsand interpretations will be most useful tocertain audiences.Report Writing as Part ofEvidence-Based AssessmentReport writing has not specifically beenaddressed in recent writings on evidencebasedassessment of children. Instead, thediscussion has focused on the use of toolsand methods that are valid and that demonstrateclinical utility (see Mash & Hunsley,2005). We feel that the move towardevidence-based assessment should, andwill, be reflected in the reports that resultfrom psychological assessments.Mash and Hunsley (2005) point out thatevidence-based assessment is not meant toreplace the clinician or the clinician’s judgment.Similarly, the clinician will continue tobe a key figure in assessment reports. That is,it is unlikely and undesirable that reports willbe completely boiler plate endeavors thatdo not allow for flexibility based on the particularassessment approach used, the client’sparticular presenting problems, or the needsof the client and allied professionals.In contrast, not unlike a scientific manuscript,reports from an evidence-basedapproach to assessment can be seen as themeans by which a client’s history and difficultiesare described, results are obtainedand interpreted, and suggestions for future(treatment) approaches to the difficultiesare discussed.The main sections of most psychologicalreports are discussed later in this chapter, butin many ways, they are analogous to sectionsof journal articles in psychology and otherscientific fields. For example, the “ReferralQuestions” section is essentially a statementof the evaluation’s purpose. “BackgroundInformation” in a report is similar to a literaturereview in a research article, whereinthe previously noted issues are mentionedand the current questions or problems arepresented to the audience. The “AssessmentProcedures” or Psychometric Summary providesthe methods used in trying to addressthe referral question(s). The results of thecurrent assessment begin to be addressedfully in the “Behavioral Observations” section.This section provides a context for theassessment results, particularly any testingthat occurred directly with the client. Theanalogy in a scientific manuscript would beinitial analyses that point to any variablesthat need to be controlled or any conditionsthat might call some results into question.Similarly, testing conditions or client factors(e.g., child was sick on the day of testing)could be important information for interpretingassessment results. The “AssessmentResults and Interpretation” sectionsare ideally a mix of what might be found inthe “Results” and “Discussion” sections ofa scientific article. In the report, the clinicianshould not present data with no interpretation,and the clinician should not makeinterpretations without clearly providingthe data on which they were made. Recommendations,which are critical in an assessmentreport, allow the clinician to suggestwhat should be done in the future to addressthe problem. Researchers routinely do thisas well in their published manuscripts.Pointing to consistencies between assessmentreports and scientific manuscripts isan oversimplification of the report writingprocess in some ways. Nevertheless, manyclinicians-in-training are also well-versedin research methodology and writing, andthis analogy may serve to make reportwriting seem less nebulous and daunting.The collection of background information,the scoring of measures, and inter-

360 chapter 16 report writingpretation of results that occur in researchare essentially the processes that take placein evidence-based assessment and reportwriting. The gathering and explaining ofevidence allow the clinician to clearly presenta case conceptualization (theory) of theclient’s difficulties that is grounded in data,as opposed to the approach evidenced inthe quote earlier in this chapter.Pitfalls of Report WritingComplaints about psychological reports persist.Norman Tallent (1993) wrote a landmarktextbook on report writing in which hesummarized the literature on the strengthsand weaknesses of reports as identified bypsychologists’ colleagues in mental healthcare, most notably social workers and psychiatrists.Some of the highlights of Tallent’sreview are outlined in the next section.Vocabulary ProblemsThe problem of using vague or impreciselanguage in report writing is commonplace.The colloquial term used to describe suchlanguage is psychobabble. Siskind (1967), forexample, studied the level of agreementbetween psychologists and psychiatrists indefining words such as the following:AbstractAffectiveAggressionAnxietyBizarreBrightCompulsiveControlDefenseDependentDepressiveEmotionalHostilityImmaturityImpulsiveNormal ConstrictionThe results of the study showed very littlecorrespondence between the definitionsoffered by the two groups of professionals.We suspect that without clear descriptorsof problems (e.g., hyperactivity, depression),such disagreement among consumersof reports persists today.Tallent (1993) refers to one aspect of thisproblem with language as exhibitionism,which seems to be a frequent criticism ofreports, particularly on the part of other psychologists.One commentator stated, “Theyare written in stilted psychological terms toboost the ego of the psychologist” (p. 33).Some other pertinent observations byTallent (1993) on the use of language by psychologistsin reports are paraphrased below:l They include complex (meaningless)words that are often used to add lengthto the report.l They are written in esoteric languageunderstood by the psychologist only.For example, it may be said that clientsmanifest overt aggressive hostility in animpulsive manner – when, in fact, theypunch you on the nose.l They are not frequently enough writtenin lay language. In particular, scores areover- emphasized, and the fit between theresults and the child’s actual behavior isunder-emphasized. Tallent (1993) arguesthat an excessive focus on multiple scoresor indices may be a method to cover upthe clinician’s lack of true understandingof the assessment findings.l They include language that is so vagueand unclear that it cannot be falsified orconsidered wrong.These latter two points are critical if reportsare to address the referral question in a mannerthat is amenable to subsequent, appropriateintervention. Of course, psychologycannot be singled out as the only professionwith a preference for its own idiosyncraticterminology, as anyone who reads a physician’sreport or a legal contract will attest.Perhaps psychologists can, however, lead theway toward competent reporting of findings.

chapter 16 report writing361Faulty InterpretationFaulty interpretations may be made on thebasis of personal ideas, biases, and idiosyncrasies(Tallent, 1993). The problem ismost readily seen when the psychologistis clearly using the same theories or drawingthe same conclusions in every report. Apsychologist may conclude that all children’sproblems are due to poor ego functioning,neuropsychological problems, orfamily system failure. A psychologist whoadheres exclusively to behavioral principles,for example, will attribute all childproblems to faulty reinforcement histories.The savvy consumer of this psychologist’sreports will eventually become wary of thepsychologist’s conclusions, as the relevanceof the favored theory to some cases is questionable.One can imagine the skepticismthat may be engendered by a psychologistwho concludes that a child whose schoolperformance has just deteriorated subsequentto a traumatic head injury merelyneeds more positive reinforcement to bringhis grades up to pre-trauma levels.Problems may also occur if a psychologistdraws conclusions that are clearly in conflictwith the data collected for a child. A psychologistmay decide not to make a diagnosis,in seeming contrast to rating scale findingsof significant T-scores on the majority ofscales. If a clear argument for resolving thisincongruity is not made, the consumer of thereport may well suspect biases. The psychologistwho routinely does not reconcile highT-scores with a lack of a diagnosis may soonbe labeled as unwilling to diagnose regardlessof assessment results. The reverse situationcan also be problematic, wherein thepsychologist makes a diagnosis without anyclear indications of significant symptomatologyor impairment. Teachers, pediatricians,or other referral sources who receive thisinterpretation consistently from the samepsychologist may eventually pay more attentionto the data presented in the reports andignore the psychologist’s conclusions, or theymay simply refer elsewhere.Report LengthPsychologists, more so than other groups,complain about the excessive length of reports(Tallent, 1993). However, length may not bethe real issue. Perhaps long reports are usedto disguise incompetence, fulfill needs foraccountability, or impress others. The possibilitythat length is a cover for other ills isoffered in the following example:A business executive likes to relate the anecdoteabout the occasion when he assigned anew employee to prepare a report for him. Indue time, a voluminous piece of writing wasreturned. Dismayed, the executive pointed outthat the required information could be presentedon one, certainly not more than two, pages.But sir, pleaded the young man, I don’t knowthat much about the matter you assigned me to(Tallent, 1988, p. 72).It may also be worth considering that theTen Commandments are expressed in 297words, the Declaration of Independence isin 300 words, and the Gettysburg Addressis in 266 words.Number ObsessionThe clinician must always keep clearly inmind that the child is the lodestar of theevaluation, and the numbers obtainedfrom personality tests and the like are onlyworthy of emphasis if they contribute tothe understanding of the child being evaluated.One way to think of the scores is asa means to an end, with the end being betterunderstanding of the child. The samenumbers for two children can mean two quitedifferent things. Just as a high temperaturereading can be symptomatic of a host ofdisorders from influenza to appendicitis,so, too, a pathognomic behavioral sign canreveal a host of possible conditions.One horrendous error often madewhen reporting test scores is a psychologistreporting a score and then saying thatit is invalid. Then why report it (Tallent,

362 chapter 16 report writing1993)? If a test score is invalid, how doesit serve the child to have this score as partof a permanent record? Reporting apparentlyinvalid scores is akin to a physicianmaking a diagnostic decision on the basisof a fasting blood test when the patientviolated the fasting requirements. In alllikelihood, the flawed results would notbe reported; rather, the patient would berequired to retake the test. We suggestthat one does not have to report scores for a testjust because it was administered. This stanceapplies to scores that are deemed invalid orcircumstances in which the psychometricsunderlying the scores are questionable. Inthese situations, disregarding the informationfrom the measure or providing onlydescriptions of the responses may betterinform case conceptualization.Failure to Address ReferralQuestionsTallent (1993) points out that psychologiststoo often fail to demand clear referral questions,and as a result, their reports appearvague and unfocused. This very obviouspoint is all too frequently overlooked.Psychologists should insist that referralsources present their questions clearly, andif not, the psychologist should meet withthe referring person to obtain further detailon the type of information that is expectedfrom the evaluation (Tallent, 1993). Manyagencies use referral forms to assist in thisprocess of declaring assessment goals. Aform similar to those used by hospitals isshown in Fig. 16.1, and one suitable for useby school systems is given in Fig. 16.2.On occasion, the referral question(s)can be insidious and, consequently, placethe psychologist in the position of disappointingthe referral source before theevaluation is even initiated. Under thesecircumstances, the psychologist may feelhelpless or even betrayed because of thenegative reaction of the referral sourceto the presentation of results and recommendations.Psychologists may often needto pursue the true referral question. Someexamples of stated and true referral questionsare shown below:Stated ReferralQuestionA child’s teacherwants to know ifchild has ADHDA parent wantsto know why achild is failing inschoolA psychiatristwants to knowif a child isdepressedA psychologistwants to knowif the child isneurologicallyimpairedTrue Referral QuestionThe teacher is convincedthat the child has ADHDand expects the psychologistto confirm itThe parent thinks thechild is depressed andwould like her to be onmedicationThe psychiatrist has madethe diagnosis ofdepression and has placedthe child on medication.The referral was madesimply because a secondopinion is required forreimbursement purposesThe psychologist isseeking a diagnosis oftraumatic brain injury inorder to bolster her courttestimonyIn all of these scenarios, it wouldbehoove the psychologist to clearly determinethe referral source’s actual needs and/or desires early on in the referral/evaluationprocess and then determine the mostappropriate way to proceed.The Consumer’s ViewVirtually no recent research has beenconducted on the consumer’s view of psychologicalreports. One study evaluatedteacher preferences for and comprehensionof varying report formats (Wiener, 1985).This study required a group of elementaryschool teachers to read and rate their comprehensionof and preferences for threedifferent reports for the same child.

chapter 16 report writing363Patient Name John Doe __________________________________________________________________Medical Record Number 00071103 ____________________ _____________________________________Attending Physician LymanType of Consultation:Patient is a 13 year-old with Type I Diabetes who has poor adherence to treatment regimen. Parents areconcerned that John is aware of the risks of th is poor adherence but seems apathetic. Patien t’s affect isflat and may be de pressed. Pati ent to be discharg ed fr om hospital following psych. consul tation.Results of Consultation: Patient appears depressed and seems knowledgeable about diabetesand his diabetes regimen. In particular, his parents noted that he appears sad most of the time,lacks energy, has reduced his contact with friends, and does not seem interested in activities thathe used to enjoy. Rating scales completed by patient and his mother showed moderate levels ofdepression. Family history of depression is significant. Outpatient therapy is recommended andhas been scheduled to begin in 1 week.Signed _____________________________________Title ______________________ DateFigure 16.1Sample referral form consultation used by hospitalsThe three reports used were a shortform, a psychoeducational report, and aquestion-and-answer format. The shortformreport was one page, single-spaced.It used some jargon, such as acronyms, toshorten length; conclusions were drawnwithout reference to a data source; andrecommendations were given withoutelaboration. The psychoeducationalreport format was three and a half single-spacedpages. It used headings suchas Reason for Referral, Learning Style,Mathematics, Conclusions, and Recommendations.Observations were statedin behavioral terms with examples usedfreely. Recommendations were givenand elaborated, and acronyms and otherjargon were only used when they weredefined in text. The question-and-answerreport was similar to the psychoeducationalreport in many ways, but it did notuse headings per se. This report listedreferral questions and then answeredeach question in turn. This report wasfour and a half pages long.Amazingly, in this study, the participantspreferred length. First, teacherscomprehended the two longer reportsbetter. Second, of the two longer reports,the teachers preferred the question-and-

364 chapter 16 report writingStudent's Name Jane Smith_____________________________ Date of Referral 10/11/07Referring School Stuart Elem.______ Age 8 _ Grade 2 _____ Grades Repeated N/AIs the student now receiving speech therapy? ___ Yes x noNever Sometimes OftenExpressive Language (problems in grammar, limited vocabulary)Receptive Language (comprehension not following directions)Speech (poor enunciation, lisps, stutters, omits sounds, infantile speech)xxxGross Motor Coordination (eye-hand, manual dexterity)xVisual (cannot see blackboard, squints, rubs eyes, holds book too close) xHearing (unable to discriminate sounds, asks to have instructionsrepeated, turns ear to speaker, often has earaches)Health (example: epilepsy, respiratory problems, etc.)xxMedications (yes) (no) (Type)Overly energetic, talks out, out of seat: SometimesVery quiet, uncommunicative: OftenActing out (aggressive, hostile, rebellious, destructive, cries easily): SometimesInattentive (short attention span, poor on task behavior): OftenDoesn ' t appear to notice what is happening in the immediateenvironment:SometimesPoor Peer Relationships (few friends, rejected, ignored, abused by peers): OftenACADEMIC PROBLEMSReading (word attack, comprehension): OftenWriting (illegible, reverses letters, doesn't write): OftenSpelling (cannot spell phonetically, omits or adds letters): OftenMathematics (computation, concepts, application): NeverSocial Science, Sciences (doesn't handle concepts, doesn't understandrerelationships, poor understanding of cause and effect): NeverSignature and position of referring personFigure 16.2Referral for consultation used by schoolsanswer report over the psychoeducationalreport. The short form was clearly leastpreferred. These are intriguing results inthat they hint that length may be overratedas a problem in report writing and thatteachers may prefer a question-and-answerreport format. This finding is interestingbecause this format is rarely used in reportsfrom clinical assessments.Do parents have different preferencesfrom teachers? In a follow-up study withparents using the same methodology, Wienerand Kohler (1986) found that teachersand parents have similar preferences. Inthis second study, the same three reportformats were used. As was the case withteachers, parents comprehended the twolonger reports significantly better than

chapter 16 report writing365the short-form report. An interestingadditional finding was that parents with acollege education comprehended reportsbetter than parents with only a high schooldiploma. Parents also tended to prefer thequestion-and-answer format to the othertwo formats, although the difference inpreference scores between the psychoeducationaland question-and-answer reportsfailed to reach statistical significance.The results of these two studies suggestthat the two most frequent consumers ofchild and adolescent psychological reports,parents and teachers, consider the clarityof reports to be more important than theirabsolute length. They also show a preferencefor reports that have referral questionsas their focus. Cognizance of thesetwo findings may benefit psychologists whowrite reports for children and adolescents.Suggested PracticesReport Only PertinentInformationOne of the most difficult decisions to makewhen writing a report involves gauging therelevance of information to include (Teglasi,1983). Clinicians happen onto a great deal ofinformation during the course of an evaluation,some of which is tangential. Say, forexample, a child is referred for an evaluationof ADHD. During the course of an interviewwith the child’s father, he recounts at lengthhis disappointment with his wife. He tells theclinician that she is dating other men, and hebelieves that she is not spending adequatetime with their children. When writing thereport on this case, the clinician has to determinewhether or not this information is pertinentto the ADHD evaluation.Clinicians must think critically aboutthe in-formation that they include inreports and consider its relevance to thecase. As discussed by Lichtenberger et al.(2004), the objectives of psychologicalreports are to “answer the referral questions;describe the person; organize thedata; and recommend interventions” (p. 3).If information is not relevant to theseobjectives, and it is very personal, the psychologistshould consider carefully thedecision to invade a family’s privacy byincluding such information in the report.Define Abbreviations andAcronymsAcronyms are part of the idiosyncraticlanguage of psychological assessment.They can greatly facilitate communicationamong psychologists, but theyhinder communication with non-psychologists.Psychologists, just like otherprofessionals, need to use nontechnicallanguage to communicate with parents,teachers, and other colleagues in themental health field. A pediatrician wouldnot ask a mother if her child had an emesis;rather, the physician would inquirewhether or not the child vomited.When writing a report, psychologistsshould limit their use of acronymsand should define any acronym used in areport. Use of the acronym SAD for separationanxiety disorder, for example, withoutdefining, is questionable practice.Emphasize Words Ratherthan NumbersParticularly in the test results section of areport, clinicians must resist a temptationto focus exclusively on numbers (i.e., testscores). Lichtenberger et al. noted that“some evaluators spend too much timewriting about the obtained test scoresrather than about what these scores mean”(p. 5). Words often communicate moreeffectively than numbers because theycommunicate more directly and in a moreaccessible manner to a variety of audiences.The typical question of a referral source has

366 chapter 16 report writingnothing to do with the obtained T-scoresbut, rather, the psychologist’s interpretationof these scores. Most laypersons willnot understand the T-score metric but canmore easily grasp clinician’s interpretationof norm-referenced scores.Reduce Difficult WordsThe issue of using simple language isby now obvious. The difficult part forreport writers is following through onthis advice. Consider the following twoparagraphs, which differ greatly. The firstexcerpt uses vocabulary that is unnecessarilycomplex for most consumers ofreports. The second example is a rewriteof the first paragraph that uses a morepractical vocabulary level.There is also evidence from the test data to suggestthat Pam is obdurate in response to anxiety.She may also tend to be very concrete and notnotice some of the subtleties of interpersonaldiscourse. Given these idiosyncrasies, shemay find it difficult to generate effective socialproblem-solving strategies and mechanisms forcoping with life’s stressors.The next paragraph tries to communicatemore clearly by using, among other things,simpler language.Pam responds to stress by withdrawing fromothers (e.g., going to her room or leaving agroup of friends on a social outing), whichseems to be the only method she uses for dealingwith stress. She also has trouble understandingand responding to messages given byothers in social situations (e.g., body languageor verbal hints). Because of these behavior patterns,Pam has trouble making friends.Related to the use of difficult words is theissue of using the correct person. We haveoccasionally seen reports where insteadof using the child’s name, he or she wasreferred to as “the child” or “the subject.”This usage sounds too mechanistic andimpersonal for a psychological report.In most cases, the use of the child’s name isbetter. It is important to also clearly differentiateamong sources of information andbetween data gathered during the assessmentand the clinician’s interpretations.Jargon or convoluted writing makes theseimportant distinctions difficult to be madeout by the reader.Briefly Describe theInstruments UsedIn many cases, it is safe to assume that thereader of the report has little knowledge ofthe tests being used. When practical, we suggestthat report writers describe the natureof the assessment devices being used.The naive reader of a report will also behelped by descriptors of the nature of a scaleor subscale that is being discussed. Thisobservation is particularly true for scalesthat are not adequately described by theirnames. Depression scales are a good exampleof scales that may be perceived inappropriately.The label depression could conjureup a variety of images in a report reader’smind including the image of a child that isincapacitated by sadness. It may well be thata Depression score indicates significancebut, de-pending on the items endorsed, maynot warrant the formal diagnosis of depression.In this case, the clinician should try todescribe the nature of the scale content and/or its interpretive meaning in order to discouragemisuse of results.Edit the ReportWe have found that a number of ourstudents do not take a critical eye towardediting their own work and not just interms of grammar and spelling. Editingis necessary to ensure the most accuratecommunication in the least amount of space.

chapter 16 report writing367Tallent (1988) provides the followingexcellent example of how an editor (andthe articulate psychologist) thinks:There is the tale of the young man who went intothe fish business. He rented a store, erected asign, FRESH FISH SOLD HERE, and acquiredmerchandise. As he was standing back admiringhis market and his sign, a friend happened along.Following congratulations, the friend gazed at thesign and read aloud, FRESH FISH SOLD HERE.Of course it’s here. You wouldn’t sell it elsewhere,would you? Impressed with such astuteness, theyoung man painted over the obviously superfluousword. The next helpful comment had to dowith the word sold. You aren’t giving it away.Again impressed, he eliminated the useless word.Seemingly that was it, but the critic then focusedon the word fresh. You wouldn’t sell stale fish,would you? Once more our hero bowed to thestrength of logic. But finally he was relieved thathe had a logic-tight sign for his business; FISH.His ever alert friend, however, audibly sniffingthe air for effect, made a final observation: Youdon’t need a sign (p. 88).Psychologists do not need to engage in suchsevere editing, but they should at least makean attempt to think critically about their wordusage in order to reduce report, sentence,and paragraph length. Judicious editing cango a long way toward clarifying meaning in areport. Sometimes new clinicians are not usedto critiquing their own writing. One readilyavailable option is to have a colleague readreports. Confidentiality, however, should bekept in mind if an editor is used.Use Headings and Lists FreelyHeadings and lists can enhance the clarityof communication (Harvey, 1989). If,for example, a clinician draws a number ofconclusions about a child, the conclusionscan sometimes lose their impact if they areembedded in paragraphs.As one would predict, the use of headingsand lists to excess has a downside. A reportthat uses too many lists, for example, appearsstilted, and it may not communicate all ofthe texture and subtleties of the child’s performance.Report writers should considerusing additional headings if a section of theirreport stretches for nearly a page (singlespaced)without a heading. Clinicians shouldconsider lists if they want to add impact tostatements and/or conclusions.Use Examples of Behaviorto Clarify MeaningBecause there is some disagreement regardingthe meanings of particular words,report writers should clarify their meaningin order to ensure accuracy. Words thatmay conjure up a variety of interpretationsinclude anxiety, cooperation, dependent, hyperactive,and low self-esteem. One way to fosterclarity is to use examples (i.e., behavioralreferents) of the child’s behavior. Here, forexample, are two ways to say that a child,Emilio, was anxious.Emilio exhibited considerable anxiety duringthe testing.Or, alternatively:Emilio appeared anxious during the testing. Hefrequently asked whether or not he had solvedan item correctly. He occasionally looked atthe ticking stopwatch during an item and thenhurried, and his face became flushed when itwas obvious to him that he did not know theanswer to a question.An additional benefit of using examplesof behavior generously is that it forces thepsychologist to consider the extent of supportingevidence for a conclusion abouta child’s behavior. If a psychologist writesthat a child is anxious but cannot think ofbehaviors to help explain this, then theconclusion should not be drawn, as it isinsupportable by evidence.Direct quotes, to some extent, are alsohelpful for clarifying meaning. If a clinician

368 chapter 16 report writingconcludes that an adolescent is suicidal, aquote from the child may help clarify thisstatement considerably. The child mayhave said, “I thought about taking somepills once” or “I feel like I want to run outin front of a car tonight and if that doesn’twork, I will steal my father’s gun and killmyself.” These statements convey varyingdegrees of suicidal intent that are mostclearly differentiated by quotes.Reduce Report LengthTallent (1988) gives the following instancesas indications of undue length:l The psychologist is concerned that ittook too long to write it (we might addthat most reports will seem that way forbeginning clinicians, but the time to writereports should decline with experience).l The psychologist has difficulty organizingall of the details for presentation.l Some of the content is not clear or useful.l The detail is much greater than can beput to good use.l Speculations are presented without agood rationale for them.l The writing is unnecessarily repetitious.l The organization is not tight.l The reader is irritated by the length orreads only a few sections such as the Summaryor Recommendations sections.The issue of length is primarily a concernof other psychologists, and it is intertwinedwith other issues, such as clarity. Hence,the psychologist in training should notassume that shorter is better. Quality maybe a more important issue than quantity.At this early point in training, the newreport writer should keep the issue oflength in mind while writing reports. Concernsabout length, however, should neverinterfere with the need to portray a child’sperformance accurately.Check ScoresAn all-too-frequent and grievous error is tore-port scores that are incorrect. Computerizedscoring represents a breakthroughthat limits errors. In fact, if the facilities areavailable, we suggest that each test protocolthat is scored by hand be checked againstcomputer scoring. If this is not possible,the test scores should at least be doublecheckedprior to finalizing a report.One way of checking scores is to bealert to in-consistencies. If, for example, anadjudicated adolescent who was referredfor conduct problems obtains an elevatedT-score on depression measures and no elevationson conduct problem scales, then thescore should be double-checked to see if ascoring error is the source of the incongruity.If a score doesn’t seem sensible, then theclinician should always check for a scoringerror in order to rule out this possibility.Check Grammar and SpellingAnother problem with reports that detractsfrom the credibility of the clinician is thepresence of spelling errors. Clinicians arestrongly advised to take the time to electronicallyand visually check their spellingand grammar.Adapting Reports toAudience and SettingThere is probably no optimal report format.Psychologists often find that theyhave to adapt their reports to meet theneeds of an ever-changing audience. Audienceshave varying characteristics, such asliteracy levels, and, more importantly, theyhave differing referral questions.In a school setting, many referrals are forlearning problems. Teachers may also beseeking information to assist them in curriculumdecisions. These are very different

chapter 16 report writing369referral questions than those that may be ofinterest in other settings. In a psychiatrichospital setting, issues such as suicide potential,safety, and coping strategies may be ofgreater concern. These questions are verydifferent than those of the school setting,requiring a focus on topics such as diagnosisand implications for pharmacological treatment.Parents are yet another audience withspecific questions. When conducting anevaluation for parents in a private practicesetting, the emphasis may be on advisingthe parents on what they can do to effectchange in their child’s behavior.The report excerpts used throughoutthis book were taken from a variety of settingswith differing referral questions. Thereader is advised to think carefully aboutthe needs of referral sources when readingthese examples and writing reports.Referral QuestionsThis section is crucial because the referralquestions dictate the design of the evaluation.This section is often brief but shouldbe descriptive so that the purpose of theevaluation is clear. The referral sourceshould also be stated (Lichtenberger et al.,2004). The lack of clear referral questionsmay lead to consumer or referral sourcedissatisfaction with the report. As notedpreviously, psychologists may have to speakmore than once with the referral source toclarify the nature of the question(s). Thereferral questions should be stated in termsof specific examples of the child’s difficultiesrather than general labels (e.g., “hyperactivity,”“academic problems,” “anxiety”).This section may also indicate (briefly) theduration, severity, and/or frequency of theproblem.The Sectionsof the PsychologicalReportIdentifying InformationMost report formats provide some identifyingin-formation on the top of thefirst page of the re-port. This section caninclude information such as name of thechild, age, grade, birth date, and perhapsthe name of the school or agency wherethe child is currently attending or beingserved. Also, most reports indicate that thereport con-tent is confidential.Assessment ProceduresThis section typically lists the assessmentmethods (both quantitative and qualitative)and tests that were used in the evaluation.Evaluation procedures can, and frequentlydo, include inter-views, reviews of records,and classroom or other observations.Background InformationThis section should include all of the pertinentinformation that may affect interpretationof a child’s scores. The key word hereis pertinent. The clinician should report onlyinformation that is relevant to the currentevaluation, not in-formation that is superfluousor an undue invasion of privacy (Teglasi,1983). Material should only be included if ithas some potential impact on the interpretationof the child’s scores in order to answerthe referral question(s). While parentaloccupation and marital status are generallyprivate subjects, these may be importantpieces of information, given what is currentlyknown about the effects of parentalvariables on child functioning. Lichentengeret al. (2004) provide a user-friendly and sensiblesummary of the types of informationto include in this section, as well as tips toprovide the information clearly.The report writer should also be clearabout the sources of information. If thefather views his son as lazy, then this statementshould be attributed to the father.

370 chapter 16 report writingStatements that could be used for makingsuch attributions include the following:According to…His father/mother stated…His mother’/father’s opinion is…His teacher’s view of the situationHer guidance counselor reported that…His parole officer acknowledges that…If care is not taken to make clear the sourcesof information, questions may arise at thetime when feedback is given to involvedparties.Sensitive background informationshould also be corroborated or excludedfrom the report if it is inflammatory andcannot be corroborated. For example, a5-year-old may say something like “Mymother shoots people,” and later, the psychologistdiscovers that the child’s motheris a police officer.Previous assessment results should alsobe included in this section (Teglasi, 1983).Also, previous experiences with psychologicalor educational interventions shouldbe noted here. The clinician may alsorefer the reader to a previous evaluation.Referring to previous evaluations, withoutfully recapitulating them, can substantiallyreduce written report bulk.Behavioral ObservationsIn this section, the behaviors that the childexhibits during the assessment are recorded.When writing this section, the number ofobservations made, the setting where theobservations were made (e.g., school, clinic,etc.), and the person who made the observationsshould be identified (Teglasi, 1983). Abrief description of the setting, particularlyif the report writer is describing classroomobservations, is also appropriate. Domainsthat routinely should be covered include“physical appearance, ease of establishingand maintaining rapport, response to failures/successes,response to encouragement,attention span, language style, distractibility,activity level, anxiety level, mood, impulsivity/reflectivity,problem-solving strategy,attitude toward the testing process, attitudetoward examiner, attitude toward self,unusual mannerisms or habits and validityof test results in view of behaviors” (Lichtenbergeret al., 2004, p. 60).Care should be taken not to confuseobservations with interpretations. In otherwords, it is appropriate, for example, tostate that the child appeared motivated toperform well, but such a statement shouldbe accompanied by the behaviors that ledto this assertion.Assessment Resultsand InterpretationThis section is where the test results for thechild are reported. Some report writers preferto integrate the results from various measuresinto a single section. Still others opt todivide this section into subsections accordingto domains assessed. The domains mayinclude: cognitive/intellectual, academicachievement, adaptive behavior, visual/motor, and behavioral/personality. This lattersection is of primary interest for this text.Organization within the behavioral/personality section can be according totheoretical orientation, training, or otherpreferences of the psychologist. We happento recommend that this section beorganized from the most important constructto least important, such that allevidence from multiple tools regardingthe most important domain of functioningfor the client (e.g., depression) is discussedfirst, followed by other comorbidissues, rather than presenting informationby each individual measure and then tryingagain to integrate the information fromvaried sources. This approach puts thefocus, in our view, where it belongs: theconstructs/domains of functioning, not thetests. Most importantly, this section shouldprovide coherent interpretations of resultsthat relate logically to one another and to

chapter 16 report writing371other sections, such as sections devotedto providing diagnostic considerations.Hence, this section should not simplyreport numerical findings that are devoidof interpretation.Diagnostic ConsiderationsThe decision about whether or not toinclude a separate portion dealing withdiagnostic issues is likely influenced by settingand referral questions. Nonmedicalsettings, for example, may discourage theinclusion of a discussion of this nature inthe psychologist’s report. The omission ofsuch a section may be in keeping with interdisciplinaryapproaches to making classification/diagnostic/eligibilitydecisions.The format for this section can be inlists or in paragraphs. A psychologist maysimply list diagnoses in a manner consistentwith the DSM-IV multiaxial approach.Others prefer to use a paragraph or twoto more fully explain the rationale for oragainst making certain diagnoses.SummaryThe final section of the report is intended togive an overview of the major findings. Thisreview helps to ensure that the reader understandsthe major points made in the report.A rule of thumb for writing summaries is touse one sentence to summarize each sectionof the report. In addition, a sentence shouldbe devoted to each major finding presentedin the test results section. In some cases, onesentence can be used to summarize multiplefindings and recommendations.One of the common pitfalls of preparingsummaries is including new information inthe summary section. If a clinician introducesa new finding in the summary, the reader islost. The reader has no idea as to the sourceor rationale behind the conclusion. We suggestthat students read their draft summariescarefully and check every conclusion made inthe summary against the body of the report.SignaturesReports typically require signatures attestingto their authenticity. An important componentof this seemingly unimportant aspectof the report is the necessity for cliniciansto use titles that represent them accurately.Some states, for ex-ample, do not have specialtylicensure, and the use of a title such asLicensed Pediatric Psychologist is not appropriate.In this case, a more generic term suchas Licensed Psychologist should be used,especially if the psychologist lacks evidenceof board certification of specialty training.Students should also be careful to representthemselves accurately. A title suchas Practicum Student, Intern, Trainee, orsomething similar should be used. Psychologicalcustom also dictates the inclusion ofthe highest degree obtained by the clinician.RecommendationsRecommendations should be specific andclear (Teglasi, 1983). A recommendationfor individual psychotherapy may be difficultto carry out, for example, if the specificproblems that need to be addressedand other aspects of the recommendationare not made explicit. Some reasons thatrecommendations are not subsequentlyfollowed may have to do with how they arecommunicated. Recommendations shouldbe understood by the individuals who willimplement them, developmentally appropriatefor the child, and practical, andshould avoid being unnecessarily complex(Lichtenberger et al., 2004).Some recommendations may also bedifficult to communicate succinctly inwriting. Therefore, one approach may beto include hand-outs for treating certainproblems that are much more specificthan can be included in the typical recommendationsection of a report. A handoutdetailing some specific recommendationsfor a teacher responding to inattentivebehaviors in the classroom may be morevaluable to the teacher than an abbreviated

372 chapter 16 report writingrecommendation. In almost all cases, theclinician should relay recommendations inperson to psychiatrists, teachers, parents,and other colleagues (Teglasi, 1983) inorder to ensure that they are followed.Psychometric SummarySome clinicians include a listing of all ofthe child’s obtained scores with the report.While this summary will be of limitedvalue to the less knowledgeable reader, itmay be of great value to another clinicianwho reviews the report. This summaryis best placed on a separate sheet(s)of paper, which makes it convenient forthe clinician to be selective about whoreceives the summary. Some psychologistsmay prefer to not send the summaryto parents and virtually always send it toother psychologists.The Report Writing Self-TestA report-writing self-test is provided in Fig.16.3. This checklist allows the psychologistto periodically and quickly review principlesof re-port writing.Item True False1. Was the report edited?TF2. Are unnecessary invasions of privacy avoided?TF3. Is the referral question(s) explicitly stated?TF4. Is the referral question answered?TF5. Does the report emphasize words over numbers?TF6. Can a person with a high school education understand the wording used? TF7. Is the report brief enough that major findings are not lost?TF8. Arethe conclusions drawn without undue hedging?TF9. Do the conclusions fit the data?TF10. Are invalid results omitted?TF11. Are percentile ranks included for the benefit of parents and clients?TF12. Were spelling and grammar checked?TF13. Aresupporting data integrated with conclusions?TF14. Are the recommendations clear and specific?TF15. Are headings and lists added to enhance space?TF16. Are acronyms defined and not overused?TF17. Is the summary free of new information?TF18. Were scores double-checked?TF19. Are examples of behavior used to clarify meaning?TF20. Are test instruments described adequately?TF21. Is the rationale for diagnoses provided?TF22. Is a conference scheduled to accompany the written report?23. Was written parental consent obtained prior to releasing the reportto interested agencies or parties?TTFF24. Is a feedback session scheduled with the child or adolescent?TF25. Are the type and paper of professional quality (e.g., laser-quality print)? TFFigure 16.3Report writing self-test

chapter 16 report writing373Communicating ResultsOrallyParent ConferencesFor the purposes of this text, parent is usedgenerically to include any consistent caregiverin the child’s life. Examples of suchcaregivers include stepparents, residentialcaretakers, and grandparents, among others.Imparting assessment results to parentsre-quires considerable savvy, as the individualdifferences between families aremyriad. Because of this diversity, there isnot a singular methodology that will beeffective with all parents. This section willpresent some ideas for sharing results withparents or other caregivers. However, it isvital for the psychologist to remain flexiblein order to adapt the format of the feedbacksession to the needs of the parent, othercaregiver, or family, as well as the setting.In an old but insightful article, Ricks(1959) summarized the heart of the parentconference dilemma.The audience of parents to which our testbasedinformation is to be transmitted includesan enormous range and variety of minds andemotions. Some are ready and able to absorbwhat we have to say. Reaching others may beas hopeless as reaching watchers with an AMradio broadcast. Still others may hear what wesay, but clothe the message with their own specialneeds, ideas, and predilections (p. 4).Regardless of the potential pitfalls, parentsmust be informed of the results of apsychological evaluation of their child (thelegal, ethical, and regulatory mandates forthis practice are given in Chapter 4).Some helpful suggestions for communicatingtest results to parents are given next:1. Avoid excessive hedging or deceit. Theproblem with hedging or failing toreport bad news is that many parentssense this deceit and respond to thepsychologist with appropriate mistrust.Honesty is also easily sensed by parents,which ultimately enhances the credibilityof the psychologist.2. Use percentile ranks heavily whendescribing norm-referenced test results.This metric is easier for parents tounderstand than other norm-referencedscores.3. Instead of lecturing, allow parentsopportunities to participate by askingabout topics such as their opinion ofthe results and how they fit with theirknowledge of their child. Moreover,listening carefully to parents helps thepsychologist determine the psychologicalneeds of the parents that are relevantto the evaluation. Similarly, it is essentialthat the parents be given frequentopportunities to ask questions (Lichtenbergeret al., 2004).4. Anticipate questions prior to the interviewand prepare responses. How woulda psychologist answer the question, “Willmy daughter outgrow her ADHD?”Psychologists can gauge the probabilitythat such questions will arise by listeningcarefully in the intake interview andthroughout the assessment process.5. Schedule adequate time for the interview.Parent conferences often becomemore involved than one has planned.Adequate time allows the psychologisttime to use counseling skills to bringa parent feedback conference to adequateclosure. Ideally, 1–2 h could beallocated for such a parent session. If asession ends early, then the psychologistis the recipient of a precious gift– extra time.6. It is often helpful to seek practice communicatingwith parents from a varietyof back-grounds. Some parents can beaddressed as if they are colleagues, whileothers may have only a limited graspof the issues being discussed. Translators,ministers, teachers, trusted family

374 chapter 16 report writingfriends, and others may serve as allies inthe feedback process.7. Avoid questionable and/or overlyexplicit predictions (Kamphaus, 2001).Phrases to be avoided would be statementslike, “She will never go to college,”or “She will always have troublewith school.” These types of statementscan be offensive to parents, not to mentioninaccurate.8. Use good, basic counseling skills. Everyparent likes to talk about the trials andsuccesses of raising a child. Give parentsat least some opportunity to do this, as itallows you to show interest in the childby listening to the parent’s perspective.9. Do not engage in counseling that isbeyond your level of expertise. Parentsare often very eager to obtain advice froma professional. It is inappropriate (andunethical by most stan-dards) for a psychologistto provide services for whichhe or she is not trained. If, for ex-ample,a parent requests marital counselingand you have no training in this area,you should inform the parent of thisfact and offer a referral. In fact, the psychologistis wise to have referral sourcesreadily available for such eventualities.10. Be aware that some parents are notready to accept some test results. Parentsmay impugn your skills becausethey cannot accept the fact that theirchild has a severe handicap. They mayleave the session angry, and you mayfeel inept. The idea that every parentconference will end on a happy note isunrealistic. Examine your skills criticallyin response to parent feedback,but realize that some parents simplywill not accept the results because oftheir own personal issues. An exampleof such a situation may involve a parentwith the same handicapping conditionas the child. If a parent was labeledhandicapped and ridiculed by peers, heor she may become defensive and angryat the suggestion that his or her childmay have a handicap. The session withsuch a parent will likely end on a tensenote. In many of these cases, however,the parent will adapt and accept thenews after developing the psychologicalresources to cope with the attendantstresses. The psychologist mayfind this same parent to interact morepositively in the next encounter.11. Maintain a positive tone throughoutthe session and discuss the child’sstrengths and competencies.Teacher ConferencesMany of the principles used in parentconferences also apply to teachers. Severalnuances, however, will be outlined in thefollowing suggestions:1. Do not monopolize a teacher’s breakfrom teaching. Some teachers get fewbreaks in a day. Most get a brief lunch,when they prefer to unwind with colleaguesand prepare for the remainderof the day. A clinician is unlikely to commanda teacher’s undivided attentionduring such breaks. If a teacher has anadditional free period, it may be a goodtime for a conference. After school isfrequently the best time to get a teacher’sundivided attention for a meeting.Teachers are generally very busy people,so the pace of the meeting will bequicker than is the case for parents.2. Teachers are interested in schoolingissues. The diagnosis of Conduct Disorderis of less concern to teachersthan getting specific recommendationsfor helping the child in the classroom(Teglasi, 1983). If a psychologist is nottrained and/or has little experience inteacher consultation, the assistance ofsomeone like a qualified school psychologistshould be enlisted to assist withthe teacher conference.

chapter 16 report writing375In any assessment with a school-agedchild, the clinician should be prepared toconduct a teacher conference, or at thevery least be available to answer any questionsand facilitate the implementationof classroom-based interventions. Such aconference is desirable because teachersare usually involved somehow in the treatmentof children and adolescents.Child FeedbackProviding assessment feedback to a childis often overlooked, but it is important todo so in most cases because the child willlikely begin some interventions or experiencesome changes in his or her environmentdirectly related to the assessment.The major decision that a clinician needsto make before giving feedback to a childregards the type of information that isappropriate for a child’s developmentallevel. Clearly, the kind of feedback given toparents is inappropriate for a 5- or 6-yearold,who may have extraordinary difficultyunderstanding the concept of a percentilerank. A child this age, however, may beable to understand the consequences ofthe evaluation. In this situation, the childmay be able to understand something like:“Remember those tests I gave you? Well,some of them seemed hard for you. Becauseof this, I suggested to your parents that yoube helped after school. So now, you will begoing to visit a teacher after school whowill help you with schoolwork.”The older the child, the more similarthe feedback session becomes to the onefor parents. One dramatic difference,however, is that negative feedback to achild or adolescent can have the oppositeof the intended effect. That is, in mostcases, the goal is to improve variables suchas peer-related social skills. A child who istold that he or she has poor social skillsmay decide to stop trying to interact withpeers. In some cases, the clinician’s honestycould harm the child. A few optionsare available in cases where a clinician isconcerned about such negative consequences.One option is to have someonewho knows the child well and has a positiverelationship with him or her help thepsychologist communicate the results in anon-threatening way to the child. A goodperson to fill this role is a teacher, otherprofessional caregiver, or possibly a parent.A second possibility, if applicable, isto have the child’s primary therapist orcounselor eventually share the results withthe child in a counseling session, when heor she could help the child cope with theresults in a supportive setting.In most cases involving feedback tochildren or adolescents, it is advisable toconsult with a fellow professional (e.g.,teacher, counselor, speech therapist, etc.)who knows the child extremely well. Thiscolleague can help the psychologist gaugethe ability of the child to deal appropriatelywith the assessment results and associatedinterventions.ConclusionsReport writing and oral reporting are central,not ancillary, considerations in the assessmentprocess. The most insightful and elegantof evaluations is lost if not translatedto usable information in written reportsand intervention planning meetings.Unfortunately, these central assessmentskills are under-emphasized in the trainingof clinicians who are left to acquire theseskills through trial-and- error. Cliniciansare advised to seek out expert supervisionin this area, if it is not readily offered. Inaddition, enlisting the aid of a competenteditor can markedly enhance the qualityof written work. Writing is not easy. Writingskills, however, can be acquired andimproved with diligence and patience.

376 chapter 16 report writingChapter Summary1. Psychological reports are frequently madeavailable to parents, judges, lawyers, andother non-psychologists, creating theopportunity for improper interpretationof the results by untrained individuals.2. Psychological reports can be useful toother clinicians who evaluate a childwho has previously been seen by a psychologist.3. Different audiences require differenttypes of written reports.4. Some of the common problems withreport writing include the following:(a) Vocabulary problems(b) Faulty interpretation(c) Report length(d) A number emphasis(e) Failure to address referral questions5. Some research has shown that teachersprefer a question-and-answer reportformat.6. Parents also tend to prefer a questionand-answerformat to other formats,although the difference in preferencescores between the psychoeducational andquestion-and-answer reports in one studyfailed to reach statistical significance.7. Suggested report writing practicesinclude the following:(a) Report only pertinent information(b) Define abbreviations and acronyms(c) Emphasize words rather than numbers(d) Reduce difficult words(e) Describe the tests used(f) Edit the report at least once(g) Use headings and lists freely(h) Use examples of behavior to clarifymeaning(i) Reduce report length(j) Check scores(k) Check spelling and grammar8. Psychological reports often includesome or all of the following headings:(a) Identifying Information(b) Assessment Procedures(c) Referral Question(s)(d) Background Information(e) Behavioral Observations(f) Assessment Results and Interpretation(g) Diagnostic Considerations(h) Summary(i) Signatures(j) Recommendations(k) Psychometric Summary9. Hints for communicating test resultsto parents include the following:(a) Be direct and honest(b) Use percentile ranks heavily whendescribing test results(c) Allow parents opportunities toparticipate(d) Anticipate questions prior to theinterview and prepare responses(e) Schedule adequate time for theinterview(f) Practice communicating with parentsfrom a variety of backgrounds(g) Avoid questionable predictions(h) Use good, basic counseling skills toconvey difficult information(i) Do not engage in counseling thatis beyond your level of expertise(j) Be aware that some parents are notready to accept some of the conclusionsoffered10. Teacher conferences are important forensuring cooperation with recommendations.11. The major decision that a clinicianneeds to make before giving feedbackto a child regards the type of informationthat is appropriate for the child’sdevelopmental level.

C h a p t e r 1 7Assessment of AttentionDeficit Hyperactivity andDisruptive Behavior DisordersChapter QuestionsIntroductionl What are some findings from researchon Attention-Deficit HyperactivityDisorder (ADHD) that have importantimplications for designing clinicalassessments for children suspected ofhaving ADHD?l What are some practical guidelines fordesigning an assessment battery forchildren suspected of having ADHDand interpreting the assessment results?l What are the implications of research onchildhood conduct problems for designingclinical assessments for children withthese problems?l What basic questions should beaddressed in clinical assessments of childrenwith conduct problems?This chapter is the first of a series ofchapters focusing on the assessment ofseveral specific types of childhood emotionaland behavioral problems. Thesechapters are designed to help an assessorapply information on the various assessmentstrategies discussed in previouschapters to the assessment of some of themore common types of psychopathologyexhibited by children and adolescents.We start with the assessment of disorders,sometimes called externalizing behaviors(Achenbach & Edelbrock, 1978) ordisorders of undercontrol (Quay, 1986).It is appropriate to start our syndromeby-syndromediscussion with this classof disorders because they tend to be themost common reason for referral to child377P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_17, © Springer Science+Business Media, LLC 2010

378CHAPTER 17 Assessment of Attention Deficit Hyperactivitymental health clinics (Frick & Kimonis,2008).This predominance in clinic referrals isout of proportion to the prevalence of thesedisorders in the general population, whereemotional difficulties are often as prevalent(Costello, Egger, & Angold, 2005). Thishigh referral rate for disruptive behaviordisorders is likely due to two factors.First, unlike adult mental health referrals,children and adolescents are rarely selfreferred.Instead, they are often referred bysignificant others (e.g., parents, teachers,physicians) in their environment. Second,disruptive behavior disorders, as the nameimplies, are syndromes of behavior thatcause significant disruptions in a child’senvironment, often directly affecting thoseresponsible for referring a child for assessmentand treatment. Thus, anyone workingin a clinical setting with children andadolescents must have a firm understandingof these behavioral disorders.To reiterate a common theme of thisbook, our recommendations for assessingchildren with ADHD and the disruptivebehaviors disorders are based on researchon the basic characteristics of these disorders.In each of the following sectionswe first provide a brief discussion of themost clinically relevant research findingsand then discuss specific recommendationsfor assessment procedures based onthese findings. We divide our discussioninto two sections corresponding to themajor subdivisions within the externalizingdisorders. The first section involvesa discussion of a syndrome of behaviorsinvolving inattention-disorganization andimpulsivity-motor hyperactivity labeledas Attention-Deficit Hyperactivity Disorder(ADHD) by the DSM-IV-TR (AmericanPsychiatric Association, 2000). Thesecond section focuses on conduct problemsand aggression subsumed under thecategories of Oppositional Defiant Disorder(ODD) and Conduct Disorder (CD)(American Psychiatric Association, 2000).Attention-DeficitHyperactivity DisorderClassification and SubtypesHistoryThe syndrome of ADHD has long beenrecognized in the medical and psychologicalliterature, with some descriptionsdating back well over 100 years (Smith,Barkley, & Shapiro, 2007). However, overthe years there has been considerable disagreementover what are the core featuresof the disorder. As a result of this confusion,there have been numerous changes indiagnostic definitions (see Frick & Lahey,1991; Smith et al., 2007). This evolutionin our conceptualization of ADHDis reflected in changes in the diagnosticcriteria for ADHD over the most recentrevisions of the Diagnostic and StatisticalManual of Mental Disorders.The second edition of the DSMincluded a syndrome of the HyperkineticReaction of Childhood to emphasize thebelief that motor hyperactivity is the corefeature of the disorder (DSM-II; AmericanPsychiatric Association, 1968). In thethird edition of the DSM the disorder wasreconceptualized to emphasize deficits insustained attention, acquiring the termAttention Deficit Disorder (DSM-III;American Psychiatric Association, 1980).Also in the DSM-III, it was explicitly recognizedfor the first time that childrencould have attention deficits in the absenceof motor hyperactivity. The next revision,DSM-III-R, softened this emphasis onattention deficits, placing it on equal footingwith motor hyperactivity in its definitionof Attention Deficit HyperactivityDisorder (American Psychiatric Association,1987). The DSM-III-R criteria alsodid away with subtypes based on the presenceor absence of motor activity.Before discussing the most recent revisionof the DSM (DSM-IV-TR; American

CHAPTER 17 Assessment of Attention Deficit Hyperactivity379Psychiatric Association, 2000), it is importantto highlight two main sources of variationin these conceptualizations of thedisorder. First, a main source of debatehas been over what are the core features ofthe disorder. For example, DSM-III proposedthree core dimensions of behaviorassociated with attention deficit disorders:(1) inattention (e.g., very distractible, difficultyfinishing things), (2) impulsivity(e.g., acts without thinking, interrupts others,loses things, makes careless mistakes),and (3) hyperactivity (e.g., fidgety and restless,running around and climbing excessively).In contrast, DSM-III-R eliminatedany distinctions among these behaviorsand considered all three dimensions to beindicative of a single domain of behavior.Research has fairly consistently suggestedthat the best method for conceptualizingthe symptoms is somewhere inbetween. Specifically, factor analyses havegenerally been able to document two partiallyindependent dimensions of behavior:inattention/disorganization and impulsivity/overactivity(Lahey, Carlson, & Frick,1997). The behaviors that are generallyconsidered to be the core features ofADHD are listed in Table 17.1.The second source of contention evidentthrough the revisions of the DSMis the debate over whether there are validsubtypes of ADHD. DSM-III proposed theexistence of two types of attention deficitdisorder: Attention Deficit Disorder withoutHyperactivity (ADD/WO) and AttentionDeficit Disorder with Hyperactivity(ADD/H). These subtypes shared the corefeatures of inattention and impulsivity butdiffered on the presence of motor hyperactivity.Lahey et al. (1997) summarize asignificant body of research attesting tothe validity of this distinction. Childrenwith ADD/H tend to exhibit more conductproblems, to be more impulsive, andto be more socially rejected than childrenwith ADD/WO. In contrast, children withADD/WO tend to be more sluggish anddrowsy (often described as unmotivated),to be more anxious and shy, and to be morelikely to show an optimal response to lowdoses of stimulant medication than childrenwith ADD/H. Based on their reviewof the research, these authors concludedTable 17.1 Core Dimensions of Attention Deficit Hyperactivity DisorderInattention-DisorganizationDifficulty organizing things a,bDifficulty finishing tasks a,bDifficulty following through on instructions a,bOften loses things a,bEasily distracted a,bOften does not listen a,bDifficulty concentrating and sustaining attention a,bMisses details and makes careless mistakes aOften avoids or dislikes tasks requiring sustainedmental effort aOften forgetful aNeeds a lot of supervision bImpulsivity-HyperactivityExcessive running and climbing a,bDifficulty playing quietly a,bExcessive talking a,bFrequently interrupts and intrudes a,bAlways on the go a,bExcessive fidgeting and squirming a,bDifficulty staying in seat a,bDifficulty waiting turn a,bFrequently blurts out answers a,bFrequently calls out in class baSvmptoms included in the DSM-IV-TR (American Psychiatric Association, 2000) criteria for ADHD, althoughwording may not be exactly as that included in the manual. b Behaviors included in factor analyses reviewed byLahey et al., 1997.

380CHAPTER 17 Assessment of Attention Deficit Hyperactivitythat the decision to eliminate the subtypesin the DSM-III-R was not consistent withthis body of research.DSM-IV-TRThe DSM-IV-TR definition of ADHD wasdesigned to reflect research findings onboth of these issues (American PsychiatricAssociation, 2000). First, there are twosymptom lists, which closely correspond tothe two dimensions of behavior describedin Table 17.1. Second, the DSM-IV-TRrecognizes the existence of subtypes basedlargely on the presence of hyperactivity.There is an ADHD Predominantly InattentiveType to designate children withproblems of inattention and disorganizationbut without problems of impulsivityand overactivity. In addition, there are theADHD Predominantly Hyperactive Typeand ADHD Combined Type to designatechildren with significant problems ofimpulsivity-hyperactivity, either in isolationfrom or in combination with problemsof inattention and disorganization.While these subtypes seem to match theresearch findings summarized previously,it is important to note that the stability ofthese ADHD subtypes over time is questionable.Specifically, in a sample of 118children with ADHD who were ages 4 to6 at the start of the study, Lahey, Pelham,Loney, Lee, and Wilcutt (2005) reportedthat it was not unusual for children tochange in their subtype of ADHD over the8-year study period. For example, 37% ofchildren with Combined Type and 50% ofthe children with Predominantly InattentiveType met criteria for a different subtypeat least twice during the study period.Children with the Hyperactive Type werethe most likely to shift subtypes, with mostshifting to the Combined Type at somepoint during the study.An inspection of the symptoms includedin the DSM-IV-TR criteria for ADHD (seeTable 17.1) indicate that the individualbehaviors that form the diagnostic criteriafor this disorder are behaviors that arequite common to some degree in normallydeveloping children and adolescents. Thisis one of the issues that has led to seriousconcerns over the potential overdiagnosisof the disorder and concomitant overuse ofstimulant medication to treat it (Angold,Erkanli, Egger, & Costello, 2000; Jensenet al., 1999). There are two critical issuesrelated to these concerns. First, at present,there is little empirical evidence tosupport the concerns about overdiagnosisand overmedication, although clearly thisis a very difficult issue on which to obtaingood data (Jensen et al., 1999). Second, thesymptoms of most childhood disorders,not just ADHD, are not qualitatively differentfrom normal behaviors shown bychildren (e.g., sadness as a symptom ofdepression). This is not to imply that this isnot an important issue in assessment but toillustrate that it is not specific to ADHD.It relates to the important issue raised inChap. 3 that classification systems mustclearly define what parameters are mostimportant for differentiating disordered(i.e., clinically impairing) manifestationsof the symptoms from more normal manifestations.The DSM-IV-TR includes severalsuch parameters for the diagnosis ofADHD, and it is imperative that assessorssystematically assess these parameters toavoid overdiagnosis.The first parameter is the frequency andseverity of the symptoms. The DSM-IV-TRsets six symptoms of either inattention-disorganizationor impulsivity-hyperactivityas the diagnostic threshold for the disorder.This level of severity was chosen basedon evidence that it seemed to designate alevel of symptomology that predictedclinically significant levels of psychosocialimpairment (e.g., poor academic performance,social rejection) for elementaryschool-aged children (Lahey, Applegate,McBurnett, et al., 1994). Using this diagnosticthreshold, 3–7% of children would

CHAPTER 17 Assessment of Attention Deficit Hyperactivity381typically meet the diagnostic criteria forADHD (American Psychiatric Association,2000). It is important to note that theappropriateness of this threshold has beenquestioned for young preschool children asbeing too liberal, because many very youngchildren show high rates of these behaviorsand eventually outgrow them (Campbell,1990), and for adolescents and youngadults as being too conservative, becausethe frequency and severity of many of thesymptoms seem to decline in adolescence(Barkley, 1997a).The second parameter that differentiatesnormal and abnormal patterns of inattention,impulsivity, and overactivity is theonset and duration of the symptoms. DSM-IV-TR specifies that “some hyperactive-impulsiveor inattentive symptoms must havecaused impairment before age 7 years”(American Psychiatric Association, 2000,p. 92). This criterion is consistent with theconceptualization that ADHD is a lifelongpattern of maladaptive behavior and not atransient reaction to a specific stressor orto the demands of a particular developmentalstage (Barkley, 1997a). While theage of onset criterion is consistent withthis conceptual framework, there are severalpractical problems in using this criterionin clinical assessments. First, it is oftendifficult to gain accurate accounts of whensymptoms became problematic, especiallywhen assessing adolescents and adults,which involves recall of events over a longperiod of time (Barkley, 1997a).Second, it is not uncommon for manyof the symptoms of ADHD, especially theinattention ones, to only become problematiconce the demands for sustainedattention and organization increase in laterelementary school years (Lahey et al., 2005;Loeber, Green, Lahey, Christ, & Frick,1992). This developmental change in inattentionsymptoms is likely the reason thatthe age of 7 onset criterion may be particularlyproblematic for the PredominantlyInattentive Type of ADHD. Specifically,in a sample of 380 clinic-referred children(mean age 8.7 years) who met DSM-IV criteriafor ADHD, almost all of those whomet the symptom cut-off for the PredominantlyHyperactive-Impulsive Type andthe Combined Type met the age of 7 onsetcriterion (Applegate et al., 1997). In contrast,only 48% of those children with PredominantlyInattentive Type met this age ofonset criterion and those who did not meetthe criterion did not differ from those thatdid on several important validity indexes,including level of impairment and level andtype of comorbidity with other disorders.Therefore, the validity of the onset criterionfor this ADHD subtype was questionable.Third, it is important to recognizethat, while the core deficits underlying thesymptoms of ADHD may be stable acrossdevelopment, how these deficits are manifestedin symptoms and secondary characteristicsmay change across development(Barkley, 1997a). In Table 17.2, we providea brief summary of some of these developmentalchanges in the symptom patternsand secondary characteristics acrossdevelopment. One criticism of the DSM-IV-TR definition of ADHD that includesthe same number and types of symptomsfor children, adolescents, and adults is thatthis static definition may not capture thesedevelopmental changes adequately (Barkley,1997a).A third parameter in the DSM-IV-TRdefinition of ADHD that is importantfor separating normative from disorderedlevels of attention and overactivity is thespecification of cross-situational consistencyof symptoms. That is, to be diagnosed withADHD, a child must show impairmentrelated to the symptoms in two or moresettings. This criterion is consistent withthe conception that ADHD should not besolely a function of a single set of environmentalcircumstances (e.g., a disorganizedclassroom, a chaotic home environment)and clearly suggests that an adequate assessmentof ADHD must involve an assessment

382CHAPTER 17 Assessment of Attention Deficit HyperactivityTable 17.2 Developmental Changes in the Core Features and Secondary Characteristics Associated with ADHD Across the LifespanAge Primary Characteristics Secondary FeaturesPreschool Restlessness, excessive activity, difficulty remaining seated,and acts without thinkingElementary-school age Poor attention span, distractibility, impulsivity, difficultyremaining seated, and can’t play quietlyAdolescence Poor attention span, distractibility, inability to finish things,careless mistakes, and lack of forethought and planningYoung adults Restlessness, distractibility, difficulty completing work,carelessness, and lack of forethought and planningNoncompliance, accidental injuries, aggression, andproblems in toilet trainingDifficulty following rules, immaturity, social rejection,needing a lot of supervision, failure to do householdchores, and poor school performancePoor school performance (dropping out), lowself-esteem, depression, substance use, delinquency,family conflict, needless risk taking, automobileaccidents, teenage pregnancy, and rebelliousnessPoor educational and occupational performance,depression, substance use, poor social/marital adjustment,and poor anger managementSources: Summarized from Barkley, 1997a; Barkley, Guevremont, Anastopoulos, & Fletcher, 1993; DuPaul & Stoner, 1994; Nadeau, 1995; Smith et al., 2007; Wender, 1995.

CHAPTER 17 Assessment of Attention Deficit Hyperactivity383of the child in several different settings.However, although this criterion appearsquite basic, it is difficult to use in clinicalassessments for several reasons. For example,it is clear that children with ADHDwill show variations in the level and severityof their behavior problems dependingon the demands of the situation (e.g., timeof day, level of structure, and complexityof the task) (Barkley, 1997a). Therefore,an assessor should not interpret the crosssituationalcriterion to imply that a childwith ADHD must show the same level andseverity symptoms across different settingswith different demands. Instead, one mustconsider whether the child shows similarbehaviors in situations with equivalentdemands, and such judgments are very difficultto make. For example, if a child ishaving trouble associated with ADHD atschool but not at home, the clinical assessormust judge whether or not this is due tothe fact that demands for sustained attentionand sitting still are not placed on thechild at home.Taken together, the level and severityof symptoms, their presence over extendedperiods of time, and their cross-situationalconsistency all are critical components ofdefining ADHD and must be systematicallyassessed in making this diagnosis. Inherentin each of these criteria is that they designatechildren who show some significantlevel of impairment in their psychosocialfunctioning due to the symptoms associatedwith ADHD. On the simplest, butpossibly the most important level, it is thissignificant impairment in functioning (e.g.,causing problems in school work, causingsocial rejection) that is most important indifferentiating normative and disorderedlevels of the symptoms that are part ofthe ADHD definition. Furthermore, thedegree of impairment in family functioning,peer relationships, and academic functioning,are often the main reasons that a childwith ADHD is referred for an evaluationand they are some of the best predictorsof a child’s long-term adjustment (Pelham,Fabiano, & Massetti, 2005).ComorbiditiesChildren with ADHD are an excellentexample of the fact that children with problemsin one area of adjustment are at risk forproblems in other areas as well. Problemsthat often co-occur with ADHD are quiteimportant clinically. They often cause moredisruptions for the child and predict pooreroutcomes than the primary ADHD symptomsthemselves (Frick & Lahey, 1991). Asa result, the secondary features are often amajor focus of intervention (Pelham et al.,2005).Conduct Problems/AggressionThe most common co-occurring problemsexperienced by children with ADHDare conduct problems and aggression,with research suggesting that 60–75% ofchildren referred to clinics with ADHDshow significant levels of these problems(Hinshaw, 1987). It is often these conductproblems that lead to a great deal of disruptionfor children with ADHD, leading tomultiple disciplinary confrontations withparents and teachers, school suspensions,and problems in peer relations. In addition,these conduct problems are often predictiveof poor outcomes in adolescence andyoung adulthood, especially for predictingdelinquency and substance abuse (Fischer,Barkley, Fletcher, & Smallish, 1993; Mannuzza,Gittelman-Klein, Konig, & Giampino,1989).Other ComorbiditiesAnother condition that often occurs withADHD is academic underachievement ora learning disability, both of which arefrequently defined as school achievementbelow a level predicted by a child’s age

384CHAPTER 17 Assessment of Attention Deficit Hyperactivityand intellectual level. Approximately 30%of children with ADHD show such learningproblems (Frick, Lahey, Christ, Loeber,& Green, 1991; Massetti et al., 2008).In addition, children with ADHD tendto show a high rate of conflict with peers(Mikami & Hinshaw, 2003), with parents(Johnston & Mash, 2001), and withteachers (Cunningham & Boyle, 2002).And not surprisingly, given the amountof difficulty and conflict the ADHD childoften experiences in his or her environment,children with ADHD often showhigh rates of anxiety (Tannock, 2000) andlow self-esteem that persist throughoutchildhood and into adolescence (Fischeret al., 1993).Conceptual ModelThere have been numerous theories ofADHD that differ in terms of identifyingthe “core deficit” that underlies thesymptoms of ADHD and the etiologicalfactors that lead to this deficit. A growingnumber of researchers in this area havebegun to focus on a failure in a child’sinhibition system that influences his orher ability to regulate attention, actions,and emotions (e.g., Nigg, 2006; Whalen& Henker, 1998). One of the more influentialand best articulated of such theoriesis one proposed by Barkley (1997b),which defines “behavioral inhibition” asthe capacity to inhibit motivated behaviors,either prior to their initiation oronce they are initiated, which creates adelay between an impulse and action. Thisdelay allows the child to “think through”his or her actions and allows the behaviorto be self-directed and guided by thedemands of any given situation. A deficitin this inhibition system would make itdifficult for a child to sustain his or herattention on a single task, it would makeforesight and planning difficult, and itwould make it difficult for the child toinhibit impulses for motor movement,thereby accounting for the core symptomsof the disorder. Barkley also outlineshow such a deficit could account formany of the other characteristics typicallyfound in people with ADHD, such as apoor sense of time, poor emotional selfcontrol,deficits in problem solving, andan inability to modulate behavior basedon changing situational demands.While there is a growing consensus that adeficit in the inhibitory control of behaviormay be a primary or at least an importantdeficit in children in ADHD, it is less clearwhat could cause this deficit to develop.Many theories focus on structural neurologicalabnormalities in parts of the nervoussystem involved in inhibitory controlof behavior (Castellanos et al., 2002). Othertheories focus on abnormalities in the functioningof these neurological regions withstudies examining the cerebral blood flowof children and adults with ADHD consistentlyshowing areas of decreased activityin the prefrontal regions of the brain(Hendren, DeBacker, & Pandina, 2000).There is evidence that these neurologicalabnormalities can result from anumber of different influences. Specifically,there is evidence that ADHD symptomsare highly heritable and, as a result,these neurological abnormalities may beinherited (Waldman & Gizer, 2006). Inaddition, the neurological abnormalitiescould result from trauma to the developingnervous system such a prenatalexposure to alcohol or other drugs, birthtrauma, or exposure to environmentaltoxins (e.g., lead) (Smith et al., 2007). Itis important to note that, although mosttheories of ADHD emphasize potentialneurological underpinnings to the disorder,there is currently no neurological testthat has proven to be useful in diagnosingADHD. Instead, the diagnosis relieson a careful assessment of the behaviorallybased diagnostic criteria usinga process outlined in the next sectionof this chapter. Although social experiencescan influence the development of

CHAPTER 17 Assessment of Attention Deficit Hyperactivity385behavioral inhibition and can changebrain functioning (Cicchetti & Walker,2001), most theories have not emphasizedenvironmental factors as primarycausal agents in the development of thecore symptoms of ADHD. Instead, mosttheories emphasize the role of environmentalfactors in determining how thecore deficit is expressed. For example, achild with problems in behavioral inhibitionwill be very difficult to socialize butsome parents will be better than othersin working with such a child to developcompensatory strategies to minimize theeffects that the inhibitory deficit mayhave on the child’s academic and psychosocialfunctioning.As a result, environmental factors canplay a large role in the development ofsome of the problems (e.g., poor schoolperformance, social rejection, conductproblems) that can develop secondarily tothe ADHD symptoms (Frick, 1994; Johnston& Mash, 2001). Therefore, althoughpsychosocial influences may not be integralto many causal theories of ADHD, itis still important to carefully assess a child’spsychosocial context to (1) determine thedegree to which the problems associatedwith ADHD have negatively impacteda child’s functioning and (2) guide interventionsdesigned to reduce or preventmany of the secondary characteristics andco-occurring problems in adjustment thatoften develop in children with ADHD(Pelham et al., 2005).Implications for AssessmentIn Table 17.3, we summarize the mainimplications of research on ADHD fordesigning an appropriate assessment batteryfor children or adolescents suspectedof having ADHD. In addition, in Boxes17.1 and 17.2, we provide two case examplesof a typical ADHD assessment battery,with Box 17.1 describing the assessmentof a child with ADHD-Combined Typeand Box 17.2 describing the assessmentof a child with ADHD-PredominantlyInattentive Type.Table 17.3 The Nature of ADHD and Implications for AssessmentFocus of ResearchClassification and presenceof subtypesPresence of multiplecomorbiditiesPotential alternative causesImplications for AssessmentAssess for presence of two core dimensions of behavior inattention/disorganization and impulsivity/overactivityAssess for subtypes based on the presence of impulsivity/overactivityAssess duration to determine if behaviors are chronic and stableAssess situational variability of behaviorsAssess level of impairment associated with symptomsAssess for the presence of conduct problems/aggressionAssess for the presence of learning problemsAssess anxietyAssess self-esteemAssess social relationships and peer social statusAssess level of parent–child and teacher–child conflictsObtain a developmental and medical historyAssess for intellectual and learning deficitsAssess for emotional difficulties

386CHAPTER 17 Assessment of Attention Deficit HyperactivityBox 17.1Case Study: Evaluation of an 8-Year-Old Girl with Attention-Deficit HyperactivityDisorder-Combined TypeClaire was 8 years, 9 months old when hermother referred her to an outpatient mentalhealth clinic for a comprehensive psychologicalevaluation. Claire’s mother was concernedabout Claire’s aggressive behavior, describingClaire as having an “uncontrollable temper”and being very defiant. Maternal report alsoindicated that Claire has had very inconsistentacademic performance throughout her first 3years of school, primarily because she failedto complete work and made a lot of carelessmistakes. To illustrate the effect of Claire’scarelessness on her school performance, hermother described an incident the previousschool year in which Claire rushed through anarithmetic test and completed all the questionsas addition, even though half of the problemswere subtraction. Claire reportedly knew howto do subtraction problems.Assessment of Core SymptomsThe core symptoms of ADHD were assessedthrough a structured diagnostic interview conductedwith Claire’s mother and her teacher(DISC-IV; Shaffer et al., 2000 and behaviorrating scales completed by her mother (CBCL;Achenbach, 2001) and her teacher (CBRSC;Neeper, Lahey, & Frick, 1990). On the DISC-IV, Claire’s mother and teacher both reportedsignificant problems of inattention-disorganization,such as difficulty sustaining attention, havingdifficulty finishing tasks, often losing things,frequently making careless mistakes, and havingvery messy work habits. In addition, theseproblems in attention were accompanied bysignificant problems of impulsivity and motorhyperactivity. Both mother and teacher indicatedthat Claire frequently interrupted others;often talked out in class; was very fidgety andrestless in class; and could not stay in her seat,either in class or at home to eat dinner.Consistent with a diagnosis of ADHD-Combined Type, Claire’s mother reported onthe structured interview that these problemshave been evident since very early in Claire’slife, especially the motor overactivity. In fact,maternal report indicated that Claire had beenasked to leave two preschools because she wastoo “rambunctious” and could not sit still.Claire’s teacher indicated that these symptomsof ADHD were currently interfering with herschool performance to a substantial degree.On the Academic Performance Rating Scale(APRS: DuPaul, Rapport, & Perriello, 1991).Claire’s teacher indicated that she was turningin less than half of the work required bythe class and it often was inaccurate, despiteClaire knowing the material.Finally, parent and teacher report on theomnibus rating scales suggested that the coreADHD behaviors were more severe thanwould be expected in children her age. Usingthe age- and gender-specific norms of CBCL,Claire had elevations on both the AttentionProblems (T-score of 75) and Thought Problems(T-score of 76) scales. Similarly, Claire’steacher on the CBRSC rated her as elevated onthe Motor Hyperactivity scale, with a T-scoreof 79 based on the entire normative sample(across ages and gender). Whereas the CBCLelevations may have been spuriously high dueto comparisons restricted to girls, the elevationon the CBRSC based on a comparison toboth boys and girls clearly indicated that herbehavior was more severe than is typical forchildren her age.Assessment of ComorbiditiesLike many children with ADHD-CombinedType, Claire also exhibited significant conductproblems. On the DISC-IV, her parent andteacher described Claire as showing frequenttemper tantrums, often arguing with adults,often refusing adults’ requests, blaming othersfor her mistakes, and being grouchy andeasily annoyed. These behaviors were ratedas severe on both parent and teacher ratingscales, with T-scores above 70 being obtainedon the Delinquent and Aggressive Behaviorscales of the parent-completed CBCL and the(Continues)

CHAPTER 17 Assessment of Attention Deficit Hyperactivity387Box 17.1 (Continued)Oppositional-Conduct Disorders scale of theteacher-completed CBRSC.On the CBCL Social Problems scale and theCBRSC Social Competence scale, mother andteacher also indicated that Claire’s behavioralproblems seemed to be affecting her peer relations.She was described by parent and teacheras being bossy and domineering in peer interactions,which had led to difficulties in makingfriends. However, a sociometric exercisedid not indicate that Claire’s social status wasnegatively affected by this behavior. In a classof 13, she was nominated as “Liked most” by 3children and “Liked least” by only 2 children.A psychoeducational evaluation did notreveal any significant learning problems.Claire’s intelligence scores were in the averageto high average range in both verbal andnonverbal abilities. She also obtained agestandardscores in the high average range onthe individually administered achievementtest. Thus, there were no indications of cognitivedeficits, and her achievement scoresindicated that she seemed to be learning ator above a level expected for her age.Ruling Out Alternative CausesClaire’s birth, medical, and developmentalhistory did not suggest the presence of anymedical or neurological disorder. As mentionedpreviously, the psychoeducationalevaluation did not reveal any cognitive orlearning problems that could account for thebehaviors. An unstructured clinical interviewdid reveal that Claire reported being sexuallyabused over a period of 1 month duringthe previous summer by a paternal uncle.This alleged abuse had been reported to thelocal child protection agency, and Claire hadbeen seen at the community mental healthcenter for 3 months following the incident.However, it did not appear that Claire’s difficultiescould be solely accounted for byan emotional reaction to sexual abuse, forseveral reasons. First, the ADHD behaviorswere more severe than would be expectedfrom such a reaction, and they clearly predatedthe alleged abuse incident. Second, shedid not show any other signs of anxiety anddepression that would suggest a significantdegree of emotional distress.Box 17.2Case Study: Evaluation of an 8-Year-Old Boy with Attention-Deficit HyperactivityDisorder-Predominantly Inattentive TypeSean was 8 years, 1 month when his teacherreferred him to a child mental health clinicfor a comprehensive psychological evaluation.Sean was failing most subjects in the thirdgrade and his teacher attributed this poor performanceprimarily to problems in concentration.At home, Sean’s mother also reportedthat he had difficulty completing things, wasoften daydreaming, and seemed to have littlemotivation for anything.Assessment of Core SymptomsThe core ADHD behaviors were assessed bystructured interviews (DISC-IV) completedby Sean’s parent and teacher and by parent(CBCL) and teacher (CBRSC, APRS)behavior rating scales. On structured interviews,both parent and teacher reported thatSean showed significant problems of inattentionand disorganization, such as beingvery distractible, frequently daydreaming,having difficulty finishing tasks, often seemingunmotivated, and seeming very sluggishand drowsy. Although Sean was described byhis mother as somewhat fidgety, neither hismother nor his teacher reported significantproblems of impulsivity or overactivity.Consistent with these reports on the structuredinterviews, Sean was rated as showing(Continues)

388CHAPTER 17 Assessment of Attention Deficit HyperactivityBox 17.2 (Continued)attentional problems on parent and teacherrating scales out of a normative range. On theCBCL completed by his mother, Sean hada T-score of 69 on the Attention Problemsscale. On the teacher-completed CBRSC, hehad T-scores of 76 and 71 on the SluggishTempo and Daydreams scales, respectively.He also had a T-score of 69 on the Inattention-Disorganizationscale of the CBRSC.Consistent with a diagnosis of ADHD-InattentiveType, his teacher’s rating of MotorHyperactivity on the CBRSC was within anormative range (T-score of 57).Assessment of ComorbiditiesOn parent and teacher structured interviews,Sean was reported as having some signs of mildanxiety, including frequent stomachaches, selfconsciousbehaviors, and concerns about hisappearance. These symptoms did not seemsevere enough to warrant a diagnosis of an anxietydisorder and they did not appear out of agenormativeranges on the CBCL, on the CBRSC,or on the self-report Revised Children’s ManifestAnxiety Scale (Reynolds & Richmond, 1985).A psychoeducational assessment did notreveal any evidence of a learning disability.Sean’s scores on both in standardized intelligencetest and achievement test were allwithin age-appropriate limits. An assessmentof Sean’s peer relations did not indicateany problems in this psychological domain,based on parent and teacher rating scales anda sociometric exercise conducted with Sean’sclassmates. Similarly, there were no indicationsfrom any assessment source that Seanexhibited significant conduct problems.Ruling Out Alternative CausesA birth, medical, and developmental historyobtained from Sean’s mother did not revealany indications of significant medical or neurologicalproblems. He reportedly had somedifficulties breathing immediately followingbirth, but he was quickly stabilized with oxygen.Also, he had some mild allergies to dustand pollen, but these were not severe enoughto warrant medication.Although Sean exhibited some anxiety symptoms,they did not seem severe enough to becausing his problems in attention. Also, his anxietyseemed to be focused largely around school(e.g., stomachaches on school days, worry abouttests). Therefore, it seemed more likely thatSean’s anxiety was secondary to the academicproblems caused by his attentional difficulties.Assessing Core BehaviorsGuided by the research on classification ofADHD, the first goal of the assessment isto assess the core features of ADHD. Manyof the behavior rating scales (Chaps. 6 and7) and structured interviews (Chap. 11) discussedin previous chapters provide scales orsections that assess for these core behaviors.These behaviors can also be assessed throughbehavioral observations (Chap. 8; see alsoPelham et al., 2005; Smith et al., 2007).In selecting specific scales to assess thecore features of ADHD, several key factorsshould be considered. First, one should payclose attention to item content. Many measuresare based on outdated conceptualizationsof ADHD or the scales were formedstrictly on the basis of statistical covariation,without any guiding theory to help in scaledefinition. As a result, many scales that purportto measure behaviors associated withADHD also include behaviors that are notcurrently viewed as being part of the primarysymptom clusters. Unfortunately,these less relevant behaviors are intermixedwith core behaviors, and, as a result, manymeasures have no pure indicator of ADHDsymptoms. Examples of several commonlyused omnibus rating scales with subscalesthat assess ADHD behaviors are presentedin Box 17.3 to illustrate this point. Evidentfrom this table is the fact that a child mayhave elevations on many scales that purportto measure ADHD without actually showingthe core features of the disorder. Therefore,

CHAPTER 17 Assessment of Attention Deficit Hyperactivity389when selecting a scale to use in an assessmentof possible ADHD, an important considerationis whether it contains scales thatare largely composed of ADHD symptoms.Further, when interpreting scale elevationson a given rating scale, one must be surethat the elevations were actually due to thebehaviors associated with ADHD.Second, many measures simply do nothave sufficient coverage of the core ADHDbehaviors to aid in the differentiation of thetwo subtypes of ADHD. Often behaviorsindicative of inattention-disorganization areintermixed with impulsive and overactivebehaviors, providing no method of determiningsubtypes. For example, from Box 17.3one notices that the ASEBA (Achenbach,2001) Attention Problems scale includesitems associated with inattention-disorganizationand impulsivity-motor hyperactivity.As a result, use of this overall clinical scaledoes not aid in distinguishing subtypes ofADHD. Typically, structured interviews thatare tied to diagnostic classification systemsand are updated as the classification systemis updated have the best symptom coverage.Structured interviews also allow one to determinethe duration and stability of ADHDbehaviors, which, as discussed previously, iscrucial in the assessment of ADHD.Box 17.3Commonly Used Behavior Rating Scales with Subscales Related to ADHD:An Illustration of Item HeterogeneityMany of the more commonly used parentand teacher rating scales have very heterogeneousitem content with respect toADHD. Most of the scales that purportto measure constructs related to ADHDcontain a large number of behaviors notconsidered to be part of the core symptomsof ADHD (see Table 17.1). Toillustrate this, we list the item content ofrelevant subscales from a few commonlyused behavior ratings scales and highlightin boldface type the items that are notdirectly tied to the two core dimensionsof ADHD.Achenbach System for Empirically Based Assessment (Achenbach, 2001)Cross Informant ScalesAttention ProblemsAttention Deficit Hyperactivity Problems(9 items) (7 items)Acts too youngFails to finish thingsFails to finish thingsDifficulty concentratingDifficulty concentratingCan’t sit stillCan’t sit stillActs impulsivelyOften confusedInnattentiveDaydreamsTalks too muchActs impulsivelyLoudPoor school performance ---------------------Innattentive14% not part of core ADHD symptoms------------------------44% not part of coreADHD symptoms

390CHAPTER 17 Assessment of Attention Deficit HyperactivityBehavior Assessment System for Children – 2nd Edition (Reynolds & Kamphaus, 2004)Parent Rating ScaleTeacher Rating ScaleAttention Problems(6 items)Hyperactivity(8-items)Attention Problems(7 items)Hyperactivity(11 items)Pays attention Cannot wait turn Pays attention Cannot wait turnHas a short attention Acts without thinking Has a short attention Acts without thinkingspanHas poor self-control spanListens to directions Interrupts parents Listens to directions Has poor self-controlwhen they are on the Does not pay attentionto lectures doingSeeks attention whilephoneschoolworkPays attention whenspoken toListens carefullyIs easily distracted------------------0% not pare of coreADHD symptomsActs out of controlInterrupts others whenthey are speakingFiddles with thingswhile at mealsDisrupts other children’sactivities------------------50% not part of coreADHD symptomsListens carefullyIs easily distractedIs easily distractedfrom class work------------------0% not part of coreADHD symptomsActs out of controlInterrupts others whenthey are speakingDisrupts the schoolworkof other childrenDisrupts other children’sactivitiesHas trouble stayingseatedIs overly activeCalls out in class------------------45% not part of coreADHD symptomsConners Rating Scales – 3rd Edition (Conners, 2008)ParentInattention (10 items)Has trouble staying focused on one thingat a timeHas a short attention spanAvoids or dislikes things that take a lot of effortand are not funHas trouble concentratingDoesn’t pay attention to details; makescareless mistakesHas trouble changing from one activityto anotherTeacherInattention (11 items)Has a short attention spanDoesn’t pay attention to details;makes careless mistakesGives up easily on difficult tasksIs sidetracked easilyAvoids or dislikes things that take a lot of effortand are not fun(Continues)

CHAPTER 17 Assessment of Attention Deficit Hyperactivity391Conners Rating Scales – 3 (Continued)ParentInattention (10 items) cont.Inattentive, easily distractedGives up easily on difficult tasksHas trouble keeping his/her mind on workor play for long10% not part of core ADHD symptomsTeacherInattention (11 items) cont.Gets boredHas trouble concentratingInattentive, easily distractedHas trouble changing from one task toanotherHas trouble keeping his/her mind on work or play-------------------------------------------------------for long 18% not part of core ADHD symptoms-------------------------------------------------------Hyperactivity/Impulsivity (14 items)FidgetingBlurts out answers before the question hasbeen completedIs constantly movingExcitable, impulsiveGets over-stimulatedBlurts out the first thing that comes to mindHas difficulty waiting for his/her turnRuns or climbs when he/she is not supposed toIs noisy and loud when playing our using freetimeLeaves seat when he/she should stay seatedHyperactivity/Impulsivity (18 items)Leaves seat when he/she should stay seatedGets overly excitedFidgets or squirms in seatRestless or overactiveBlurts out answers before the question has beencompletedExcitable, impulsiveActs as if driven by a motorRuns or climbs when he/she is not supposed toTalks out of turnInterrupts others (e.g., butts into conversations orgames)Fidgets or squirms in seatRestless or overactiveIs noisy and loud when playing or using free timeInterrupts others (for example, butts into conversationsor games)Gets over-stimulated or “wound up”Talks too muchFidgeting---------------------------------------------------7% not part of core ADHD symptoms Is constantly movingGets up and moves around during lessonsHas difficulty waiting for his/her turnTalks non-stop-----------------------------------------------------11% not part of core ADHD symptoms(Continues)

392CHAPTER 17 Assessment of Attention Deficit HyperactivityPersonality Inventory for Children -2(Lachar & Gruber, 2001)Impulsivity and Distractibility (27 items)My child’s manners sometimes embarrass meSchoolteachers complain that my child cannotsit stillMy child’s behavior often makes othersangryMy child often does not finish things thathe/she startsMy child jumps from one activity to anotherMy child often acts without thinkingMy child is often restlessI cannot get my child to do his/her schoollessonsMy child often forgets to do thingsMy child often nags and bothers otherpeopleMy child cannot wait for things like otherchildren doMy child cannot keep attention on anythingMy child does not learn from his/hermistakesMy child is almost always on time andremembers what he/she is supposed to doMy child cannot sit still in school becauseof nervousnessThe school says that my child needs helpin getting along with other childrenMy child usually runs rather than walksMy child tends to swallow food withoutchewing itRecently the school has sent home notesabout my child’s bad behaviorMy child seems more clumsy than otherchildren his/her ageMy child will do anything on a dareMy child brags about being sent to theprincipal at schoolNothing seems to scare my childMy child likes to show offMy child tends to bragMoney seems to be my child’s biggestinterest--------------------------------------------63% not part of core ADHD symptomsStudent Behavior Survey (Lachar, Wingenfeld,Kline, & Gruber, 2000)Attention Deficit Hyperactivity (16 items)Completes class assignmentsDemonstrates logical approach to learningFollows teacher’s directionsMaintains alert and focused attentionPersists even when activity is difficultRemembers teacher’s directionsStays seatedWaits for turnWorks independently without disturbingothersListens to other studentsDaydreamsInterrupts othersImpulsiveMisbehaviors unless closely supervisedOveractiveTalks excessively--------------------------------------------25% not part of core ADHD symptoms

CHAPTER 17 Assessment of Attention Deficit Hyperactivity393There are several ratings scales that havebeen designed solely to assess behaviorsassociated with ADHD and provide muchmore extensive coverage of behaviors associatedwith this disorder. They include theADHD Rating Scale IV (DuPaul, Power,Anastopoulos, & Reid, 1998) and the DisruptiveBehavior Disorders Rating Scale(Pelham et al., 2005), both of which weredesigned to provide a complete coverageof the diagnostic criteria for ADHD. TheBrown Attention Deficit Disorder Scalesfor Children (Brown, 2001) and the AttentionDeficit Disorder Evaluation Scales(ADDES-3; McCarney, 2004) also provideextensive coverage of ADHD symptoms,although they were not explicitly tied toDSM-IV-TR criteria.Third, when selecting measures toassess the core features of ADHD, onemust obtain information from multiplesources (parents and teachers). This helpsone determine the situational variabilityof behaviors. Also, if the core features ofADHD can be assessed through multiplemodalities (e.g., rating scales and behavioralobservations), this negates the needto rely on any single imperfect assessmentinstrument. Therefore, one must choosea set of assessment instruments that providesa multi-informant and multi-methodassessment of the core ADHD behaviors.Fourth, one should have informationthat allows comparison to age norms. Mostdefinitions of ADHD either implicitly orexplicitly state that the symptoms shouldbe inconsistent with a child’s developmentallevel (American Psychiatric Association,2000). Therefore, an assessment mustprovide information that allows one tocompare a child’s behaviors to the behaviorsof other children of a similar developmentallevel. Typically, behavior ratingscales are best suited for this task becauseof their extensive normative base. However,one important caution is in order inusing norm-referenced rating scales. Manyscales (e.g., ASEBA) often only providenorm-referenced scores broken down byage and gender. Definitions of ADHD donot make the restriction that the behaviorsbe inconsistent for a child’s gender. Infact, it is well accepted that boys are fourto six times more likely to show ADHDthan girls (American Psychiatric Association,2000). Using gender-specific normsignores this widely found and acceptedgender ratio and artificially equates thenumber of girls and boys with significantlevels of ADHD behaviors. This will leadto more girls and fewer boys being consideredto have significant ADHD behaviorsthan if cross-gender norms are used.Fifth, researchers studying children withADHD and attempting to define the coredeficit that may underlie this disorder (e.g.,response inhibition, sustained attention)have frequently used laboratory measures ofinattention and impulsivity in their assessments.These measures place children instandardized conditions and attempt toquantitatively measure their inattentiveness,impulsiveness, or related behaviorsunder these conditions. For example, oneof the most frequently used laboratorymeasures for the assessment of ADHD isthe Continuous Performance Test (CPT).There are many variations of the CPTbut a prototypical CPT is the one developedby Gordon (1983). In the GordonCPT, a child is told to view a screen onwhich numbers are presented. The childis instructed to press a button each time apredetermined number is presented. Tworesponses are measured. Omissions are thenumber of times the designated numberis presented to the child and the child failsto press the button. Number of omissionsis considered to be a measure of sustainedattention, especially increases in omissionsover time. Commissions occur when thechild incorrectly presses the button whenthe designated number is not presented andare considered a measure of impulsivity.There are many variations of the CPT(e.g., Conners, 1995), with some CPTtasks presenting auditory rather thanvisual stimuli and other tasks presenting

394CHAPTER 17 Assessment of Attention Deficit Hyperactivitydistracters (e.g., numbers flashing oneither side of the target stimuli that needto be ignored). Also, in addition to theCPT, there are many other laboratorymeasures that have been used in researchwith children who have ADHD. Rapport,Chung, Shore, Denney, and Isaacs (2000)review 142 studies using over 40 differentlaboratory measures to assess characteristicsassociated with ADHD. They identifyseveral characteristics of the tasks thatmost consistently differentiate childrenwith ADHD from other children. First,the tasks that most consistently showgroup differences rely on recognition,recall, or both (e.g., the letter or word thata child is supposed to look for is not continuouslydisplayed for the child), and theyoften involve some speed of processingcomponent. Also, most of the tasks thatreliably differentiate children with ADHDfrom other children place special demandson a child’s working memory and eachof the tasks is experimentally paced, notallowing the child to control the speed atwhich stimuli are presented. Rapport et al.(2000) provide an interesting discussionas to how these different task parametersmay provide clues to the specific deficitsdisplayed by children with ADHD.It is evident from this rather extensiveliterature that these tasks have been quiteuseful in studying children with ADHD.These tasks also have a number of appealingqualities for clinical assessments aswell. For example, they can help bridgethe gap between assessment instrumentsbeing used in research that guide our currentconceptualizations of ADHD andthose commonly used in clinical practiceto make the diagnosis (Frick, 2000). Evenmore appealing, however, is that these laboratorytasks provide a potential means ofassessing the symptoms of ADHD that arenot based on the perceptions of others thatcould be biased.Unfortunately, despite this promisefor their clinical utility, there are a numberof limitations in their developmentthat make their clinical usefulness in thediagnosis of ADHD somewhat limited atpresent (Pelham et al., 2005; Smith et al.,2007). For example, the primary validationof these tasks has been to differentiateADHD children from non-ADHDchildren or to show the effects of stimulantmedication on the task performanceof children with ADHD (Rapport et al.,2000). Unfortunately, both types of studiesrely on group-level data that are difficultto translate into findings that aremeaningful for interpreting an individualchild’s score (e.g., How many children withADHD do not score high on the laboratorymeasure? What percentage of treatedchildren show the response to medicationon the task?). Furthermore, it is not clearwhether performance differences on thesetasks are specific to children with ADHDor whether children with other types ofbehavioral or emotional disturbances alsoshow similar problems in performance.Without such information, it is difficultto use these measures to make differentialdiagnoses between ADHD and otherforms of psychopathology.Finally, scores from the laboratorymeasures generally show low correlationswith behavioral observations and parentand teacher reports (e.g., interviews andrating scales) of ADHD symptoms (Barkley,1991; DuPaul, Anastopoulos, Shelton,Guevremont, & Metevia, 1992). Furthermore,there is very little information as towhether or not the laboratory measuresadd any clinically useful information tothese more ecologically valid measuresof behavior (Pelham et al., 2005; Rapportet al., 2000). For example, if a childis reported as showing significant problemswith ADHD symptoms accordingto parent and teacher reports and basedon behavioral observations in the classroom,interventions for his behavioralproblems in his natural environment willlikely be the same, irrespective of his performanceon the laboratory task. Becausetreatment decisions are largely based on

CHAPTER 17 Assessment of Attention Deficit Hyperactivity395the more ecologically valid measures, therole of laboratory measures in clinicalassessments is uncertain at this point. InBox 17.4, we summarize Barkley’s (1991)review of laboratory measures of ADHDsymptoms, which provide a more indepthdiscussion of these measures’ ecologicaland incremental validity.Box 17.4Research Note: Ecological Validity of Laboratory Measures of Attention andImpulsivityBarkley (1991) provides a critical discussionof laboratory measures (LM) used to assessattention and impulsivity, two aspects of thecore symptoms of ADHD. The focus of muchof Barkley’s discussion is on reaction timetasks (RTT), continuous performance tasks(CPT), and the Matching Familiar FiguresTest (MFFT). Although Barkley also discussesanalogue observations of motor activity, thesewere discussed previously in Chap. 8.Concept of Ecological ValidityThe first issue addressed in this article is theconcept of ecological validity. Barkley definedecological validity as “the degree to whichLMs represent the actual behaviors of interest(i.e., inattention and impulsivity) as they occurin naturalistic settings” (p. 150). This aspectof validity is crucial for clinical assessments. Incontrast, most research projects have the goal ofdetermining the “core deficit” in ADHD, whichmay not be manifested in the natural setting.Therefore, a LM “need not be ecologically validto be useful in research on ADHD” (p. 151).Evidence for Ecological ValidityOne type of research on the ecological validityof LMs consists of studies testing group differenceson LMs between ADHD and controlchildren. In general, Barkley’s review suggestedthat all of the LMs have consistently differentiatedADHD from normal control children.However, most studies indicate that these effectsare weakened or altogether eliminated whendifferences in intellectual level are controlled.Also, LMs do not differentiate ADHD childrenfrom other clinic-referred children. As a result,scores on the LMs are not useful in makingdifferential diagnoses within clinic referrals.A second type of research investigates theeffect of stimulant medication on a child’s performanceon the LMs. Barkley’s conclusion wasthat the research evidence was mixed. Severalstudies found significant effects for stimulantmedication on commonly used LMs, whereasmany studies found no such reliable effect.A third type of research investigates thecorrelations between scores on LMs and parentand teacher ratings of ADHD behavior.Studies have generally found significant butmodest (.21–.51) correlations between LMmeasures of attention and impulsivity andparent and teacher ratings of behaviors consideredto be indicative of these constructs.ConclusionsBarkley concludes that the ecological validity ofLMs should be considered limited based on theexisting research. As a result, clinical researchshould avoid using LMs as the “gold standard”against which other measures of attention andimpulsivity are judged. For clinical assessments,one should recognize the limited usefulness ofLMs in a battery of assessment tests. Most clinicaldecisions are based on the more ecologicallyvalid measures of parent and teacher report of achild’s behavior in the naturalistic setting. Barkleyconcludes, “Where LM results conflict withthose obtained from other sources, such as parentand teacher behavior ratings, history, andobservations in natural settings, the LM resultsshould probably be disregarded in favor of thesemore ecologically valid sources” (p. 173). As aresult, the incremental benefit of adding LMsto an assessment battery seems to be minimal.Source: Barkley, R. A. (1991). The ecological validity of laboratory and analogue assessmentmethods of ADHD. Journal of Abnormal Child Psychology, 19, 149–178.

396CHAPTER 17 Assessment of Attention Deficit HyperactivityOne final note is in order for assessing thecore features of ADHD. Despite widespreadcautions against the practice (e.g., Barkley, 1990;Kamphaus, 2001; Kaufman, 1994), clinical assessorscontinue to use the Freedom from Distractibility(FD) factor from the Wechsler IntelligenceScales for Children (Wechsler, 1991) as an indicatorof the presence of ADHD symptoms. Thisinappropriate use of FD is largely maintainedsimply because of the name given to the factor,which is a significant source of distress to the personwho first applied the FD label to the WISCsubtests. Alan Kaufman states, “The label shouldhave been trashed years ago. I cringe whenever Iread it” (Kaufman, 1994, p. 212).As Kaufman goes on to explain, theuse of the FD as a measure of inattentionor distractibility is inappropriatebecause these scales are affected by multipleemotional and behavioral factors, notjust distractibility. To illustrate the point,Kaufman reviewed 19 studies in which theFD was lower in samples of children withlearning disabilities, children with leukemia,children with emotional difficulties,heterogeneous psychiatric samples ofinpatient or outpatient children, childrenwith autism, children with schizophrenia,children with Conduct Disorder, and childrenwith muscular dystrophy. The authorthen makes the cogent point that the FDfactor is important because of its robustability to differentiate the abnormal (medically,behaviorally, and educationally) fromthe normal population, but it is basicallymeaningless for identifying a specific typeof exceptionality, such as ADHD.Assessing Comorbid ProblemsMany of the problems that often co-occurwith ADHD can be assessed in conjunctionwith the assessment of the core ADHDbehaviors. For example, many of the omnibusrating scales and structured interviewsdiscussed in previous chapters includeitems that assess for conduct problems,anxiety, depression, self-esteem, and socialcompetence. Like the ADHD behaviorsthemselves, potential co-occurring problemsare best assessed through multipleinformants and using multiple formats.Given the overlap with learning problems,psycho-educational testing should also be apart of most ADHD assessments. Learningdifficulties are not reliably assessed throughrating scales, interviews, behavioral observations,or projective testing. As a result,standardized intelligence and achievementtests are often required as part of a comprehensiveevaluation for ADHD.Some omnibus behavior rating scalesinclude items that assess the degree of familyconflict and other aspects of a child’s familycontext. However, as discussed in Chap. 12,some assessments may require a more indepthassessment of specific areas of familyfunctioning that are better obtained throughmethods that specifically focus on the child’sfamily context. Each aspect of family functioningthat was highlighted in Chap. 12(parenting styles and behaviors, parentingstress, marital conflict, and parental adjustment)is important in understanding thefamily context of a child with ADHD.Assessing Potential AlternativeCausesOne of the most difficult aspects of assessingfor ADHD is ruling out alternative causes forthe symptoms. Probably, the most importantpiece of information for this purpose hasalready been discussed. If one has adequatelydetermined that the symptoms of ADHD areof sufficient number, severity, and durationand cause enough impairment to warrant adiagnosis, most of the alternative explanationsfor symptoms can be ruled out as a soleexplanation for the behaviors. Many of thealternative explanations for ADHD symptomsresult in behaviors similar to ADHD,but of lower intensity and of shorter durationthan is typical for children with ADHD.Given that some medical and neurologicaldisorders can manifest in problems of

CHAPTER 17 Assessment of Attention Deficit Hyperactivity397inattention-disorganization and impulsivityhyperactivity,and given that such behaviorscan be side effects of certain medications, itis important that a thorough developmentaland medical history be obtained on a childwhen assessing for ADHD. When the historyis suggestive of the possible presenceof a medical or neurological disorder, thechild can be referred to a physician for amore comprehensive medical and/or neurologicalexam. However, most experts onADHD feel that a full medical exam neednot be a standard part of a diagnostic assessmentof ADHD, because the diagnosis isprimarily based on behavioral data (Barkley,1997a; Pelham et al., 2005).The greatest difficulty in ruling outalternative explanations for ADHD symptomsis the problem of determining whatis primary and what is secondary. Basedon the research on comorbidities, weknow that ADHD children are at risk formany other problems in adjustment, mostof which can also mask as ADHD (e.g.,learning disabilities, emotional disorders).The distinction between primary and secondaryis largely a clinical one, based ona complex weighing of various pieces ofassessment information. To summarize oursuggestions from Chap. 15, several typesof information may be helpful in makingthis difficult case formulation. Consideringthe level of impairment associated withdifferent areas of dysfunction (i.e., whichproblem areas seem to be causing the mostproblems for the child) and the temporalsequencing of problem behaviors (i.e.,which problems seem to predate others)can help in distinguishing primary or secondaryareas of dysfunction. In addition,viewing family history data and determiningif a child might be at risk for a certaintype of problem given its occurrence inrelatives may also aid in making differentialdiagnoses. Specifically, for ADHD,determining if a child’s parents or otherfirst-degree relatives had childhood historiesof ADHD symptoms can providevaluable information in making an ADHDdiagnosis. In contrast, a positive familyhistory for bipolar illness might warrantfurther assessment to determine if a child’smotor restlessness and problems of impulsivecontrol are early indicators of an affectivedisturbance (Smith et al., 2007).ADHD in the Schools: SpecialEducation PlacementADHD is a psychological syndrome thatrequires collaboration between professionalsacross multiple specialties (i.e., psychology,education, medicine) for assessmentand treatment. Clinical assessors must beable to effectively utilize the expertise ofother professionals and tailor their assessmentsto provide useful information tomany different disciplines. Because ADHDoften has its most dramatic and noticeableeffect on a child in the school setting, collaborationwith educators is particularlycrucial. It is essential that assessments forADHD be designed to provide informationhelpful to educators in developingappropriate interventions.We feel strongly that not all childrenwith ADHD require special educationplacement. There is ample evidence thatsome children with ADHD can successfullyfunction in a regular education classroomwith medication, with specific modificationsof the classroom environment, and/or with structured behavioral interventionsdesigned to reduce the disruptive behaviors(see Abramowitz & O’Leary, 1991). However,designing a successful interventionprogram for a child with ADHD requires aclear documentation of a child’s strengthsand weaknesses (behaviorally, emotionally,cognitively, and academically). Simplyknowing that a child has ADHD gives educatorsonly limited information on whichto base interventions, given the greatdiversity within children with ADHD. As

398CHAPTER 17 Assessment of Attention Deficit Hyperactivitya result, clinical assessments should clearlyoutline an individual child’s competenciesand deficits, with suggestions on how theseabilities will influence a child’s functioningin a classroom setting.Unfortunately, some children withADHD may require more intensive educationalservices, such as those offered inspecial education programs. To serve childrenin special education programs, schoolsystems must operate under federal andstate guidelines, the former of which mayvary in implementation from state to stateand the latter of which may vary both incontent and implementation. It is imperativethat clinical assessors understand theguidelines under which a school system isoperating, in order to design assessmentsand make recommendations that enhancea school’s ability to appropriately meet theeducational needs of a child.The primary piece of federal legislationthat guides provision of special educationservices to children and adolescents is theIndividuals with Disabilities Act (IDEA),which was originally passed in 1975 asPublic Law 94–142 and amended andrenamed in 1990 and 2004. IDEA, and themore recent IDEIA, mandates that childrenwith disabilities be provided speciallydesigned instruction and related services inthe least restrictive environment (i.e., closecontact with children without disabilities)necessary for a child to learn. Providingservices to children with ADHD underIDEA has been a source of contention inmany school systems, because ADHD isnot listed as one of the disabilities explicitlycovered under IDEA. However, manychildren with ADHD have secondary featuresthat may allow them to be servedunder IDEA guidelines, such as speech orlanguage impairments, emotional disturbance,or specific learning disabilities. Infact, most of the children with ADHD whorequire intensive special education servicesdo so, not because of the ADHD itself,but because of the additional disruptionscaused by these secondary features.However, it is possible that some childrenwith ADHD may need special educationservices but do not qualify underIDEA guidelines. These children can beserved under a civil rights law, Section 504.Section 504 is part of the RehabilitationAct Amendments of 1973 (PL93–112) andwas designed to protect individuals withhandicaps. Section 504 specifically mandatesthat,No otherwise qualified individual with handicapsin the United States, shall, solely by reasonof her or his handicap, be excluded fromparticipation in, denied the benefits of, or besubjected to, discrimination under any programor activity receiving Federal financial assistance(29 U.S.C. Sec. 794).It is generally accepted that ADHD qualifiesas a handicap under Section 504 (Madsen,1990) and, therefore, schools mustmake accommodations for the individualneeds of children with ADHD under thisstatute. Unfortunately, unlike IDEA, Section504 does not allocate federal fundingfor educational interventions.This discussion of special education lawsmay seem irrelevant or at least peripheralfor clinical assessors who operate outside ofthe school system. In fact, many such assessorsprefer to remain ignorant of such legalguidelines, so as not to be confined to thelimits delineated in such laws. However,if one wants to aid a child with ADHD inreceiving needed educational services, oneshould understand the legal guidelines soas to be able to work collaboratively withschool personnel in developing an appropriateeducational plan.Conduct ProblemsClassification and DiagnosisLike ADHD, conduct problems in childrenrepresent a critical mental health concern.These problems are highly disruptive

CHAPTER 17 Assessment of Attention Deficit Hyperactivity399to others in a child’s environment can bepredictive of problems later in life, includingcriminal behavior (Frick & Kimonis,2008). Also like ADHD, there is considerableagreement that children with conductproblems are a heterogeneous group (Frick,2006). Therefore, a substantial body ofresearch has been directed at determiningthe most appropriate method of classifyingconduct problems into meaningful subtypes.The most commonly used method ofclassifying conduct problems in childrenis a two-dimensional approach originallydescribed in DSM-III and continuedwith some modifications in the laterrevisions of this manual. This systemdivides conduct problems into two syndromes:Oppositional Defiant Disorder(ODD) and Conduct Disorder (CD).ODD refers to a pattern of negativistic,oppositional, and stubborn behaviors,whereas CD refers to more severeantisocial and aggressive behaviors thatinvolve serious violations of others’rights (e.g., aggression, destruction ofproperty) or deviations from major ageappropriatenorms (e.g., running awayfrom home, truancy). A summary of thebehaviors indicative of these two dimensionsis provided in Table 17.4.A detailed discussion of the validity ofthe ODD/CD distinction is beyond thescope of this chapter (see McMahon &Frick, 2007). However, the relationshipbetween ODD and CD behaviors is importantfor a number of reasons. First, thereappears to be a hierarchical relationshipbetween the two diagnoses. That is, mostchildren with the more severe symptomsof CD also show the symptoms of ODD(Lahey & Loeber, 1994; Spitzer, Davies,& Barkley, 1990). However, the converseis not true. There are many children withODD who do not show the more seriousconduct problems associated withCD. Second, there seems to be a developmentalrelationship between ODD andCD. A 3-year longitudinal study of clinicreferredboys found that 82% of the newcases of CD (n = 22) that emerged duringthe study period had received a diagnosisof ODD in the preceding year (Lahey,Loeber, Quay, Frick, & Grimm, 1992).Therefore, ODD behaviors can be viewedas a risk factor for the development of themore severe CD.Table 17.4 Two-Dimensional Classification of Conduct ProblemsOppositional Defiant DisorderConduct DisorderLoses temperBullies or intimidates othersArgues with adultsInitiates physical fightsActively defies adultsHas been physically cruel to othersRefuses adults’ requests or rules Steals things of nontrivial valueDeliberately annoys others Forced someone into sexual activityBlames others for mistakes Stays out after dark without parental permission, before age 13Is angry and resentfulLies to obtain goods or favorsIs spiteful and vindictiveHas been physically cruel to animalsIs touchy and easily annoyed Has deliberately destroyed others’ propertyHas set fires with intention of causing serious damageHas run away from home overnight more than onceIs often truant from school, beginning before age 13Has broken into someone’s house, building, or car

400CHAPTER 17 Assessment of Attention Deficit HyperactivityComorbiditiesThe most common problem co-occurringwith conduct problems is ADHD. In ameta-analytic study of community studies,Waschbusch (2002) reported that 36% ofboys and 57% of girls with conduct problemshad comorbid ADHD. When studyingpre-adolescent children and childrenreferred to clinics, the rate is even higherand often ranges from 75 to 90% (Abikoff& Klein, 1992; Hinshaw, 1987). Thisdegree of overlap has led to a debate as towhether or not ADHD and conduct problemsshould even be considered separatepsychological domains (see Rutter, 1983).We feel that the research indicates thatthese domains are at least partially independent(Frick, 1994; Hinshaw, 1987).Research does suggest that the developmentof ADHD usually precedes the developmentof conduct problems (Waschbusch,2002) and it often signals the presence of amore severe and more chronic form of conductproblems in children (Frick & Loney,1999). Also, there is growing evidence thatconduct problems improve when childrenwith ADHD have been treated with stimulantmedication (Hinshaw, 1991). Therefore,clinical assessments of children withconduct problems should routinely assessfor the presence of ADHD.Children with conduct problems alsofrequently have a comorbid anxiety disorder,and this seems to be especially thecase for girls (Loeber & Keenan, 1994).Children with conduct disorders alsoshow a high rate of depression (Harrington,Fudge, Rutter, Pickles, & Hill,1991). Importantly, there is evidence thatchildren with both conduct problems anddepression show a high rate of suicidalideation (Capaldi, 1992). The combinationof suicidal ideation, depression, andpoor impulse control that often is presentin children with conduct problemshas been associated with increased risk forsuicide (Shaffer, Garland, Gould, Fisher,& Trautman, 1988). Finally, childrenwith CD are at higher risk for substanceabuse, especially those children with bothADHD and CD (Lynskey & Fergusson,1995; Thompson, Riggs, Mikulich, &Crowley, 1996).Research also indicates that approximately20–25% of children with CD areunderachieving in school relative to a levelpredicted by their age and intellectualabilities (Frick et al., 1991). The reasonfor this association is not clear, possiblybecause the mechanisms involved maydiffer depending on the age of the samplestudied (Hinshaw, 1992). For example, inelementary school-age samples, much ofthe overlap between CD and academicunderachievement seems to be due tothe presence of ADHD (Frick et al.,1991). However, learning difficultiesseem to predict adolescent-onset conductproblems independent of other factors(Hinshaw, 1992). Despite the lack of adefinitive explanation for the correlationbetween learning and conduct problems,the simple fact that they co-occur so consistentlywarrants the assessment of learningproblems when assessing children andadolescents with conduct problems.Correlates with PotentialCausal RolesMost researchers agree that conduct problemsare the result of a complex interactionof multiple causal factors (Frick, 2006).Identifying the important causal agentsand how they interact to cause conductproblems is still an area in need of moreresearch. Past research has uncovered severalfactors that are associated with conductdisorders and that likely play a role in theirdevelopment. These factors can be summarizedin five categories: biological factors,cognitive correlates, family context,social ecology, and peers. The research onthe biological correlates of conduct problemsin children, while crucial for developingcausal theories, is not reviewed here

CHAPTER 17 Assessment of Attention Deficit Hyperactivity401because the current state of knowledge isnot sufficiently developed to have clearimplications for assessment (see Dodge &Pettit, 2003; Lahey, Hart, Pliszka, Applegate,& McBurnett, 1993; Raine, 2002).In contrast, there are several aspects ofthe child’s cognitive and learning styles thathave been associated with conduct problemsand aggression that may be importantto the assessment process (see Frick& Loney, 2000). First, in general, childrenwith conduct disorders tend to score lowon intelligence tests, especially in the areaof verbal intelligence (Loney, Frick, Ellis,& McCoy, 1998; Moffitt, 1993). Second,many children with serious conduct problemstend to show a learning style that ismore sensitive to rewards than punishments.This has been labeled as a rewarddominantresponse style that could explainwhy many of these children persist in theirmaladaptive behaviors, despite the threatof serious potential consequences (Fisher& Blair, 1998; O’Brien & Frick, 1996).Third, many children with conduct problemsshow deficits in their social cognition,which is the way they interpret social cuesand use them to respond in social situations.For example, children with conductproblems are more likely to attribute hostileintent to the actions of peers and areless able to develop nonaggressive responsealternatives in situations involving peerconflict, both of which could make thechild more likely to respond aggressivelyin social situations (Crick & Dodge, 1996;Dodge & Pettit, 2003).While these cognitive correlates haveplayed an important role in many theoriesof how conduct disorders develop,and they are important targets of interventionfor many treatment programs(e.g., Lochman, Wells, & Lenhart, 2008),there is probably no set of correlates thathas been as important to theory and treatmentas family dysfunction. There seemto be at least three dimensions of familyfunctioning that are consistently relatedto childhood conduct problems: parentalpsychiatric adjustment, marital instability/divorce, and parental socialization practices(see Frick, 1994). A meta-analysis ofthe research on the relationship betweenfamily functioning and conduct disordersin youth found that parental socializationpractices were especially important(Loeber & Stouthamer-Loeber, 1986). Tobe specific, parental involvement in theirchild’s activities, parental supervision oftheir child, and the use of harsh or inconsistentdiscipline tended to show the strongestrelationships with conduct problemsin children of all the variables included inthe meta-analysis.Another clinically important class ofcorrelates is comprised of factors withinthe child’s larger social ecology that mayplay a causal role in the development ofconduct problems. One of the most consistentlydocumented of these correlateshas been low socio-economic status (Frick,Lahey, Hartdagen, & Hynd, 1989). However,several other ecological factors, manyof which are related to low socio-economicstatus, such as poor housing, poor schools,and disadvantaged neighborhoods, havealso been linked to the development of conductproblems in children (see Frick, 1998;Peeples & Loeber, 1994). In addition, thehigh rate of violence, witnessed by childrenwho live in impoverished inner-cityneighborhoods, has also been linked to thedevelopment of conduct problems (Osofsky,Wewers, Hann, & Fick, 1993).Finally, research has documented a relationshipbetween peer rejection in elementaryschool and the later development ofconduct problems (Roff & Wirt, 1984). Inaddition, peer rejection in elementary schoolis predictive of an association with a deviantpeer group (i.e., one that shows a highrate of antisocial behavior and substanceabuse) in early adolescence (Fergusson,Swain, & Horwood, 2002). This relationshipis important because association witha deviant peer group leads to an increase

402CHAPTER 17 Assessment of Attention Deficit Hyperactivityin the frequency and severity of conductproblems (Patterson & Dishion, 1985) andit has proven to be a strong predictor oflater delinquency and other negative outcomes,such as substance abuse (Dishion,Capaldi, Spracklen, & Li, 1995; Fergussonet al., 2002). Therefore, peer rejection maybe directly related to the development ofconduct problems but also may indirectlyinfluence conduct problems by increasingthe chance that the child or adolescent willassociate with deviant peers.Conceptual ModelFrom the preceding section, it is clear thatthere has been a great deal of research documentingmany characteristics of childrenwith severe conduct problems that areimportant to consider in conducting assessmentswith these children. Unfortunately,there has not been much agreement as to agood framework for organizing these manydiverse characteristics into a clear conceptualframework for understanding howconduct problems develop. Many theoristshave tended to focus on either one correlate(e.g., community violence) or oneclass of correlates (e.g., family dysfunction)without attempting to integrate the manydiverse and potentially interacting influencesthat play a role in the developmentof conduct problems. Others have viewedthese problems from a “cumulative risk”perspective, which acknowledges that anysingle factor will be limited for explainingthe development of conduct problems andthat their development in any child is likelythe result of the additive influence of manydifferent causal factors (Loeber, Burke,Lahey, Winters, & Zera, 2000).An alternative framework that is gainingsupport in research is a “developmentalpathway approach,” which proposesthat children may develop conduct problemsthrough many different causal trajectories,each involving somewhat differentinteractions of causal processes (Frick,2006). This approach recognizes that, forany child, the development of serious conductproblems is likely the result of multipleinteracting causal factors. In addition,a developmental framework explicitlyrecognizes that there may be distinct subgroupsof children with severe conductproblems with different causal processesunderlying their behavior. This approachis consistent with a long history of tryingto divide antisocial and delinquent youthinto distinct subgroups that differ in termsof behavior, associated characteristics,outcomes, and response to treatment (seeFrick & Marsee, 2006 for a review).Consistent with this research, the DSM-IV-TR specifies two subtypes of ConductDisorder (CD; American PsychiatricAssociation, 2000). Children in the childhood-onsetsubtype begin showing severeconduct problems prior to adolescence,often as early as preschool or early elementaryschool, and their behavioral problemsincrease in rate and severity over thechildhood years (Lahey & Loeber, 1994).In contrast, youth in the adolescent-onsetsubtype do not show significant behavioralproblems in childhood, but begin to exhibitsignificant conduct problems as they enteradolescence (Moffitt, 1993, 2003). Oneof the key differences between these twogroups of antisocial youth is that the childhood-onsetgroup is much more likely tocontinue to show antisocial and criminalbehavior into adulthood compared to theadolescent-onset group (Frick & Loney,1999; Moffitt, 2003). However, in additionto the differences in prognosis, researchhas uncovered several other characteristicsthat could suggest that the causal processesunderlying the antisocial behavior of thetwo groups are also different.Specifically, children in the childhoodonsetgroup are characterized by moreaggression, higher rates of cognitive (e.g.,lower verbal intelligence) and neuropsychological(e.g., executive functioning

CHAPTER 17 Assessment of Attention Deficit Hyperactivity403deficits) dysfunction, more disturbances intheir autonomic nervous system functioning,and more severe problems of impulsecontrol, often leading to higher rates ofdiagnosis of Attention-Deficit HyperactivityDisorder than children with the adolescent-onsetpattern of CD (Frick, 2006;Moffitt, 1993; 2003). The fewer pathogenicbackground factors, as well as thebetter adult outcome for the adolescentonsetsubtype, suggests that their conductproblems may be an exaggeration of a normativepattern of rebellious and antisocialbehavior related to the important tasksinvolved in identity development that takeplace in adolescence (Frick, 2006; Moffitt,2003; Silverthorn & Frick, 1999). In contrast,the childhood-onset group appearsto show a more severe type of dysfunctionthat extends beyond a single developmentalstage.Importantly, research has also indicatedthat there are some important distinctionsthat can be made within this childhoodonsetgroup in terms of the types of dysfunctionalprocesses that may be operating.The distinction is based on differentiatingbetween children who show a callous andunemotional interpersonal style and thosewho do not. Callous-unemotional traitsrefer to a lack of guilt over misdeeds, a lackof empathy toward others, and other deficitsin their emotional responses (Frick &Dickens, 2006; Frick & White, 2008).Children with conduct problems, whoalso show callous-unemotional traits,tend to show a more severe, more aggressive,and more stable pattern of conductproblems (Frick & Dickens, 2006). Further,these children with callous-unemotionaltraits tend to be more thrill andadventure seeking, they are less sensitiveto the effects of punishment compared tothe effects of rewards, and they are lessreactive to emotionally distressing stimulithan other children with childhoodonsetCD (Frick & White, 2008). All ofthese characteristics are consistent with atemperamental style associated with lowemotional reactivity that can (1) placea child at risk for missing some of theearly precursors to empathetic concernthat involve emotional arousal evokedby the misfortune and distress of others;(2) lead a child to be relatively insensitiveto the prohibitions and sanctionsof parents and other socializing agents;and (3) create an interpersonal style inwhich the child becomes so focused onthe potential rewards and gains involvedin using aggression to solve interpersonalconflicts that he or she ignores the potentiallyharmful effects of this behavior onhim- or herself and others (Frick & Morris,2004).In contrast to those youth with callousand unemotional traits, children withinthe childhood-onset group who do notshow these traits tend to show the oppositeextreme of emotional reactivity. Theytend to be highly reactive to emotional andthreatening stimuli and to respond morestrongly to provocations in social situations(Frick, 2006; Frick & Morris, 2004). Also,their aggressive and antisocial behavior ismore strongly associated with dysfunctionalparenting practices and with deficitsin verbal intelligence than the group thatis high on callous-unemotional traits (Frick& White, 2008). These findings suggestthat children with childhood-onset antisocialbehavior, but who do not show highrates of callous-unemotional traits, mayhave problems more specifically associatedwith poor behavioral and emotionalregulation, characterized by very impulsivebehavior and high levels of emotionalreactivity. Such poor emotional regulationcan result from a number of interactingcausal factors, such as inadequate socializationfrom families, deficits in their verbalintelligence that make it difficult for themto delay gratification and anticipate consequences,or temperamental problems inresponse inhibition such as those discussedin the previous section on ADHD. The

404CHAPTER 17 Assessment of Attention Deficit Hyperactivityproblems in emotional regulation can leadto very impulsive and unplanned aggressiveacts for which the child may be remorsefulafterward but that he or she still hasdifficulty controlling. It can also lead to achild being susceptible to becoming angry(i.e., emotionally aroused) due to perceivedprovocations from peers, leading to violentand aggressive acts within the context ofhigh emotional arousal.This developmental framework has anumber of implications for the assessmentprocess for children with severe conductproblems (McMahon & Frick, 2005). Itsuggests that, not only do interventionsfor children with severe conduct problemsneed to be comprehensive in targeting alarge number of diverse causal influences,but these interventions also need to be tailoredto the unique needs of specific subgroupsof children with conduct problems(see Frick, 2006 for a more a extendeddiscussion of this issue). In order to implementsuch a comprehensive and individualizedapproach to treatment, there needsto be a comprehensive assessment of thechild that identifies the most appropriatetargets of treatment, given the individualchild’s specific developmental history.Implications for AssessmentThis body of psychological researchforms the basis for designing an appropriateassessment for children with conductproblems. In Table 17.5, we summarizethe critical areas of research and their relevanceto the assessment process. In Box17.5, we provide a case study of a comprehensiveevaluation of a child with severeconduct problems. As is evident fromTable 17.5, assessment of conduct problemsshares several important characteristicswith the assessment of ADHD. Thecomplex and pervasive nature of conductproblems requires a comprehensive evaluationthat assesses many aspects of thechild’s functioning and psychosocial environment.Further, conduct problems andother relevant aspects of a child’s psychosocialfunctioning should be assessed usingmultiple informants and multiple assessmenttechniques.Table 17.5 Key Research Findings and Their Implications for the Assessment of Childrenwith Conduct ProblemsResearch Findings1. Core symptoms and subtypes: Conduct disordersrepresent a heterogeneous categorywith widely varying levels of impairmentand many important subtypes2. Common comorbidities: Conduct disordersare often accompanied by several comorbidtypes of problems that influence thecourse and treatment of conduct disorders3. Correlates with potential causal roles:Conduct disorders develop through acomplex interaction of numerous factorswithin the child and his or her psychosocialenvironmentImplications for Assessment1a. Assess a wide range of conduct problems1b. Assess the level of impairment associated withthe disorder2a. Assess for the presence of ADHD2b. Assess for the presence of anxiety anddepression (including suicidal ideation)2c. Assess for substance use and abuse3a. Assess important aspects of a child’s oradolescent’s family environment3b. Assess child’s intellectual level, academicachievement, learning style, and socialproblem solving(Continues)

CHAPTER 17 Assessment of Attention Deficit Hyperactivity405Table 17.5 (Continued)Research Findings4. Multiple developmental pathways: Conductdisorders develop through multiple differentpathways each involving distinctdevelopmental mechanismsImplications for Assessment3c. Assess child’s peer interactions, social status,and associations with a deviant peer group3d. Assess critical aspects of child’s social ecology(e.g., economic disadvantage, witnessing ofviolence)4a. Assess the developmental sequence of onset ofconduct problem behavior, especially whethersevere conduct problems onset prior to adolescence4b. Assess for callous-unemotional traitsBox 17.5Case Study: 14-Year-Old Adolescent Male with Severe Conduct ProblemsPatrick was 14 years, 1 month old when hismother requested a comprehensive psychologicalevaluation from a university-basedoutpatient psychological clinic. His motherwas concerned about Patrick’s poor gradesin school and his frequent lying. His motheralso expressed concerns about his frequentfights both at school and in his neighborhood.Because of his fighting at school, Patrick hadbeen placed in a full-time class for childrenwith behavioral problems.Assessment of Core Featuresand SubtypesOn the DISC-IV, both Patrick and his motherreported the presence of a number of severeconduct problems. They both reportedrepeat instances of lying, repeat instances ofstealing items from stores (shoplifting), andseveral school suspensions for physical fights.His mother also reported several instancesof truancy, and Patrick admitted to breakinginto a neighbor’s house to steal thingsand using a knife in a fight. Patrick furtheradmitted to occasional use of marijuana. Theseverity of these conduct problems is supportedby parental report on the CBCL, onwhich Patrick had T-scores of 79 on both theDelinquent Behavior and Aggressive Behaviorscales. On the Personality Inventory forYouth (PIY; Lachar & Gruber, 1994), Patrickobtained a T-score of 69 on the Delinquencyscale.According to both Patrick and his mother,much of Patrick’s aggressive and antisocialbehavior occurred alone. In fact, Patrickreported problems in peer relations, as evidentby T-scores of 68 and 75 on the PIY SocialWithdrawal and Social Skills scales. Patrick’smother reported that his aggressive and antisocialbehavior is of longstanding duration.He had averaged about two suspensions peryear since the first grade, mostly because offighting, indicating that his aggression startedwell before adolescence.Assessment of ComorbiditiesBoth Patrick and his mother reported that Patrickgoes through frequent periods of depression,often lasting for as long as a month. Atthe time of the assessment, Patrick had beenexperiencing significant periods of sadnessfor the past 3 weeks. He had also lost interestin activities, he had not been sleeping well atnight, and he had lost his appetite resulting insignificant weight loss. His mother had also(Continues)

406CHAPTER 17 Assessment of Attention Deficit HyperactivityBox 17.5 (Continued)noticed a decrease in energy and a decrease inhis ability to concentrate. Patrick reported onthe DISC-IV that he had, twice in the past 2years, thought of killing himself by cutting hiswrist and on one occasion had actually startedto use a knife but was stopped by a classmate.Both Patrick and his mother reported thatthe episodes of depression seemed to coincidewith disciplinary confrontations, such asbeing suspended from school.Patrick and his mother reported someproblems of attention and concentration,but these seemed to occur during periods ofdepression, and therefore, did not seem to beassociated with ADHD. Also, a psychoeducationalassessment revealed that Patrick wasfunctioning in the low average range of intellectualabilities, with a particular weakness inhis verbal abilities. He scored in a range commensuratewith this intellectual level on anachievement screener.Assessment of Correlates with PotentialCausal RolesPatrick lived alone with his mother, whoworked full-time outside of the home asa secretary. Patrick’s parents had divorcedwhen he was 6 years old and he had had minimalcontact with his father, who accordingto maternal report, was in and out of prisonand had a substance abuse problem. Patrick’smother reportedly had limited social contacts.Her extended family mostly lived in adifferent region of the country and she hadnot developed a good network of friends.Patrick’s mother also reported that she hadgreat difficulty disciplining Patrick. Becausehe was so moody, she rarely tried to make himdo things. Also, whenever he did somethingwrong (e.g., getting suspended from school),she did not punish him because it would simplymake him angrier and more difficult tolive with. There was no indication that Patrickwas involved with a deviant peer group.In fact, his mother was concerned with hislack of connectedness with any peers.Assessment of Multiple DevelopmentalPathwaysInformation from both the DISC-IV and anunstructured interview both suggested thatPatrick’s serious conduct problems have beenevident, at least since the first grade. Thus,the problems clearly had a childhood-onset.Also, a measure of callous-unemotional traits(Frick & Hare, 2001) did not suggest thatPatrick showed high rates of these traits.This would also be consistent with the highrate of depression, a weakness in his verbalintelligence, and the apparent associationbetween his conduct problems and ineffectiveparenting practices in the home, all of whichare typically more common in youth withoutcallous-unemotional traits.Assessment of Core FeaturesThe first goal of the assessment is to carefullyand thoroughly assess the number,types, and severity of the conduct problemsand the level of impairment that the conductproblems are causing for the child oradolescent (e.g., school suspensions, policecontacts, peer rejection). This assessmentis important given that research has showngreat variability in these dimensions amongchildren with conduct problems, and ithas shown that these dimensions may besome of the best predictors of outcome forchildren with conduct problems (Frick &Loney, 1999). The primary methods ofassessing the core symptoms are structuredinterviews, behavior rating scales, andbehavioral observations (see McMahon& Frick, 2005 for a summary of specificmeasures).

CHAPTER 17 Assessment of Attention Deficit Hyperactivity407Many structured interviews and behavioralrating scales provide good coverage ofthe conduct problem behaviors and allowfor multi-informant assessments. However,they each offer unique advantages inother respects. Behavior rating scales aremore time-efficient and provide some ofthe best norm-referenced information fordetermining the severity of conduct problems.In contrast, diagnostic interviewsoften provide important information onthe degree of impairment associated withthe conduct problems and a structuredmeans of assessing the age of onset of theproblem behaviors. As discussed in Chap.8, behavioral observations provide a thirdway of assessing conduct problem behaviors.Behavioral observations in a child’snatural setting can make a unique contributionto the assessment process by providingan assessment of a child’s behaviorthat is not filtered through the perceptionsof an informant and by providing anassessment of the immediate environmentalcontext of a child’s behavior. Unfortunately,for older children and adolescents,many of the common conduct problemsare by nature covert (e.g., lying and stealing)or only occur infrequently (e.g., fighting),which makes them difficult to capturethrough some observational technique.One important advantage that manystructured interviews have over behavioralrating scales and behavioral observationsis that they provide a structuredmethod for assessing when a child firstbegan showing serious conduct problems,thereby providing an important source ofinformation on the developmental trajectoryof the child’s problem behavior. Forexample, in the DISC-IV (Shaffer, Fisher,Lucas, Dulcan, & Schwab-Stone, 2000),any question related to the presence of aconduct problem that is answered affirmativelyis followed by questions askingthe parent or child to estimate at whatage the first occurrence of the behaviortook place. Obviously, such questions canalso be integrated into an unstructuredinterview format as well.Assessment of ComorbiditiesCo-occurring problems can be assessedconjointly with the assessment of the conductproblems themselves, through theuse of omnibus rating scales, structuredinterviews, or multi-domain observationalsystems. Given the association betweenconduct problems and learning disabilities,a psychoeducational evaluation thatincludes a standardized intelligence test andacademic achievement screener should alsobe a part of most evaluations of childrenand adolescents with conduct problems.Assessment of Correlates withPotential Causal RolesUncovering which of the many potentialcausal factors may be operating on a childreferred for an evaluation could be crucialfor making recommendations for treatment.Of primary importance are correlateswithin the child’s family environment(Frick, 1994). In Chap. 12, we discussedsome basic issues and methods in assessingcrucial elements in a child’s family environment(see also McMahon & Frick, 2007).Research also indicates that peer rejectionis predictive of the development ofconduct problems and is associated withan adolescent’s association with a deviantpeer group. Therefore, assessing a child oradolescent’s peer relationships is a criticalassessment goal. Several of the omnibusrating scales discussed in previous chaptersprovide an assessment of a child’s peerfunctioning. Also, several rating scalesfocus specifically on a child’s or adolescent’ssocial functioning (see also Cavell, Meehan,& Fiala, 2003). Social status can alsobe assessed directly through a sociometricexercise, if a child is in elementary school

408CHAPTER 17 Assessment of Attention Deficit Hyperactivity(see Chap. 9). Given that a child’s associationwith a deviant peer group is associatedwith an increase in the severity of conductproblems, some method of assessing thisaspect of a child’s social functioning (e.g.,Elliott, Huizinga, & Ageton, 1985) shouldalso be included in a comprehensive assessment.Another important class of potentiallyimportant correlates to severe conductproblems involves specific cognitive deficitsand learning styles. Specifically, deficits inintelligence, especially verbal intelligence,have been associated with conduct problems,necessitating a standard intellectualevaluation as part of most assessment batteriesfor children with conduct problems.In addition, as discussed previously, manychildren with conduct problems, especiallythose who also show a callous and unemotionalinterpersonal style, show heightenedsensitivity to rewards compared topunishments. There are computerizedtasks developed to assess this learningstyle in young children; however, the clinicalutility of such information and somemajor limitations in the development ofthese tasks make their usefulness in manyclinical assessments somewhat limitedat the present time (see Frick & Loney,2000 for a review). There are also laboratorymeasures, typically involving a childbeing provided a hypothetical vignette ofa social situation and asked to state how heor she would respond if the situation wasreal, that assess several deficits in socialcognition that have been associated withconduct problems, such as a hostileattributional bias. Examples include theIntention-Cue Detection Task, the Problem-SolvingMeasure for Conflict, and theWALLY Game (Conduct Problems PreventionResearch Group, 1999; Dodge &Coie, 1987; Webster-Stratton & Lindsay,1999). The recently developed Social-CognitiveAssessment Profile (SCAP; Hughes,Meehan, & Cavell, 2004) shows promise asa brief (15–20 min), clinically useful interviewwith elementary school-age children,designed to assess social-cognitive deficitsassociated with conduct problems. Whilethese measures are also not without somelimitations in their clinical usefulness (Frick& Loney, 2000), interventions for the deficitsthat are assessed by these measuresare part of many treatment programs forconduct problems (Lochman et al., 2008).Therefore, the information they providecan be useful in treatment planning.Finally, a child’s social ecology is oftencrucial for understanding the developmentof conduct problems in many cases.Therefore, it is important to assess suchvariables as the economic situation of thefamily, the level of social and communitysupport provided to the child and his orher family, and other aspects of a child’ssocial climate (e.g., neighborhood, qualityof school and, degree of exposure toviolence) (McMahon & Frick, 2007). Forexample, the Neighborhood Questionnaire(Greenberg, Lengua, Coie, Pinderhughes,& the Conduct Problems PreventionResearch Group, 1999) is a brief parentreportmeasure used to assess the parent’sperception of the family’s neighborhoodin terms of safety, violence, drug traffic,satisfaction, and stability.Assessment of ImportantDevelopmental PathwaysA key area of research for guiding theassessment process is the research documentingvarious potential developmentalpathways to conduct problems. As reviewedpreviously, children with conduct problemscan fall into childhood-onset or adolescentonsetpathways, depending on when indevelopment their level of severe antisocialand aggressive behavior started. Also,within the childhood-onset group, thereseem to be important differences betweenthose who do and do not show high levelsof callous-unemotional traits. Knowledge

CHAPTER 17 Assessment of Attention Deficit Hyperactivity409of the characteristics of children in thesedifferent pathways, and the different causalmechanisms involved, can serve as a guidefor structuring and conducting the assessment(McMahon & Frick, 2005). Further,interventions can be tailored to the uniqueneeds of youth in these different pathways(Frick, 2006).Specifically, knowledge of the developmentalpathways can provide a set ofworking hypotheses concerning the natureof the conduct problems, the most likelycomorbid conditions, and the most likelyrisk factors (McMahon & Frick, 2005). Forexample, for a youth whose conduct problemsappear with the onset of adolescence,one would hypothesize based on the availableliterature that he or she is less likelyto be aggressive, to have intellectual deficits,to have temperamental vulnerabilities,and to have comorbid ADHD. However,the youth’s association with a deviant peergroup and factors that may contribute tothis deviant peer group affiliation (e.g., lackof parental monitoring and supervision)would be especially important to assess foryouth in this pathway.In contrast, for a youth whose seriousconduct problems began prior to adolescence,one would expect more cognitiveand temperamental vulnerabilities, comorbidADHD, and more serious problemsin family functioning. For those youthsin this childhood-onset group who donot show CU traits, the cognitive deficitswould more likely be verbal deficitsand the temperamental vulnerabilitieswould more likely be problems regulatingemotions, leading to higher levels ofanxiety, depression, and aggression involvinganger. In contrast, for a youth withchildhood-onset conduct problems whoshows high levels of callous-unemotionaltraits, the cognitive deficits are morelikely to involve a lack of sensitivity topunishment and the temperamental vulnerabilitiesare more likely to involve a preferencefor dangerous and novel activitiesand a failure to experience many types ofprosocial emotions (e.g., guilt and empathy).Further, assessing the level and severityof aggressive behavior, especially thepresence of instrumental aggression, wouldbe critical for children and adolescents inthis group (Marsee & Frick, 2007).As most clinicians recognize, people donot often fall neatly into the prototypesthat are suggested by research. Therefore,these descriptions are meant to serve ashypotheses around which to organize anassessment based on the available research.They also highlight several specific importantpieces of information that are neededwhen assessing children and adolescentswith conduct problems. One of the mostcritical pieces of information in guidingassessment, and perhaps ultimately intervention,is determining the age at whichvarious conduct problems began. Thisinformation provides some indication as towhether or not the youth may be on thechildhood-onset pathway. As noted previously,unstructured and structured interviewsare often the most common methodsfor obtaining this information.Unfortunately, there has been littleconsistency in the literature concerning themost appropriate operational definition ofchildhood- vs. adolescent-onset. For example,the DSM-IV-TR makes the distinctionbetween children who begin showing severeconduct problems before age 10 (i.e., childhood-onset)and those who do not showsevere conduct problems before age 10 (i.e.,adolescent-onset) in its definition of ConductDisorder. However, other researchstudies have used age 11 (Robins 1966) orage 14 (Patterson & Yoerger, 1993; Tibbetts& Piquero, 1999) to define the start of adolescentonset. Thus, onset of severe conductproblems before age 10 seems to be clearlyconsidered childhood-onset and onset afterage 13 clearly adolescent-onset. However,how to classify children whose conductproblems onset between the ages of 11 and13 is less clear and probably dependent on

410CHAPTER 17 Assessment of Attention Deficit Hyperactivitythe level of physical, cognitive, and socialmaturity of the child.In addition to the difficulty in determiningthe most appropriate way to dividechildren based on their age of onset, thereis also concern about how accurate theparent or youth is in reporting the timingof specific behaviors. There are threefindings from research that can help ininterpreting such reports. First, the longerthe time frame involved in the retrospectivereport (e.g., a parent of a 17-year-oldreporting on preschool behavior vs. a parentof a 6-year-old reporting on preschoolbehavior), the less accurate the report islikely to be (Green, Loeber, & Lahey,1991). Second, although a parental reportof the exact age of onset may not be veryreliable over time, typical variations inyears are usually small (Green et al., 1991).As a result, these reports should be viewedas rough estimates of the timing of onsetand not as exact dating procedures. Third,there is evidence that combining informants(e.g., such as a parent or youth) orcombining sources of information (e.g.,self-report and record of police contact),and taking the earliest reported age ofonset from any source, provides an estimatethat shows somewhat greater validitythan any single source of information alone(Lahey et al., 1999).If the youth’s history of conduct problemsis consistent with the childhoodonsetpathway, then additional assessmentto examine the extent to which callousunemotionaltraits may also be present isimportant. There have been several reviewsand critiques of the available methodsfor assessing these traits (Sharp & Kline,2008; Vincent, 2006). The two most commonlyused methods are the PsychopathyChecklist: Youth Version (PCL-YV; Forth,Kosson, & Hare, 2003) and the AntisocialProcess Screening Device (APSD; Frick &Hare, 2001). The PCL-YV is a cliniciancompleted checklist for adolescents ages12 to 18 years. It is completed based on a60 to 90 min semi-structured interview andreview of all available collateral information(e.g., psychosocial histories, institutionalrecords). The APSD includes parent andteacher ratings scales (Frick & Hare, 2001)and a self-report questionnaire (Munoz &Frick, 2007). Although there is research tosupport the usefulness of both the PCL-YV and APSD (see Frick & Dickens, 2006;Frick & White, 2008), both measuresinclude only limited items specificallyrelated to callous-unemotional traits (4 and6 items, respectively). A scale that providesa more comprehensive assessment of thesetraits, the Inventory of Callous-Unemotionaltraits, has been developed and hasshown some initial promise in a large(n = 1443) community sample of young adolescentsin Germany (Essau, Sasagawa, &Frick, 2006) and moderate size (n = 248)sample of detained juvenile offenders inthe United States (Kimonis et al., 2008). InTable 17.6, a summary of the items assessingCU traits from these three measuresare provided.ConclusionsIn this chapter we discuss two specificapplications of the assessment proceduresand techniques reviewed in previous chapters.We discuss the assessment of tworelated types of childhood psychopathology:Attention-Deficit Hyperactivity Disorderand conduct problems. To continueour basic premise that clinical assessmentsshould be guided by basic psychologicalresearch, we provide a brief overviewof some of the more important researchfindings with particular relevance to theassessment process. For both domains,assessments should be structured aroundour current knowledge of the core featuresof each domain. In addition, the mostfrequent co-occurring problems associatedwith both types of psychopathology

CHAPTER 17 Assessment of Attention Deficit Hyperactivity411Table 17.6 Items Assessing Callous-Unemotional Traits from Three Commonly UsedMeasuresPsychopathy Checklist-YouthVersion(Forth et al., 2003)Antisocial ProcessScreening Device(Frick & Hare, 2001)Inventory of Callous-UnemotionalTraits (Essau et al., 2006)Lacks guilt and remorse Feels bad or guilty (I) Feels bad or guilty (I)Shallow affectCallous use of othersFails to acceptresponsibilityDoes not show emotionConcerned about thefeelings of others (I)Concerned about school work (I)Keeps promises (I)Keeps the same friends (I)Does not feel remorseful when doingsomething wrongEasily admits to being wrong (I)Tries not to hurt others’ feelings (I)Feelings of others are unimportantDoesn’t care who he/she hurts to getwhat he/she wantsConcerned about feelings of others (I)Apologizes to persons he/she hurts (I)Does not care if he/she gets in troubleSeems very cold and uncaring to othersWorks hard on everything (I)Always tries best (I)Does not care about doing things wellCares about how well he/she does atschool or work (I)Does things to make others feel good (I)Tries not to hurt others’ feelings (I)Does not like to put the time into doingthings wellWhat he/she thinks is right and wrongis different from what others thinkDoes not show emotionsExpresses feelings openly (I)Hides feelings from othersIt is easy to tell how he/she is feeling (I)Very expressive and emotional (I)Does not care about being on timeNote: (I) designates items that are inversely scored.should be routinely assessed, because thesecomorbidities often have important prognosticand treatment implications.For ADHD, a difficult part of theassessment is ruling out other medical orpsychological disorders that could solelyaccount for the ADHD symptoms. Also,in assessing ADHD, one must be knowledgeableof educational laws related tolegally mandated services for children withADHD so that one can work with educatorsin designing a treatment plan for thechild or adolescent with ADHD.For conduct problems, the myriad ofpotential causal factors should be assessedto determine which ones may be operating

412CHAPTER 17 Assessment of Attention Deficit Hyperactivityfor a given child and which ones should,therefore, be a focus of intervention. Also,different causal factors may be involvedin the various subgroups of children withconduct problems. Understanding the differentpathways through which childrenand adolescents develop serious conductproblems can be critical for designingassessments and interpreting the informationprovided by the evaluation.Chapter Summary1. Externalizing behaviors are the mostcommon reason for referral to childmental health clinics.2. Based on research on ADHD suggeststhat:(a) Assessments should include a multiinformantand multi-source assessmentof the core ADHD behaviors:inattention-disorganization andimpulsivity-hyperactivity; assessmentof these core features must be placedwithin a developmental perspective.(b) Assessments should screen for thepresence of the most common cooccurringproblems that may accompanyADHD: conduct problems/aggression, emotional disturbance,low self-esteem, problematic socialrelationships, learning difficulties,and family conflict.(c) Assessments should rule out alternativecauses for the core symptoms:medical/ neurological disorders,mental handicaps, learning disorders,and adjustment reactions toenvironmental stressors.(d) Because ADHD often has a majorimpact on a child’s or adolescent’sschool functioning, the assessmentshould be conducted in collaborationwith school personnel and witha knowledge of local educationalstatutes relevant to services for studentswith ADHD.3. Based on research on severe conductproblems in children and adolescents:(a) Assessments should provide a multisourceand multi-method assessmentof conduct problems includingdetermining the types and severityof conduct problems and the age atwhich they began.(b) Assessments should screen for themost common co-occurring problemsthat often accompany conductproblems: ADHD, emotionaldisturbance, substance abuse, andlearning disabilities.(c) Assessments should assess knowncorrelates to conduct problems thatcould play a role in causing or maintainingthe problem behavior, andtherefore, should be a major focusof intervention: family functioning,cognitive deficits, social ecology,peer relations, and associations witha deviant peer group.(d) The age at which the serious conductproblems began and the presenceof callous-unemotional traitsshould be assessed because of theirimportance in designating uniquepathways to the development ofconduct problems.

C h a p t e r 1 8Assessment of Depressionand AnxietyChapter Questionsl What are the core symptoms of childhooddepression and anxiety?l Which informant(s) and assessmentstrategies are most useful for the diagnosisof depression and anxiety?l What developmental factors need to beconsidered in assessment of child/adolescentdepression and anxiety?l What factors contribute to better orworse prognosis for childhood depressionand anxiety?Internalizing DisordersIt is commonplace to consider symptomsof depression and anxiety under theterm internalizing problems, a tendencyreflected in the structure of many commonrating scales (e.g., Achenbach & Rescorla,2001; Reynolds & Kamphaus, 2004).More specifically, rating scales often havean internalizing symptom composite thatconsists of depression, anxiety, and often,somatic complaints. Historically, childrenwith internalizing difficulties have beendescribed in a number of ways, includingas having a personality problem syndromewith difficulties of anxiety, withdrawal, andfeelings of inferiority (Peterson, 1961);anxious-fearful (Behar & Stringfield, 1974);inhibited (Miller, 1967); anxious-immature(Conners, 1970); and overcontrolled (Edelbrock,1979). Children solely with internalizingproblems are easily distinguishedfrom children solely with externalizingproblems, but this does not imply that suchdifficulties are easily diagnosed and treated.In fact, it can be quite the contrary.Internalizing disorders are among themost difficult to diagnose because of thenature of the symptomatology. The child’ssymptoms take more of a toll on the child’s413P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_18, © Springer Science+Business Media, LLC 2010

414 chapter 18 Assessment of Depression and anxietysubjective mental state than on significantothers or on the child’s adjustmentto school or other settings. Furthermore,children with internalizing problems havebeen found to be better adapted than childrenwith externalizing difficulties, achievinghigher reading and intelligence testscores, and performing better in school(Cohen et al., 1985). The following quotefrom Reynolds (1990) eloquently summarizesthe central assessment challenge thatis presented by the most common internalizingsyndromes of depression and anxiety.“Internalizing disorders, including depressivedisorders, anxiety disorders, obsessivecompulsive disorder, somatic disorders, andsuicidal behaviors, among others, are associatedwith overcontrolled behaviors. Phenomenologically,internalizing disordersare characterized by covert, inner-directedsymptomatology. The covert nature of thesedisturbances presents challenges to professionals,particularly with regard to assessmentand treatment. As noted in this article,internalizing disorders as a function of theirinsidious nature, do not readily come tothe attention of psychologists. Because ofthis, professionals need to be vigilant to thepotential existence of internalizing disordersin children and adolescents” (p. 137).Within the broad category of internalizingproblems, depression and anxietysyndromes can be reliably differentiated,although it has been theorized that theremay be a third disorder that is symptomaticallyand etiologically overlapping withboth these disorders (see Klein et al.,2005). Still, research has supported a twofactormodel of internalizing problems, oneindicative of depression and one that can beconceptualized as anxiety (Ollendick et al.,2003). Klein et al. (2005) provide a clearreview of the ways in which the symptomsof either problem can be mutually influential,and while it is beyond the scope of thischapter to consider the evidence for a commondisorder for anxiety and depression,it is important for the reader to be awarethat although this chapter will discussthese syndromes as if they differ in manyrespects (including etiology and course),there is substantial comorbidity betweenthe two (Klein et al.).The present discussion is influencedprimarily by the structure of the DSM-IV,which focuses on anxiety and depressiondisorders as separate entities, althoughcomorbidity is recognized in a little understooddiagnosis of Mixed Anxiety-DepressionDisorder (American PsychiatricAssociation, 2000). Ultimately, it will beup to the clinician to determine, throughthe evidence gathered, if interventionstargeting both anxiety and depression arewarranted. The important reminder is thatboth areas of concern be assessed as eitherprimary or secondary influences on thechild’s functioning.Childhood DepressionDiagnostic CriteriaIn order to meet DSM-IV-TR diagnosticcriteria for Major Depressive Disorder, atleast five of nine symptoms must be presentfor at least a two-week period, and oneof the symptoms must be either depressedmood or loss of interest or pleasure (AmericanPsychiatric Association, 2000).These symptoms include:1. depressed or irritable mood2. loss of interest in daily activities3. significant weight loss or failure to makeexpected weight gains4. frequent insomnia or hypersomnia5. motor agitation or retardation6. frequent fatigue7. feelings of worthlessness or guilt8. impaired concentration9. suicidal ideation or suicide attempt.

chapter 18 Assessment of Depression and anxiety415For many years, the need for diagnosticcriteria for childhood depression was notclear, as researchers debated whether ornot the disorder could exist in childhood.There is increasing evidence that childrenand adolescents display much of the samesymptomatology as adults, including cognitivedistortion (Haley et al., 1985). Unfortunately,there is also increasing evidencethat depression is quite persistent andimpairing in childhood. In fact, the existenceof childhood depression is no longera point of contention (Kovacs, 1989).Although there are some developmentalinfluences on the manifestation of depression(see below), most evidence points tothe continuity of depression from childhoodto adulthood (Klein et al., 2005).Despite some of the previous controversyregarding the existence of childhooddepression, clinicians are fairly reliable intheir diagnosis of the condition. Whenusing structured diagnostic interviewssuch as those described in Chap. 11, theinter-rater reliability of the diagnosisranges from about .75 to .90 (Ambrosini,2000; Christ, 1990). Emerging agreementregarding the existence of the depressionsyndrome in childhood, the widespreadapplication of the DSM criteria, and evidenceof inter-rater reliability have allcontributed to forming a consensus thatchildhood depression is an important publichealth problem that warrants significantresearch attention.Regardless of the accuracy of diagnosis,as Reynolds (1990) observes, the needto assess for the presence of depressionis not always obvious. Consequently, thepsychologist must develop some strategiesand/or cues that trigger a search fordepression. Strategies for assessing fordepression can be gleaned from an understandingof the nature of the syndrome,including its risk factors and course. Somemethods for identifying the need to assessfor the presence of depression are discussednext.Characteristics of ChildhoodDepressionAlthough prevalence rates are oftendebated, research has generally concludedthat less than 3% of preadolescent childrenexperience depression but that ratesof depression increase substantially in adolescenceto around 15%, although a largerproportion of youth suffer from symptomsof depression at some point (see Hankinet al., 2008 for review). A prevalence ratethis large warrants vigilance for the existenceof depression in referral populations.Also, a higher rate of depression may beexpected in adolescence because of theattendant stressors associated with thisage range (Cooper, 1990). Adolescents arepresented with increasing developmentaldemands and stressors for which theymay often be ill-prepared (Petersen et al.,1993).Several studies have suggested that theexpression of depression differs somewhatby age. While children and adolescentsboth report dysphoria as a central symptomof depression, younger children morefrequently exhibit sleep disturbances.Less than half of children and adolescentsreport suicidal attempts, increasedappetite, hypersomnia, hallucinations, ordelusions (Christ, 1990). Adolescents aremore likely than young children to experiencefeelings of hopelessness as part ofdepression (Weiss & Garber, 2003). Onestudy found that a community sample ofadolescents reported significantly moresomatic complaints on the Child DepressionInventory (CDI) than preadolescents(Cooper, 1990). Further analyses revealed,however, that this finding was primarilydue to gender differences: adolescent girlsacknowledged more somatic complaintsthan preadolescents (boys and girls) andadolescent boys. However, this findingdoes suggest that excessive somatic complaintsshould trigger further assessmentfor depression.

416 chapter 18 Assessment of Depression and anxietyDepression is also distinctive as a syndromebecause of the crucial hereditaryfactor that has been identified. Heritabilityestimates vary widely, but the proportion ofyouth with depression who have a positivefamily history of depression is somewherearound one third (Rice, Harold, & Thapar,2002). In addition, family history of depressionis a significant predictor of recurrenceof depression for children and adolescents(Klein et al., 2005). Therefore, to say theleast, knowledge of family history is oftenhelpful for conceptualizing cases. Familyresemblance data may affect assessment andintervention planning for a child who hasseveral symptoms of depression but doesnot meet diagnostic criteria. If this child’smother has a history of depression, moreassessment and intervention would be warrantedthan if no family resemblance waspresent. This information indicates that thepsychologist should routinely and carefullygather family history data regarding a historyof depression for the child suspected ofhaving an internalizing disorder.The duration of a depressive disordercan be substantial. In an interesting study,105 children who entered a special schoolfor children who lost a parent due to death,divorce, or separation were divided intodepressed and non-depressed groups basedon a CDI cut score of 13 or greater indicatingdepression. The elevated group wasfound to, for the most part, exceed the cutscore again four years later (Mattison et al.,1990). Mattison et al. found the depressedgroup to show considerable impairmentover the four-year duration of the study.They concluded that:“They further showed significantly pooreracademic performance, received significantlymore counseling, and more oftenseparated from the school under negativecircumstances. The most pathological scoresoverall were demonstrated by the children inthe original depressed group who separatedfrom the school during the 4 years undernegative circumstances” (p. 169).This study did not consider actual diagnosesof depression, but the results point topotentially prolonged difficulties for youthwith higher levels of depressive symptoms.Research has shown that episodes of majordepression tend to average around sevento eight months in length for children andadolescents (Birmaher, Arbelaez, & Brent,2002).The chronicity of depressive symptomatologyhas even been documented infirst-graders. In a study of 677 first-gradechildren from the Baltimore, Marylandpublic schools, CDI scores tended to bemost stable for children with scores inthe upper quartile of the scale. Interestingly,all of the children who scored in theupper quartile in the fall of the academicyear remained in the upper quartile whenretested in the spring. Children who scoredbelow the upper quartile showed considerablymore fluctuation from the fall to thespring (Edelsohn et al., 1992). Furthermore,Luby et al. (2002) have found supportfor a depression syndrome/diagnosisin preschool-aged children, if the durationrequirement from the DSM-IV is reducedAs discussed elsewhere in this text, thedevelopmental onset and course are criticalareas to assess for understanding the manifestationof psychological problems, theassociated features, and the potential outcomes.This issue is illustrated quite wellfor youth depression in a study by Hammen,Brennan, Keenan-Miller, and Herr(2008). They considered the outcomes atage 20 for individuals with an early onsetof depression (i.e., by age 15) versus a lateonset (i.e., after age 15), with an additionalconsideration of whether or not the depressivesymptoms were recurrent by age 20. Ingeneral, the early onset, recurrent group hadthe worst outcomes in many psychologicaland social domains of functioning. In particular,these individuals:l Were more likely than individuals tohave had comorbid anxiety by age 20

chapter 18 Assessment of Depression and anxiety417l Were more likely to have comorbid eatingdisorders and externalizing problemsl Were more likely than individuals withrecurrent depression to report not havinga close friend, not being involved ina romantic relationship, and report havingproblems in family relationshipsl Were less likely to have a history ofclose friendships and good family relationships.These findings suggest that early onsetof depression, and particularly recurrentdepression during adolescence, have a significantimpact on many areas of functioning.Thus, knowledge of the developmentalmanifestations of depression, recognition ofheterogeneity of depressive presentations(Klein et al., 2005), and comprehensivenessin the assessment of depressive symptomsand associated features are essentialfor providing appropriate and meaningfulrecommendations.ComorbidityWhen a child is diagnosed with depression,there is a high probability (i.e., the majorityof cases; Klein et al., 2005) that the childwill also meet criteria for another disorder,especially an anxiety problem (Angold,Costello, & Erkanli, 1999). One studyfound that every child who was diagnosedwith depression over a four-year periodwas at some point also diagnosed with ananxiety disorder (Christ, 1990). This overlapis striking, but it does not indicate thatdepression and anxiety comprise the samesyndrome. In addition to being reliably differentiatedwith diagnostic measures, thetwo problems differ in family resemblanceand course (Stavrakaki & Ellis, 1989).Also of interest is the comorbidity ofdepression and disruptive behavior disorders.Angold et al. (1999) found a greaterthan chance probability that youth diagnosedwith depression also meet criteriafor ADHD, ODD, and CD. The patternof comorbidity also appears somewhat differentbased on developmental level. Morespecifically, younger children with depressionare likely to have comorbid anxietydisorders, whereas adolescents with depressionare more likely to have comorbid substanceabuse problems or eating disorders(see Hankin et al., 2008).The comorbidity of youth depressionalso extends into the realm of academicfunctioning. Hammen and colleaguesfound youth depression was associatedwith greater academic problems and poorerschool attendance (Hammen et al., 1999).As would be expected, depression is associatedwith greater social/peer problems(e.g., Klein, Lewinsohn, & Seeley, 1997;Marcotte, Fortin, Potvin, & Papillon,2002) and a greater number of stressful lifeevents, including interpersonal stressorsthat may have a particularly negative impactfor young girls (Hankin, Mermelstein, &Roesch, 2007). Although a strong case canbe made for depression being either thecause or the effect in these relations, theimportant issue in assessment is to approachreferrals for depression comprehensively soas to determine all of the relevant targetsof intervention. The research on childhooddepression clearly indicates a need to assessfor symptomatology and difficulties in avariety of domains, as well as to assess indicatorsof impairment related to depressivesymptoms (e.g., truancy).Furthermore, some research suggeststhat one should often screen for depressionwhen a medical illness is present. Inone study, depression and anxiety had a41% higher prevalence rate in individualswith medical illness. Depression that is secondaryto medical illness may have differentoutcomes, thus also requiring differingapproaches to intervention (Cassen, 1990).Issues of comorbidity can also be confusedwith the notions of primary and secondaryconditions. The primacy of the conditioncan be especially important for internalizing

418 chapter 18 Assessment of Depression and anxietydisorders because it may be important todifferentially intervene based on knowledgeof primacy. A child with both depressionand anxiety may require intervention fordepression first, for example, if the child ispresenting risk factors for suicidal behavior.Relatedly, a child suffering from anxiety anddepression may require intervention for anxietysymptoms initially if compulsive handwashingbehavior has resulted in impairedfunctioning in school, at home, and amongpeers. Determining primacy of conditionis fraught with problems (Winokur, 1990),and the implications of doing so are not yetentirely clear at this stage of research on thecomorbidity of internalizing disorders inchildren. Some evidence suggests that anxietyor externalizing problems often precededepression rather than the other way around(see Hankin et al., 2007). The clinicianshould attempt to distinguish the primacyof anxiety and depression for the purpose ofdeveloping recommendations for targets ofintervention. Some characteristics that maybe useful in such work include symptomatology,course of symptoms and condition,family resemblance and background, testresults, and response to prior treatment/interventions (Winokur, 1990).Specialized Measures ofDepressionAside from a general approach (i.e., likelythrough clinical interview) for gatheringevidence on symptomatology, family history,stressors, comorbidity, and developmentalcourse in the assessment ofdepression, the clinician also has the abilityto select specific tools that assess depressivesymptoms in an in-depth, efficientmanner. Several examples of such tools arereviewed in this section.Children’s Depression Inventory (CDI;Kovacs, 1992)The Children’s Depression Inventory(CDI; Kovacs, 1991) is a 27-item depressionself-report scale for ages 7 through17. A derivative of the Beck DepressionInventory (Semrud-Clikeman, 1990), theCDI enjoys a long history of clinical use,particularly in research investigations.Scale ContentThe 27 items of the CDI assess a widerange of depressive symptomatology, muchof which is included in popular diagnosticsystems such as the DSM. Items assess sadness,cognitive symptoms, social problems,somatic complaints, and acting-out behaviors.In addition to a total score, there arefive subscales available: Negative Mood,Interpersonal Problems, Ineffectiveness,Anhedonia, and Negative Self-esteem.Administration and ScoringContributing to the popularity of the CDIis its ease of administration and scoring.The scale can be completed in minutesby a competent reader. Each item is composedof three stems from which the childmust choose one. The three stems typicallyrepresent differing levels of severityof depressive symptomatology. Items areassigned scores of 2, 1, or 0, where a higherraw score reflects more severe symptomatology.Linear T-scores are available for the CDItotal and subscale scores, with the authorusing a T-score of 65 or higher as indicativeof clinical significance (Kovacs, 1992). Separatenorms are available for 7–12-year-oldboys, 7–12-year-old girls, 13–17-year-oldboys, and 13–17-year-old girlsNormingThe CDI possesses essentially local normsbased on a sample of 1,266 Florida schoolchildren in grades 2 through 8 (Kovacs,1992). Seventy-seven percent of childrenin this normative sample were white,whereas the remaining 23% were AfricanAmerican, Hispanic, or Native American(specific percentages for each group werenot reported). The normative samplegives little evidence of national representation;hence, the degree to which CDInorms can be recommended as a national

chapter 18 Assessment of Depression and anxiety419or some other standard is not known.However, it is apparent that at least someminority groups were underrepresented.ReliabilityThe reliability of the overall CDI score isgood with internal consistency coefficientstypically in the .80s. The coefficients forthe scales are modest at best, ranging from.59 (Interpersonal Problems) to .68 (NegativeSelf-esteem). Test-retest coefficientsfor the total score are adequate based ondata reported from a number of studiesemploying a range of time frames (i.e., 1week to 1 year; see Kovacs, 1992).ValidityInternal evidence of validity has beenassessed via numerous factor-analyticinvestigations. Kovacs (1991) suggestsusing a five-factor solution for conceptualizingsubscales. The presence of more thanone factor, however, does not necessarilysupport the validity of the CDI. The factorstudies do suggest dominance by a largefirst factor (Cooper, 1990).Criterion-related studies are generallysupportive of the CDI as a measure ofinternalizing symptomatology. The CDIcorrelates significantly with other measuresof anxiety and depression (see Myers& Winters, 2002 for review).Some of the validity evidence associatedwith the CDI has been described as“mixed.” Although, for example, the CDIhas shown some validity for differentiatingbetween nonclinical children and samplesof children with depression, several studieshave shown that the CDI cannot accuratelydistinguish between samples withdepression and other psychiatric groups(Myers & Winters; Silverman & Rabian,1999).Strengths and WeaknessesThe CDI continues to be a widely usedmeasure. Among the CDI’s strengths are:1. A long research history that has contributedto considerable trust by thepsychological re-search community2. Ease of administration and scoring3. Relatively low cost4. Evidence of concurrent validity withmeasures of internalizing symptomsThe CDI, however, has some noteworthyweaknesses that caution against overinterpretation.1. The wisdom of offering such rigid cutscores for screening or diagnostic purposesin the manual is questionable, asthese suggestions are often applied rigidlyby the user (Kamphaus, 2001).2. The scale scores of the CDI have fairlylow internal consistency for clinicalpurposes.3. The norm-referenced scores from theCDI should be interpreted cautiouslygiven their lack of representativeness.In fact, a crucial flaw such as this suggeststhat the CDI may be more usefulfor research purposes than for clinicalassessment and diagnostic decisions.The CDI, in fact, does have animpressive history of research utilityfor which norm-referenced scores areoften of little interest.4. The ability of the CDI to assist withdifferential diagnosis is questionable(Silverman & Rubian, 1999).Reynolds Adolescent Depression Scale,2nd Edition (RADS-2; Reynolds, 2002)The Reynolds Adolescent DepressionScale, 2nd edition (RADS-2; Reynolds) isdesigned to assess symptomatology associatedwith depression via self-report in adolescentsages 11–20. The RADS-2 is closelybased on its predecessor (Reynolds, 1986),although it may be used with a slightlywider age range than was recommendedfor the original version. The RADS-2 isnot designed provide a diagnosis of a specificdepressive disorder. However, it isdesigned for use as a screening measure orfor research.

420 chapter 18 Assessment of Depression and anxietyScale ContentThe RADS-2 provides a thorough samplingof depressive symptoms that areincluded in the DSM and other nosologiesand takes approximately 5–10 min toadminister.Administration and ScoringThe RADS-2 uses 30 items to which theclient responds with one of four choices offrequency: almost never, hardly ever, sometimes,and most of the time. Six of theseitems are considered critical items. A TotalDepression score, as well as scores on fourscales: Dysphoric Mood, Anhedonia/NegativeAffect, Negative Self-evaluation, andSomatic Complaints. The items are placedon a single page where a template is usedto easily sum up item scores to obtain rawscores. Raw scores are then converted topercentile ranks based on the total sampleand by gender and grade, and unlike theprevious version of the RADS, T-scoresare offered. The T-scores are based on theoriginal distributions of the total score andscale scores in the norm sample (i.e., linearT scores). This change represents a significantimprovement and brings the RADS-2in line with the scores available for manyother measures of child functioning. Reynolds(2002), in line with other approachesto interpreting rating scales, suggests thata T score of 65 or higher be consideredclinically significant and worthy of closeconsideration (along with other assessmentdata) in case conceptualization.NormingThe RADS-2 was normed on 9,052 participantsfrom seven states and one Canadianprovince recruited from schools. Thesample is described as socioeconomicallydiverse, with representation reported byage, gender, and ethnicity. The match ofthe sample to meaningful national criteriasuch as census data is not given. Theauthor reports increased representativenessof Hispanics in the RADS-2 normsample; however, Hispanics (i.e., 4.5% ofthe norm sample) and African Americans(7.4% of the norm sample) both appearto be underrepresented. A clinical sampleconsisting of 297 participants was alsorecruited. This sample was mostly female(55.6%), with the vast majority of the clinicalsample being Caucasian (86.1%).ReliabilityReliability indices for the RADS-2 totalscore are good, with the internal consistencycoefficients of the total score andscales ranging from .86 to .93 in the normingsample (Reynolds, 2002). Two-weektest-retest reliability was also good, rangingfrom .77 for Somatic Complaints to .85for the Total Depression score.ValidityLike the RADS, the RADS-2 has good contentvalidity in that it systematically assessesmany of the symptoms commonly associatedwith the syndrome. To date, there are limiteddata on the construct validity of the RADS-2, although the manual reports higher correlationsbetween the RADS-2 and HamiltonDepression Rating Scale Clinical Interviewthan were enjoyed by the RADS. In addition,the RADS-2 total and scale scores showedmoderate to high correlations (i.e., rangingfrom the mid .40s to high .70s) with theMajor Depression and Dysthymic Disorderscales of the Adolescent PsychopatologyScale (Reynolds, 1998).Many studies show reasonablecorrelations between the original RADSand measures of depression and similarinternalizing constructs. Research (e.g.,Krefetz, Steer, Gulab, & Beck, 2002) hasshown good convergence between theRADS and Beck Depression Inventory(BDI), with consistently significant correlationsranging from .68 to .76.Additional research has shown that theprevious version of the RADS correlatewell with a measure of suicide risk (Kaczmarek,Hagan, & Kettler, 2006) and to bepredictive of later suicidality in hospitalizedadolescents (Huth-Bocks et al., 2007).

chapter 18 Assessment of Depression and anxiety421Strengths and WeaknessesThe RADS-2 has many apparent strengthsincluding:1. A thorough assessment of depressivesymptomatology rooted in contentvalidity2. Norm-referenced scores, unlike its predecessor3. Ease of administration and scoring4. Good recommendations regardinginterpretation in the manualWeaknesses may include:1. More information of the norming sampleis provided than was the case for theRADS, but it still does not appear to berepresentative of the U.S. population, particularlyin regards to ethnicity.2. A quite limited research based on theRADS-2, particularly in comparison tothe CDI.3. The response scale (i.e., “almost never,”“hardly ever,” “sometimes,” “most ofthe time”) may be difficult for someadolescent respondents to interpret.Reynolds Child Depression Scale(RCDS; Reynolds, 1989)The Reynolds Child Depression Scale(RCDS; Reynolds) is a downward extensionof the RADS designed for ages 8 through12. It is designed with the same rationale andintended uses as the RADS.Scale ContentThe scale includes 29 items that use a Likerttypefour-choice format. The four responseoptions measure frequency of symptomatologywith the following options: almost never,sometimes, a lot of the time, and all the time.A 30th item uses a rebus response formatwith five faces depicting a range of happyto sad faces. As with the RADS, some itemsare reverse-scored in order to discourageresponse sets. The items were selected withdiagnostic criteria clearly in mind, ensuringthat the RCDS assesses a good range ofdepressive symptoms.Administration and ScoringAn attempt was made to make all items readableby second-graders to enhance its utilitywith this younger age range. The RCDS isalso self-contained on one sheet for convenience,and a scoring template is provided.A machine-scannable form is also availablefor group or other large-scale administrationof the RCDS. Although a cut score isprovided, the author cautions against its usefor diagnostic purposes (Reynolds, 1989).Of course, cut scores are always questionable,even for screening purposes. Again,only percentile rank scores, with their attendantproblems, are offered for the RCDS inlieu of standard scores.NormingThe RCDS was normed with a sample of1,620 children in the Midwest and otherportions of the country. Reynolds (1989)presents data showing similar raw scores forsamples tested in Sacramento, California andBeloit, Wisconsin and concludes that geographicrepresentation is adequate for thesample. Unfortunately, SES was not specificallyindicated for the norming sample, withthe exception that some students had teachersestimate their SES. Gender, ethnicity,and grade data are reported for the sample.Like the RADS-2, however, the representationof the sample in comparison to criteriasuch as census data is not given.ReliabilityThe internal consistency of the RCDS isstrong, with coefficients at most grade levelsof .90. A coefficient of .87 was obtainedeven at grade 3. Test-retest at 2- and 4-weekintervals is strong, with coefficients rangingfrom .81 to .92 (see Myers & Winters, 2002for review; Reynolds, 1989).ValidityReynolds (1989) makes a good case forcontent validity. In addition, five studies ofthe relationship of the RCDS to the CDI

422 chapter 18 Assessment of Depression and anxietyproduced significant correlations, mostly inthe .70s. The RCDS has also shown goodcorrelations with a variety of other indicatorsof childhood depression (see Myers& Winters, 2002). Correlations with theRCMAS (see below) are also significant,ranging from .60 to .67 in three studies.A five-factor solution is reported for theRCDS, with the first factor clearly dominant.The presence of a strong first factorargues (lays?) for an emphasis on total scoreinterpretation.Strengths and WeaknessesThe RCDS is strikingly similar to theRADS in terms of its strengths and weaknesses(see above). However, the RCDS,unlike the RADS-2, uses only percentilesas norm-referenced scores. We have someconcern regarding the use of norm-referencedscores with the RCDS, given thelack of information regarding the normingsample. The RCDS may be a good way toefficiently gain self reports on depressivesymptoms, but interpretation of the RCDSshould be descriptive rather than reliant onordinal scale data (i.e., percentiles).Peer Nomination Inventory forDepression (PNID)The PNID includes 20 items that assessthree aspects of depression: depression,happiness, and popularity (Lefkowitz &Tesiny, 1980). It is designed for classroombasedassessment. The PNID is quite differentfrom the other measures reviewedin this section in that it is based on peernominations. A child’s score on each itemis the number of nominations divided bythe number of students who rated the childin the classroom. The item scores are thensummed to yield a total score. Normativedata for the PNID have been collected forgrades 3 through 5. Reliability data for thePNID are adequate, but correlations withself-report measures of depression havebeen low (Merrell, 1994) to moderate(Semrud-Clikeman, 1990). Although nota strong measure of clinical depression perse, the PNID could serve as an additionaldata source in the evaluation of children bygauging a child’s popularity, social interaction,and happiness as perceived by peers.Information of this nature could assist intargeting for intervention classroom behaviorsthat improve a child’s adaptation in thisimportant setting.An Assessment Strategyfor DepressionAs we have in previous chapters, we recommenda five-stage assessment processfor the assessment of depression: screening,classification, co-morbidities, alternativecauses, and treatment considerations(see Table 18.1).In light of the prevalence of depression, itscommon comorbidity with other problems,its adverse effects on development in avariety of domains, and its less-than-flagrantsymptomatology, we recommendthat every child who is seen by a clinicianshould be screened for depression.Research findings support the need forscreening all referrals that are seen notonly by psychologists, but also by otherprofessionals (e.g., pediatricians, teachers).The screening process can be timeefficientand may include:l Administering a brief, self-report inventoryto the child, who has adequatereading ability.l Asking parents whether or not their childexhibits symptoms given in the DSM-IVcriteria.l Querying teachers, nurses, grandparents,or others about symptoms of depression.Any one of these screening methods takesonly a few minutes. In fact, querying both

chapter 18 Assessment of Depression and anxiety423Table 18.1 The Assessment Questions Related to Childhood Depression and Implications forAssessmentAssessmentQuestionScreeningClassificationComorbiditiesAlternativeCausesTreatmentConsiderationsImplications for AssessmentAdminister screening measures or screen during interview. Assess forcritical symptoms (i.e., suicidality, psychosis). Determine need for furtherassessment.Assess for presence of symptoms that meet DSM-IV criteria. Determinestability and duration of symptoms.Determine onset.Determine family resemblance.Assess for presence of anxiety disorders, ADHD, LD/MR, eating disorders,or Oppositional Defiant Disorder/Conduct Disorder. Determine the influenceof depression on school performance, absenteeism, etc.Assess social relationships and peer social status.Assess for presence of substance abuse, particularly in adolescents.Obtain developmental and medical histories.Rule out dysthymia and Post-traumatic Stress Disorder.Rule out medical problems that are associated with depression.Assess for presence of maladaptive cognitions. Assess for presence ofchronic stressors. Assess parental depression and parenting style.Evaluate response to previous interventions (e.g., psychotropics).the child and a parent would take only a fewminutes. Screening efforts may allow the clinicianto implement intervention in order toavert considerable suffering (see Box 18.1).Classification of a child as depressedrequires meeting the DSM-IV-TR criteria,and by this we mean meeting all criteriabased on apparently valid information.Possessing the necessary symptomatologyis only a first step in meeting criteria. Thepsychologist has to be sure that malingering,response sets, or other threats to validityhave not had an effect on the report ofsymptoms. An adolescent, for example,may be asked if he/she has experienceddecreased appetite, fatigue, sleeplessness,agitation, and suicidal ideation. He/shemay discern that it is wise to deny all ofthese difficulties if he/she thinks that hemay be a candidate for inpatient or partialhospitalization treatment. Clinicians whotreat adolescents have seen cases in whichthe youth denies suicidal ideation, althoughhe/she may have been transferred from ahospital emergency room because of a suicideattempt. An array of valid assessmentmethods is necessary in order to ensureadequate documentation of symptoms:l Structured and semi-structured interviews(e.g., K-SADS; see Chap. 11) withchildren, their parents, teachers, andother caregivers are necessary to assessfor the presence of symptomatologyand their effects on the child’s functioningin different environments. Indeed,semi-structured interviews, because oftheir flexibility and comprehensive havebeen described as “best practice” byKlein et al. (2005; p. 427).l Self-report inventories and parent andteacher rating scales can provide furtherdocumentation of symptoms, screenfor comorbid problems, and assess for

424 chapter 18 Assessment of Depression and anxietyBox 18.1A Case of Individual Screening for DepressionScreening for depression is crucial because ofthe nature of internalizing symptomatology.Such screening is time-efficient in most settings.This case example illustrates the importanceand practicality of screening.Matt was referred for an evaluation becauseof his parents’ concerns about his academicprogress. He is a 16-year-old junior in highschool. His parents and teacher suspect thathe has a reading disability.Matt’s mother reported a long history of academicproblems for Matt, including considerablespeech delays as a youngster. He was reportedlyenrolled in speech therapy throughout his elementaryschool years. He was often ridiculed byhis peers for misarticulations and stuttering. Hisspeech problems have reportedly declined somewhatin high school. He has, however, had longstandingproblems with reading and spelling.Matt’s parents expressed some frustrationbecause they feel that Matt has been passedalong from grade to grade because he is veryquiet and never causes any problems.He continues to have trouble in most of hisclasses. He is performing adequately in math,but he complains that he has trouble takingnotes in class and comprehending them afterhe takes them. He reportedly failed in hisEnglish class during the first semester of thecurrent academic year. He is very distraughtabout these failures because of his desire to goto college following graduation and his concernsabout disappointing his parents. Mattalso reported some other recent stressors in hislife. He recently broke off a 1-year relationshipwith a girlfriend, and he is receiving rehabilitationfor a severe back injury that he incurredin a bicycle accident six months ago. Matt alsohas two sisters who were reportedly diagnosedwith developmental disorders, one with a readingdisability and the other with ADHD.Matt gave the impression of being welladjustedand socially skilled. He was impeccablygroomed and a good conversationalist.He even asked the examiner questions abouthis interests and offered his full cooperationwith the assessment process. Matt describedmany successes in his life, including beingchosen as captain of the baseball team at hisschool. He apparently has an active and successfulsocial life. Aside from athletics, heis reportedly involved in a variety of otherextracurricular activities.The results of the assessment producedclear evidence of below-average achievementin reading, even though Matt’s intellectualfunctioning was slightly above average.Following reports of moderate levels ofdepression on the parent and self-report versionsof the BASC-2, Matt was asked to completethe Child Depression Inventory (CDI).Matt acknowledged enough symptoms on theCDI to fall into the mild/moderate range ofdepressive symptoms. He acknowledged guiltabout past failures, uncontrollable sadness,insomnia, hopelessness, decreased appetite,increased fatigue, low self-esteem, concernsabout his appearance, lack of interest in socialactivities, and occasional thoughts of hurtinghimself, among other symptoms. The incongruencebetween the symptoms endorsed onrating scales and his presentation was striking.During the feedback session, Matt and hisparents were apprised of these findings andasked to verify their validity, which they did.He was then referred for follow-up assessmentand intervention for his depression.This case illustrates how time-efficientscreening for depression can be of potentialimportance. In this scenario, depression wasnot suspected by the referral sources, theexamining psychologist, or the client.the presence of response sets or otherthreats to validity.l History- taking from the child and/orparents is necessary in order to establishadequate duration and pervasiveness ofsymptoms as well as to obtain valuableinformation in areas such as family psychiatrichistory (Klein et al., 2005).

chapter 18 Assessment of Depression and anxiety425Anxiety disorders are especially importantto consider when ruling out comorbiditiesof depression. Anxiety symptomatologyshould be assessed concurrently with all ofthe data collection necessary for classification.Omnibus rating scales that include awell-validated anxiety scale and interviewschedules that also assess for anxiety arehelpful in this regard. The full realm ofanxiety problems should be routinely ruledout. Somatization problems, phobias,milder fears, obsessive-compulsive disorder,and separation anxiety disorder shouldbe distinguished.One of the alternative causes thatlooms large in cases of depression is amedical difficulty that can cause depression.We recommend that a clinicianassessing a child who displays the symptomsof depression should have access tothe results of a recent physical exam, or anexam should be scheduled if one has notbeen conducted within a few months ofthe psychological evaluation.The psychological evaluation must alsorule out other conditions that may appearto be a major depressive episode. In thisregard, thorough history-taking is necessaryin order to rule out transient statessuch as bereavement (e.g., depression dueto a child’s mother being called away toactive duty in the armed forces), adjustmentdisorder with depressed mood, ordysthymia.Finally, assessment methods mustbe selected that help direct interventionefforts. Rating scales may be used tobroadly sample behavior in home, school,and other settings. A broad sampling ofbehavior allows the clinician to considerdepressive problems and other difficultiesthat may in some way adversely affectthe child’s adaptation. If, for example, thechild also shows considerable evidence ofworry about school difficulties, an adjustmentin the child’s curriculum may have apositive effect on the child’s perceived levelof stress, which, in turn, could help abatedepressive symptoms.An assessment of academic achievementmay also be necessary in order to determinewhether or not a child is suffering fromacademic under-achievement associatedwith long-term depressive symptomatology.Teacher interviews and norm-referencedand curriculum-based measures maybe helpful in this regard.Self- and other informant ratings arealso critical for initial assessment and formonitoring response to interventions. Theprevious research citing long-term adaptationdifficulties for children with depressiveproblems suggests that careful monitoringof the effects of treatment is necessary, andpost-treatment follow-up is warranted inorder to prevent or effectively deal withindications of relapse.A case study of an adolescent girl withdepression is provided in Box 18.2, illustratingthis approach to assessing depression.Anxiety Disordersof ChildhoodThe term anxiety can be traced to the Latinroot angere, which means to cause distressor strangle. That is, unpleasant feelingsof anxiety, including potentially intensefeelings of fear, dread, and worry, evokeemotions associated with strangulation(Stavrakaki & Ellis, 1989). In there, discussionof evidence-based assessment of anxietyin children, Silverman and Ollendick(2005) highlight Barlow’s relatively recentdefinition of anxiety as being particularlyuseful for assessment and subsequentintervention. For Barlow (2002), anxietycan be described as being dominated byconcerns about future unpredictability orloss of control which is accompanied by a“shift in attention to the focus on potentiallydangerous events or one’s own affectiveresponse to these events” (p. 104).Problems that are commonly discussedunder the general rubric of anxiety include

Box 18.2Sample Case of an Adolescent with DepressionSeveral assessment methods are applied in thecase of Corin. The report is given as it wouldbe written in a typical psychological report,with commentary regarding interpretation.Corin, a 16-year-old girl, was referred for apsychological evaluation by her mother due toacademic problems. According to Mrs. Jacobs(mother), Corin has trouble concentrating inschool and she thought that Corin’s schoolproblems may be caused by ADHD.Corin lives with her mother and youngerbrother. Corin’s parents are divorced, and shereportedly rarely sees her father.Mrs. Jacobs stated that there were no difficultiesduring her pregnancy with Corin. Duringbirth, however, Corin was in fetal distress,and Mrs. Jacobs had an emergency Cesareandelivery. Corin was born full-term andweighed 8 pounds, 7 ounces. Corin’s motherdescribed her development as normal in thatshe reached all motor and language developmentalmilestones at age-appropriate times.Corin is, reportedly, in very good health. Sheis supposed to wear contact lenses, but hermother reports that Corin rarely wears them.Corin’s parents divorced when Corin was5 years old. In addition, Corin was reportedlyvery close to her maternal aunt who diedearlier this year. Corin stated that her motheris constantly on her back and that she has todo too much work around the house. Thereis a family history of alcohol abuse on bothher mother’s and father’s sides of the family.Corin’s mother and father also both have areported history of alcohol abuse.Corin is currently in the 10th grade. Corin’smother noted that Corin did well in elementaryschool, always receiving A’s or B’s. Startingin the seventh grade, Corin’s academic performancereportedly deteriorated. Her mothernoticed that she had problems reading aloudand problems with mathematics.Corin received her first failing grade in theninth grade and was moved from algebra backto pre-algebra, which she had in eighth grade.According to her mother, her grades continuedto deteriorate, and currently she is receiving an“F” in history and English, a “C” in science, algebra,and Spanish, and an “A” in photography.Corin’s teachers feel that she is smart but thatshe does not put forth significant effort. Theyreported that she exhibits behavioral problemsin the classroom. She is reportedly constantlytalking and disrupting the class. Her algebrateacher stated that she is fond of Corin but thatshe plays around with her classmates too much.She reportedly talks back to her teachers, whichis consistent with the reports from last year’steachers. Corin’s teachers also reported that shefrequently sleeps through class.According to records kept by the schoolfor the past two years, Corin is often excessivelytardy and absent from class. In the fallof this year, she served two days of suspensionin study hall for being rude to a new teacher.Subsequent to this incident, a parent conferenceat school resulted in Mrs. Jacobs agreeingto have Corin evaluated in order to ruleout ADHD.Corin has reportedly been seen in counselingat a community clinic for two years. Her currentcounselor reported that Corin remains under agreat deal of stress due to her parents’ divorce,her aunt’s death, and her continuing academicvproblems. Mrs. Jacobs mentioned that Corinalso often has stomachaches and dizzy spells,which she attributes to stress. Corin complains ofnot being able to go to sleep; once asleep, she frequentlyhas nightmares about dying. She reportedthat she worries about getting cancer. She toldthe examiner that one time, she felt a bump onher leg and thought that she had cancer.Mrs. Jacobs reported that Corin drinksalcohol regularly, and that at times, she drinksquite heavily (i.e., 12–15 beers). Corin told theexaminer that she sometimes blacks out on theweekends from drinking. She also admitted tofrequent use of marijuana, and she reportedlysmokes at least one pack of cigarettes a day. Afew weeks before the evaluation, Corin wasdetained by police because of skipping schooland being intoxicated.In two interview sessions, Corin describedherself as being depressed. She told the examinerthat, for the past three weeks, she hasfelt angry, sad, and hopeless. Throughout theinterview Corin appeared to be agitated.Corin’s intellectual functioning was estimatedas being above her same-aged peers.She did well on tests that were timed andinvolved physical manipulation of materialsand the results indicate that her nonverbalreasoning abilities are better developed thanher abilities in verbal reasoning.

chapter 18 Assessment of Depression and anxiety427Generalized Anxiety Disorder, SeparationAnxiety Disorder, Agoraphobia, SocialPhobia, Simple Phobia, School Phobia,other fears, Obsessive-Compulsive Disorder,Post-traumatic Stress Disorder, andPanic Disorder, among others.Although they share obvious features,these anxiety problems also compriseunique facets that have clear implicationsfor assessment, differential diagnosis, andtreatment planning. According to theDSM-IV-TR (2000):(a) Generalized Anxiety Disorder is characterizedby excessive, unfocused worry;(b) Separation Anxiety Disorder is centeredaround significant distress whenseparated from primary caretakers orhome;(c) Agoraphobia features anxiety regardingbeing in places or situations fromwhich escape is difficult or perceivedas difficult;(d) Social Phobia is characterized by fear ofsituations that involve social evaluation;(e) Simple Phobias are marked by significant,intense fear of particular objectsor stimuli, whether the exposure isdirect or anticipated;(f) Obsessive Compulsive Disorder consistsof persistent, recurrent thoughtsand behavioral compulsions associatedwith impaired functioning;(g) Post-traumatic Stress Disorder is basedon re-experiencing of a traumatic eventand increased arousal; and(h) Panic Disorder features a sudden occurrenceof intense physiological and cognitivesymptoms of anxiety that tend tooccur unexpectedly.The measures of anxiety that often usedin child assessments typically screen aspectsof all of these problem areas broadly. It isthen up to the clinician to obtain the necessaryinformation to best conceptualize thespecific anxiety problem or disorder andthen make recommendations accordingly.Characteristics of ChildhoodAnxiety DisordersComparatively speaking, much less isknown about childhood anxiety disorderscompared to the knowledge base fordepression. Classification schemes and theassessment instruments themselves are primarilybased on research on adult anxietydisorders (Silverman & Ollendick, 2005).There are issues from research on childrenand adolescents specifically that haveimportance for assessment practice:l In many respects, anxiety disorders canbe conceptualized as exaggerations ofnormal responses to developmentaldemands (e.g., separation anxiety inyoung children; anxiety in social situationsfor adolescents). It is when theanxiety response is impaired that interventionis needed (Silverman & Ollendick,2005).l Evidence regarding the stability andprognosis of anxiety disorders in childrenis quite mixed (Weems & Stickle,2005).l Research using the Anxiety DisordersInterview Schedule for Children (ADIS)has found strong parent-child interraterreliability for individual anxiety disorderdiagnostic categories (Lyneham, Abbott,& Rapee, 2007) but also substantialparent-child disagreement (Grills &Ollendick, 2003). These results shouldcaution the clinician to attend to reasonsfor convergence and divergence amonginformants.l The phenomenology of childhood anxietydisorders appears to be very similarto that of adults (Silverman, 1993).However, the assessment methods usedto make diagnoses of childhood anxiety

428 chapter 18 Assessment of Depression and anxietydisorders have often not been developedwith potential developmental influenceson the manifestation of anxiety in mind(Silverman & Ollendick, 2005).l While anxiety often is apparent in casesof depression, anxiety disorders alsooften occur in the absence of depression(Silverman, 1993). Anxiety with comorbidproblems is associated with pooreroutcomes than anxiety alone (see Saavedra& Silverman, 2002).l Separation Anxiety Disorder typicallyoccurs in children less than 11–13 yearsof age (Strauss, 1993; Weems, Hammond-Laurence,Silverman, & Ginsburg,1998); however, social phobiabecomes more common in adolescence(Weems et al., 1998).l Fears of animals, the darkness, andheights are more likely to occur atyounger ages than social phobias which,in turn, occur at younger ages than agoraphobia(Strauss, 1993).l There are two main types of school phobia:those that come on acutely in a childwho has functioned reasonably normallybefore onset, and those that arise in achild who has had similar problems fromthe preschool years and has never developedthe social skills that would permitnormal functioning (Berg, 1993).l The average of onset for childhoodObsessive-Compulsive Disorder (OCD)is approximately 9–12 years, with familyhistory of OCD placing a youngster atriskfor an even younger onset of OCDsymptoms (see Baldwin and Dadds,2008 for review).l The central symptoms of PosttraumaticStress Disorder (PTSD),increased arousal, psychic numbing,and re-experiencing the trauma, occurin children as well as adults who arediagnosed with the condition (Last,1993). However, the frequency andintensity of these symptoms may beimportant indicators of impairmentdue to PTSD symptoms for children(Carrion, Weems, Ray, & Reiss, 2002).Specialized Measures of AnxietyRating ScalesRevised Children’s ManifestAnxiety Scale (RCMAS; Reynolds &Richmond, 1985)The Revised Children’s Manifest AnxietyScale (RCMAS; Reynolds & Richmond)measures the expression of anxiety symptomatologywhether or not the constructis conceptualized as being a state or a trait.We review the RCMAS in relative depthbecause it is widely used and has manyunique features for a single constructrating scaleScale ContentThe RCMAS includes 37 items distributedamong four scales: PhysiologicalAnxiety (10 items), Worry/Oversensitivity (11 items), Social concerns/Concentration(7 items), and Lie(L, 9 items). The content of the subscalesappears diverse. Items from the physiologicalanxiety subscale, for example, rangefrom “difficulty making decisions” (notethat the physiological nature of this itemis not apparent) to “awakening scared fromsleep” to “having sweaty hands.”The L scale is a rather unique featureof a single-construct measure. TheRCMAS L scale measures children’s tendencyto portray themselves in a favorablelight with items like, “I always have goodmanners.” To obtain a high score on thisscale, children would have to deny evergetting angry and liking everyone theyknow. This scale is likely to be transparentto many adolescents. Research has shownthat younger children tend to score higheron the RCMAS Lie scale than older youth(Pina, Silverman, Saavedra, & Weems,2001).

chapter 18 Assessment of Depression and anxiety429Administration and ScoringChildren have simply to respond “yes” or“no” to RCMAS items that are read froma one-page response form that is accompaniedby a template for scoring. Groupadministration is feasible for older children.A high score is indicative of higheranxiety.The Total Anxiety score is a T-scoreconversion of a raw score. Standard scoreswith a mean of 10 and standard deviationof 3 are also provided for the subscales. Astandard score of 13 on the L scale shouldlead the clinician to suspect the child’sreport (Reynolds & Richmond, 1985).Norm-referenced scores are provided bygender, age, and ethnicity, giving the usera variety of interpretive options. Nongender-basednorms are not offered.NormingThe RCMAS was normed using 4,972children age 6 through 19 years. Thesample was collected from 13 states. Age,gender, and ethnic in-formation is givenin the manual (Reynolds & Richmond,1985). The representation of Hispanicstudents is lacking, and no formal measureof SES is included. Consequently,while the sample size is large and thegeographic representation is better thanfor some other single-trait measures, therepresentativeness of the RCMAS sampleis still open to question.ReliabilityInternal consistency reliability of the compositescore is good with overall coefficientsin the low .80s. Subscale reliabilitiesare somewhat lower, particularly for thePhysiological Anxiety and Social Concerns/Concentration scales. The PhysiologicalAnxiety scale produced a coefficient of .67for the total score in the norm sample. TheSocial Concerns/Concentration estimatewas similar at .64. The Worry/Oversensitivitycoefficient was somewhat higherat .76. Clearly, the Worry/Oversensitivitysubscale is the most trustworthy and may,in some instances, be more useful for identifyingchildren with anxiety problems perse than the RCMAS total score (see Silverman& Ollendick, 2005).ValidityConsiderable validity evidence is providedin the RCMAS manual. Factor studies ofthe standardization data do not suggest thatthe RCMAS is dominated by a single factor.A review of the factor loadings suggest thatsome of the scale placements were based onminimal loadings. The item about havingdifficulty making decisions, for example,had a .26 loading on both the first and secondfactors for the norming sample, yet theitem is placed on the Physiological Anxietyfactor, not Worry/Oversensitivity. Theoreticalor logical considerations may havehelped with the item-placement process.The RCMAS demonstrates consistentrelations with other measures of anxiety. Acorrelation of .65 is reported between theRCMAS Total Anxiety score and the StateTrait Anxiety Inventory for Children TraitAnxiety score.Like the CDI, the ability of the RCMASto identify true cases of specific anxietydisorders is questionable (Silverman &Rabian, 1999). Similarly, in their review ofthe relevant research, Silverman and Ollendick(2005) point out that the RCMAS doesnot consistently distinguish children withanxiety disorders from children with otherforms of psychopathology (e.g., depression,Conduct Disorder, ADHD), although itmay distinguish children with anxiety fromchildren without any significant psychologicaldifficulties. Clinicians are, therefore,advised to use diagnostic interviews orother measures to identify disorders per se,with the RCMAS serving as one potentialindicator of problems with anxiety.Strengths and WeaknessesStrengths of the RCMAS include:1. A broad sampling of anxiety symptomatology

430 chapter 18 Assessment of Depression and anxiety2. Ease of administration and scoring3. A larger and more geographicallydiverse norming sample than is typicalfor single-construct measures4. A reliable Total Anxiety score5. The presence of an L scaleWeaknesses of the RCMAS include:1. A lack of information about relevant ethnicand SES representation of the normingsample2. Subscales with questionable externalevidence of validityAdditional Self-ReportMeasuresIn addition to the RCMAS, there is a vastarray of rating scales for child anxiety. Silverman& Ollendick (2005) provide a usefulreview of many such measures.Three self-report measures that haveenjoyed wide use and extensive researchare the Fear Survey Schedule for Children-Revised (FSSCR; Ollendick, 1978), theMultidimensional Anxiety Scale for Children(MASC; March et al., 1997), and theState-Trait Anxiety Inventory for Children(STAIC; Spielberger, 1973). The STAIChas a companion parent report version.Some of the assessment domains offered bythese measures are provided in Table 18.2.Two relatively new self-report ratings ofanxiety focus on aspects of the individual’scognitions as they contribute to, or be a signof, marked anxiety. The Anxiety ControlQuestionnaire for Children (Weems, Silverman,Rapee, and Pina, 2003) contains 30items that assess the child’s perceived controlover threatening events or stimuli. TheChildren’s Automatic Thoughts (Schniering& Rapee, 2002) assesses whether ornot the respondent has experienced particularnegative thoughts about anxietyprovokingsituations. Initial research onTable 18.2 Domains Assessed by SeveralRating Scales Assessing AnxietyFSSC-R DomainsTotal fear scoreFailure/criticismFear of unknownInjury/small animalsFear of danger/deathMedical FearsMASC DomainsPhysical symptomsSocial anxietyHarm avoidanceSeparation/panicSTAIC DomainsTotal State AnxietyState-cognitiveState-somaticTotal trait anxietyTrait-cognitiveTrait-somaticthe psychometrics of both of these scaleshas had promising results (see Silverman &Ollendick, 2005), and the unique focus ofthese scales also appears to translate wellto cognitive-behavioral interventions forchildhood anxiety.Interviews and Clinician RatingsThe typical way for anxiety to be assessedin children is through interviews (includingsome of the diagnostic interviewsreviewed in Chap. 11) or subscales withinbroad-band ratings (usually completed byparents, teachers, or youths themselves).The Anxiety Disorders Interview Schedulefor Children (Albano &Silverman, 1996)is a unique-structured interview in that itprovides an in-depth assessment of symptomsand impairment across anxiety disordercategories.

chapter 18 Assessment of Depression and anxiety431The Yale-Brown Obsessive CompulsiveScale for Children (Goodman et al., 1989)is but one example of a clinician rating ofanxiety symptoms. The adult version of thisscale is well-known and extensively usedin research and practice. Such a tool maybe useful if a clinician wants to zero in onspecific areas of anxiety symptomatologyin a quick, efficient manner (the YBOCShas ten items) rather than conducting anextensive multi-domain interview.Cognitive-BehavioralAssessmentKendall and Ronan (1990), among others,emphasize the importance of a child’sinformation processing as a mediatingvariable in the development and expressionof anxiety symptoms. Some of the cognitivevariables that may mediate anxiety areself-statements, irrational beliefs, currentconcerns, images, problem-solving capacities,expectancies, and attributions. Methodsfor assessing these cognitive variablesinclude (Kendall & Ronan, 1990):l Self-monitoringl Recordings of spontaneous verbalizationsl Think aloud methodsl Role playingl Imagery assessmentl Thought sampling and thought listingtechniquesl Interviewsl Endorsement measuresIn addition, numerous other observationalmethods, including parent-child interactionsand social interaction scenarios havebeen used in the assessment of youth anxiety(see Silverman & Ollendick, 2005).These informal methods of assessmentvary in the degree to which they havebeen developed for clinical use. Cognitive-behavioralmethods potentially servethe clinician, even in experimental form,by providing corroborating data for testbasedfindings, and by providing insightsthat may be useful for intervention design.Still, the incremental validity obtained byincorporating such procedures into assessmentbatteries for childhood anxiety isuncertain (Silverman & Ollendick, 2005).The preceding discussion in no wayprovides an exhaustive description of theavailable methods for assessing anxiety inchildren. Instead, we have attempted tohighlight those measures that are wellknown,widely used, and/or unique in theirimplementation or focus. The reader isencouraged to remain up-to-date on theburgeoning body of research on evidencebasedassessment of anxiety (as well as otherforms of psychopathology) to determinewhich approaches possess clinical utility.An Assessment Strategyfor AnxietyWe again recommend a five-stage assessmentprocess for the assessment of anxiety:screening, classification, comorbidities,alternative causes, and treatment considerations(see Table 18.3). Screening foranxiety can be accomplished using bothinformal and formal means. A few questionsabout anxiety symptoms may be moreappropriate for many settings than the useof self-report or other more formal psychometricdevices. As noted elsewhere in thistext, many omnibus rating scales provide ascreening of anxiety, but they do not differentiateamong different types or classesof anxiety problems. The most importantscreening principle to remember is to doit, because failure to screen for anxiety mayresult in under-serving children with lessobvious problems.Classification of anxiety difficulties iscomplicated by debate regarding the

432 chapter 18 Assessment of Depression and anxietyTable 18.3 The Assessment Questions Related to Childhood Anxiety and Implications forAssessmentAssessmentQuestionScreeningClassificationand SubtypesComorbiditiesAlternativeCausesTreatmentConsiderationsImplications for AssessmentAdminister time efficient assessment instruments (i.e., rating scales)Determine need for further assessment.Assess for presence of symptoms that meet may DSM-IV criteria throughspecific rating scales or interviews. Determine stability and duration ofsymptoms.Determine onset.Assess for presence of depression, academic difficulties, externalizingproblems, or other anxiety problems.Determine the influence of anxiety on school performance, absenteeism,etc.Assess social relationships and peer social status.Obtain developmental and medical histories.Rule out stressful life events or other psychopathology as primaryexplanation for anxiety symptoms.Rule out medical problems that are associated with anxiety.Assess for presence of maladaptive cognitions.Assess parental depression and parenting style.Evaluate response to previous interventions (e.g., pharmacotherapy,counseling, etc.).appropriateness of existing criteria, includingthe DSM-IV. These issues notwithstanding, itis critical for treatment planning purposes forthe clinician to determine the phenomenologyof the client’s anxiety and the situations/stimuli that tend to elicit or alleviate the anxiety.Silverman and Ollendick (2005) note thatsome available measures are better able thanothers for assessing different types of anxiety.Anxiety disorders are most likely to becomorbid with other internalizing disorderssuch as other anxiety disorders or depression.Moreover, the nature of the symptomatologymay vary over time. A child may, forexample, initially be diagnosed with SADand not have school difficulties. Schoolrefusal and subsequent academic problems,however, may appear at a later time.Medical problems should be ruled outas alternative causes because of the role thatphysical illness and medical procedures canplay in initiating or exacerbating a child’sanxiety symptoms. In fact, one of the firsthypotheses to consider when somaticsymptoms are present (e.g., headaches,stomachaches, diarrhea) is the possibilitythat the child is physically ill. Althoughit seems obvious, it should be stated that amedical evaluation should be conductedwhen somatic complaints are evident.Further assessment should be conductedon the factors that may influence treatmentoutcome. Research on this area of assessmentis limited, but an approach that incorporatesbehavioral observation, rating scales,and/or self-monitoring is recommended(Silverman & Ollendick, 2005). Maladaptivecognitions, family stressors, and traumaticevents are but some of the variables that mayplay a role in the client’s manifestation andmaintenance of anxiety symptoms and thatwill undoubtedly play a role in the progressof treatment.A sample case illustrating this assessmentapproach for anxiety is provided inBox 18.3.

chapter 18 Assessment of Depression and anxiety433Box 18.3A Sample Case Illustrating the Assessment of AnxietyThomas is a 9-year-old male who was referredfor an assessment due to academic difficultiessubsequent to being transferred to a newschool. He was referred in May of the schoolyear. His grades are reportedly lower in his newschool than in the previous school. He makesfrequent self-deprecating statements andreportedly requires much attention and supervisionto complete his homework. He tendsto give up easily on academic activities, oftenrefusing to complete them. He reportedly hastrouble separating from his mother, and heoften seeks attention in the classroom. On thepositive side, he is described as endearing toteachers, creative, and artistically talented.Thomas was reportedly born three monthspreterm and weighed approximately threepounds at birth. He was treated in neonatalintensive care for two months after delivery.His mother reported that he did not walkuntil 16 months, and he did not speak in singlewords until 24 months of age.Thomas was reportedly diagnosed at age5 with a motor delay, ADHD, and ODD. Hecontinues to receive occupational therapy forhis motor problems. His teachers have reportedlyalways indicated that he responds better ina structured classroom. His grades have usuallybeen above grade level with the exception ofmathematics. He has not been retained nor hashe been served in special education. Accordingto his mother, Thomas often complains ofheadaches and stomachaches at school.Thomas’s test behavior was also remarkablein that he refused testing initially. Heclung to his mother in the waiting room.He cried and requested that his mother joinhim for the testing. During testing (with hismother in the room), he cried, tantrumed,threw objects, refused to answer questions,and responded impulsively. Several times, heremarked, “I’m an idiot.” He was rescheduledto complete testing a week later.During the second assessment session,Thomas cooperated fully. He again pleadedwith his mother to join him, but she refused.He did not display any of the previous behaviorproblems, and he did not make self-deprecatingstatements.Thomas’s intellectual functioning was inthe Low Average range as was his measuredacademic achievement in reading, math, andwriting. His academic difficulties do not appearto be the result of a specific learning disability,although it is reasonable to expect that Thomasmay struggle with some academic tasks, unlesshe is given some assistance in acquiring furtheracademic skills.Based on findings from rating scales andhistorical information, Thomas was diagnosedwith Separation Anxiety Disorder. This diagnosiswas based on the rationale that he displayedthe minimum of three symptoms for a minimumof four weeks. His symptoms included:1. Excessive distress when separated fromfamily members2. Worry about the well-being of familymembers3. Somatic complaints when separatedAn interesting aspect of this case is that theprevious diagnoses of ADHD and ODD werenot confirmed based on detailed interviewinformation and rating scales. For example,many of Thomas’s oppositional behaviorsseemed better explained by his refusal to separatefrom his mother and his difficulty workingon tasks to their completion when he was upset.His teacher reported that Thomas only seemsinattentive and distracted when faced with difficulttasks. On new or difficult tasks, Thomasseems to adjust better when given some initialguidance and feedback on his performance.Treatment recommendations were madethat centered around ways for Thomas tocope with his anxiety, particularly regardinggoing to school and ways for his mother tofacilitate his progress and not maintain hisdesire to avoid separation through tantrumingor refusal to perform tasks. The results ofthe evaluation were also shared with schoolpersonnel, and a behavioral intervention planfor Thomas was developed at school.

434 chapter 18 Assessment of Depression and anxietyConclusionsThe assessment of child anxiety and depressionrequires a well-trained clinician. Theless obvious nature of their symptomatologyand the range of disorders related toanxiety and depression make identificationof these problems a professional challenge.The nature of the symptomatology alsopoints to the necessity of gathering selfreportinformation – an additional challengein the assessment of young children.Many child anxiety and depression measurescombine the two sets of symptoms.As such, scales are not likely to differentiatebetween the two syndromes, necessitatingefforts at scale refinement (Silverman& Rabian, 1999).Although our understanding of childdepression and anxiety has increasedexponentially of late, our understanding ofthese problems still lags behind the availableknowledge base for adults. Cliniciansare advised to make a concerted effort toremain abreast of changes in the emergingresearch base. We recommend thediscussion of evidence-based assessmentof childhood depression by Klein et al.(2005) and the article on evidence-basedassessment of childhood by Silverman andOllendick (2005) as useful resources forhighlighting important issues in this areaof assessment.Chapter Summary1. Depression and anxiety difficulties canbe subsumed under the more global setof adjustment difficulties of childhoodcommonly referred to as internalizingdifficulties/disorders.2. Internalizing disorders are among themost difficult to diagnose because of thenature of the symptomatology.3. In order to meet the diagnostic criteriafor Major Depressive Disorder, at leastfive of nine symptoms must be presentfor at least a two-week period, and one ofthe symptoms must be either depressedmood or loss of interest or pleasure(American Psychiatric Association,2000). The nine symptoms include: (1)depressed or irritable mood, (2) loss ofinterest in daily activities, (3) significantweight loss or failure to make expectedweight gains, (4) frequent insomnia orhypersomnia, (5) motor agitation orretardation, (6) frequent fatigue, (7)feelings of worthlessness or guilt, (8)impaired concentration, (9) suicidality.However, commonly used assessmenttools vary in the degree to which theirdepression scales match these criteria(See Box 18.3.)4. Approximately 2–3% of preadolescentchildren may be suffering from depressionat any point in time, with ratesincreasing significantly in adolescence.5. Specialized measures that may be usefulfor assessing childhood depression includethe Children’s Depression Inventory, theReynolds Adolescent Depression Scale-2, the Reynolds Child Depression Scale,and the Peer Nomination Inventory forDepression, among many others.6. A five-step method for assessing fordepression and anxiety involves (1)screening, (2) classification, (3) comorbidities,(4) alternative causes, and (5)past treatment.7. Each of the anxiety disorders describedin the DSM-IV has a different set ofcore symptoms or problems. Thus, differentialdiagnosis is often quite importantfor optimal treatment planning.8. Some findings regarding anxiety disordersinclude: many anxiety problemscan be conceptualized as exaggerationsof normal developmental processes(e.g., separation anxiety for young children);research is mixed on the degree

chapter 18 Assessment of Depression and anxiety435of parent-child agreement for anxietysymptoms on structured interviews;the phenomenology of childhood anxietydisorders is very similar to that ofadults; anxiety disorders often occur inthe absence of depression.9. Some relatively popular self-reportmeasures of anxiety symptomatologyin children include the Revised Children’sManifest Anxiety Scale, the FearSurvey Schedule for Children-Revised,the Multidimensional Anxiety Scale forChildren, and the State-Trait AnxietyInventory for Children.10. Cognitive-behavioral assessment,behavioral observations, and clinicianratings have also been espoused asimportant assessment tools for anxietythat have implications for interventionplanning and that enable progressmonitoring. Structured diagnosticinterviews are often useful for differentialdiagnosis and treatment planningas well.

C h a p t e r 1 9Assessment of Autism SpectrumDisordersChapter Questionsl How are autism-spectrum disordersdefined by the DSM-IV?l How is autism differentiated from mentalretardation?l What types of scales or strategies aremost useful for the assessment of childrenwith these syndromes?l What is the recommended approach forthe assessment of autism spectrum disorders?Definitional IssuesAutism is believed to affect 1 in 500 children(Bertrand et al., 2001), yet it is oftennot recognized until after the child is 3years of age (Filipek et al., 2000). Hence,early screening is important for devisingthe most appropriate intervention strategy.In addition, because of the difficulty in recognizingautism or related problems, newassessment methods for infants and toddlersare needed. However, a discussion ofthe development of such tools is beyondthe scope of this chapter. Importantly, thearray of assessment strategies available forautism spectrum assessment referrals hasgreatly improved and more easily translateto treatment recommendations. This chapterwill give an overview of the measuresand noteworthy issues in the assessment ofchildhood autism.DSM-IV-TR CriteriaIn the DSM-IV-TR, Autistic Disorder ispart of a class of problems known as PervasiveDevelop-mental Disorders (PDD;American Psychiatric Association, 2000).437P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_19, © Springer Science+Business Media, LLC 2010

438 chapter 19 Assessment of autism spectrumHowever, the term “Autism SpectrumDisorders” (ASD) is now generally usedin lieu of PDD (see Ozonoff, Goodlin-Jones, & Solomon, 2005). Other disordersthat are included under the rubric of ASDinclude Childhood Disintegrative Disorder,Asperger’s Disorder (often referred toas “Asperger’s Syndrome”), Rett’s Disorder,and PDD Not Otherwise Specified (AmericanPsychiatric Association, 2000).The prototypical ASD is, of course,Autistic Disorder, a condition that has longbeen of interest to psychologists. AutisticDisorder will be the primary focus of thischapter, with some discussion of other ASDissues, particularly regarding the assessmentof Asperger’s Disorder. The DSM-IV-TR describes Autistic Disorder as “thepresence of markedly abnormal or impaireddevelopment in social interaction and communicationsand a markedly restricted repertoireof activity and interests.”Autistic Disorder is also known by otherterms such as autism or early infantile autism.Autistic Disorder is known to be frequentlycomorbid with mental retardation (AmericanPsychiatric Association, 2000). Withco-morbidity estimates ranging as high as75%, the differential diagnosis of autismand mental retardation can be challenging.The DSM-IV emphasizes three classesof symptoms that are central to the disorder:social interaction problems, communicationproblems, and repetitiveand stereotyped behaviors. The onsetof these symptoms must occur by age 3.The DSM criteria require that a total ofsix symptoms with a minimum of one fromeach of the three classes of problems bepresent in order to make a diagnosis. Furthermore,at least two of the symptomsfrom the social interaction class of problemsmust be present.The social interaction symptoms include:1. Failure to engage in appropriate nonverbalbehaviors such as appropriate eyecontact, facial expressions, body postures,and social gestures2. Poorly developed peer relationships3. Failure to share pleasurable interests,activities, or achievements with others4. Lack of social reciprocityCommunication symptoms include:1. Delayed development of spoken language2. When speech is present, inability tobegin or maintain a conversation3. Stereotyped or repetitive use of language4. Lack of imaginative playProblems with repetitive or stereotypedplay include:1. Preoccupation with one or more behaviorpatterns2. Preoccupation with nonfunctional routinesor rituals3. Inappropriate motor movements4. Preoccupation with specific objectsCharacteristics of AutismAutism Spectrum Disorders are oftendescribed as etiologically diverse disorders.Evidence points to genetic (e.g.,Bailey et al., 1995) and neurological (e.g.,Courchesne et al., 2001) underpinnings ofASDs; however, specific etiological agentshave not been identified. Much remains tobe learned about risk factors that may beetiologically involved in autism. The lackof scientific clarity implies that cliniciansshould not place undue importance on etiologicalagents when conceptualizing casesof autism or when providing feedback toothers. Instead, the focus of assessmentshould be on symptomatology, developmentalhistory (e.g., language development),strengths, and impairments.Differentiation of autism from otherASDs and from Mental Retardation can be

chapter 19 Assessment of autism spectrum439quite difficult. One study found that childrenwith language disorder could be more easilydifferentiated from children with autism, butautism and PDD-NOS were more difficultto differentiate (Mayes, Volkmar, Hooks, &Cicchetti, 1993). Similarly, differentiation ofhigh functioning children with autism (i.e.,those with average or higher intelligence)from children with Asperger’s Syndromecan be quite difficult in that such youth withautism may not appear to have communicationdifficulties (Howlin, 2003).Some of the assessment tools discussedin this chapter may provide useful informationin making distinctions among ASDs aswell as related problems. However, thereremains a relative lack of research on thisissue and no measure to date has clearlyshown the ability to differentiate withinASDs in meaningful ways (Ozonoff et al.,2005). Therefore, it is incumbent on theclinician to use a variety of methods to gaina comprehensive picture of the client’s difficultiesand history such that the most appropriateinterventions can then be sought.concern, particularly for high-functioningindividuals with autism, is in the internalizingproblem domain (e.g., Kim, Szatmari,Bryson, Streiner, & Wilson, 2000). However,the validity of commonly used measuresof depression and anxiety – such asthose discussed in the preceding chapter– for individuals with ASDs is unknown(Ozonoff et al., 2005).Regular, comprehensive assessmentacross domains of functioning will aid intracking treatment gains as well as otherimportant areas of difficulty apart from theprimary problems associated with ASDs.For example, many children with ASDsmay also present with significant externalizingbehavioral problems that occur whenexpectations are not met, routines change,or due to frustration from communicationdifficulties. These behaviors may themselvesbe a primary focus of treatment. Inaddition, sleep difficulties that may accompanyASDs may warrant specific interventions(Ozonoff et al., 2005).ComorbidityThe most frequent comorbid problemfor autism is mental retardation, with themajority of children with autism also meetingcriteria for mental retardation (AmericanPsychiatric Association, 2000). On theother hand, the majority of individuals withmental retardation do not meet criteria forautism or another ASD. Needless to say,the inclusion of cognitive functioning inassessments of ASDs is critical for caseconceptualization and treatment planning.The same, then, can be said for the inclusionof a measure of adaptive functioningin the assessment battery.As noted by Ozonoff et al. (2005), newdifficulties may emerge over the course ofdevelopment for youth with ASDs suchthat initial diagnosis as well as treatmentplanning and progress monitoring arecomplicated. A common comorbid area ofSpecialized Measures ofAutismInfant Behavioral SummarizedEvaluation (IBSE)The IBSE (Adrien, Barthelemy, Perrot, &Roux, 1992) is a unique rating scale in thatit is appropriate for infants and preschoolers.One study evaluated some psychometricproperties of the scale for a sample of39 children with autism, 33 with MentalRetardation, and 17 with other handicaps(Adrien et al., 1992). Good inter-raterreliability was found. A factor analysis of31 items found two factors, with the firstfactor accounting for most of the items.Adrien et al. also concluded that first factorscores were capable of differentiatingthe group of children with autism groupfrom the other two groups. The IBSE was

440 chapter 19 Assessment of autism spectrumalso able to differentiate among preschoolerswith autism and other children withdevelopmental delays by age 26 months(Desombre et al., 2006).Autism DiagnosticInterview-Revised (ADI-R)The ADI-R is a semi-structured diagnosticinterview designed to be used with a child’sprimary caregiver and can be used for anyonewith a mental age of 2 years or higher(Lord, Rutter, & Le Couteur, 1994). TheADI-R assesses behaviors relevant to thedifferential diagnosis of pervasive developmentaldisorders in individuals from 18months to adulthood. The ADI focuses onthree domains of behavior that are largelyconsistent with the DSM-IV criteria: reciprocalsocial interaction; communication,and repetitive, restricted, or stereotypedbehavior. Scores on the ADI-R havebeen found to differentiate children withautism from children with other forms ofdelay (Mildenberger, Sitter, Noterdaeme,& Amorosa, 2001). Another study againfound the ADI-R to reliably differentiatebetween samples of children withautism and children with mental retardation(Lord, Storoschuk, Rutter, & Pickles,1993). This investigation, however, identifieddifficulties when trying to differentiatethese samples among children with mentalages of less than 18 months, reflecting thecontinuing dilemma of diagnostic accuracywith low-functioning children.One study evaluated inter-rater reliabilityof the ADI-R across seven examinersfor one case (Cicchetti, Lord, Koenig,Klin, & Volkmar, 2008) and found perfectagreement for 74% of the items and “poor”agreement on less than 10% of the items.A large scale study of the ADI-R foundsupport for a two-factor model (i.e., socialcommunication and stereotyped behaviors)as well as a three-factor model (i.e., socialcommunication, peer relationships/play,and stereotyped behaviors; Frazier, Youngstrom,Kubu, Sinclair, & Rezai, 2008).The strength of the ADI-R is its strongresearch base, including the validity evidenceof the current version and its predecessorwhich goes back to two decades (LeCouteur et al., 1989). However, the ADI-Ris quite lengthy, taking anywhere from 90min to 3 h to administer and score (Ozonoffet al., 2005). Therefore, it may be lesspractical than other tools in many clinicalsettings.Parent Interview for Autism(PIA)The PIA (see Stone, Coonrod, Pozdol, &Turner, 2003) is a newer alternative to theADI-R. It is unique in its focus on trackingchanges in symptoms over time, withparents rating the frequency of symptomson a five-point scale. Research has showngood internal consistency and differentialvalidity for the PIA. In addition, changesin PIA ratings, particularly in the areasof social and communicative functioning,were related to changes reflected in otherratings of the child symptomatology (Stoneet al., 2003). While generally based on thethree core dimensions of autism, the PIAhas items that were developed and rationally/theoreticallygrouped into the followingscales (Stone & Hogan, 1993):l Social Relating Affective ResponsesMotor Imitationl Peer Interactions Object Playl Imaginative Play Language Understandingl Nonverbal Communicationl Motoric Behaviorsl Sensory Responses Need for SamenessWhile the PIA has shown some promise,research is limited, and the validity of the12 subscales is not documented.

chapter 19 Assessment of autism spectrum441Childhood Autism Rating Scale(CARS)The CARS (Schopler, Reichler, & Renner,1988) is a rating form that is typicallycompleted by an observer of the behaviorof a child who has reached a developmentalage of approximately 2 years or higher.In some cases, an observation period ofonly about 30 min is used (Sevin, Matson,Coe, & Fee, 1991). Items are rated on a7-point continuum for the 15 subscales.The CARS is composed of nine portions.It includes items such as a puzzle-like taskand a set of toy objects designed to assessmake-believe play activity.The CARS produces a moderately highcorrelation of.67 with the ABC (Eaves &Milner, 1993). Although this correlation issignificant, the two instruments did classifychildren at differing rates. The CARS correctlyidentified 98% of the children withautism, whereas the ABC did so for 88% ofthe sample (Eaves & Milner, 1993).In an investigation by Sevin et al.(1991), 24 children were assessed withthe CARS. The inter-rater reliability ofthe CARS for these subjects was foundto be highly variable. Coefficients wereas low as 0.10 (Activity Level) and 0.14(Intellectual Response), and as high as0.85 (Relating to People). The inter-raterreliability coefficient for the total scorewas 0.68.In spite of such low to modest reliabilitycoefficients, the CARS results showedfairly good concordance with cliniciandiagnoses of autism (Sevin et al., 1991).The CARS classified 19 of the 24 subjectswith autism correctly. Only two ofthe subjects were incorrectly classifiedwith autism, although this study provideda weak test of differential validity becauseof the small sample of subjects withoutautism. Subsequent research has shownthat the total score on the CARS correlateshighly with the ADI-R but that italso may overidentify autism among childrenwith mental retardation (Saemundsen,Magnusson, Smari, & Sigurdardottir,2003).The CARS is a well-known, widely usedassessment tool for autism; however, Ozonoffet al. (2005) caution that the content isbased on conceptualizations of autism datingprior to DSM-IV.Autism Diagnostic ObservationSchedule (ADOS)The ADOS (Lord, Rutter, DiLavore, &Risi, 2002) is designed to assess symptomsof ASDs in a wide rang of individuals basedon age and language skills. The ADOSconsists of four modules, with the moduleselected based on the child’s current expressivelanguage and developmental level (i.e.,Module 1 for no speech; Module 2 forphrase speech; Module 3 for verbally fluentchildren and young adolescents; Module 4for verbally fluent adolescents and adults;Lord et al., 2002). Module 2 is describedin Box 19.1.The ADOS emphasizes the assessmentof reciprocal social interaction andcommunication rather than restricted orrepetitive behavior, and it is structured inthat the environment for the observationis structured in a standardized way, andthe examiner presents predefined tasks tothe child in this structured setting. In thisway, the examiner serves as both the stimulusfor interaction in try to elicit particularresponses from the subject and as therecorder of behavior.A study on the development of the currentversion of the ADOS revealed stronginterrater reliability, internal consistency,and (Lord et al., 2000). The differentialvalidity of the ADOS for distinguishingchildren with autism vs. other delays wasalso good. Another study found substantialagreement between the ADOS andADI-R in the classification of children ashaving autism or not in children younger

442 chapter 19 Assessment of autism spectrumBox 19.1A Closer View of the Autism Diagnostic Observation Schedule (ADOS)Together with the ADI-R, the ADOS has beenreferred to as the current “gold standard” forthe diagnosis of ASDs (Ozonoff et al., 2005,p. 524). The ADI-R, of course, allows the clinicianto gain a detailed history of the child’sdevelopment and symptomatology in a semistructuredinterview format. This approachis not particularly unique among assessmentpractices. However, the structured observationformat of the ADOS is unique to childassessment in that there are no well-establishedanalogous procedures for assessment ofinternalizing or externalizing problems.To further illustrate the content of the ADOS,Module 2, designed for children who speak inphrases, is described below. Module 2 consists of14 tasks as follows (see Lord et al., 2002):1. Construction task: Designed to assess howthe child communicates a need for moreobjects to complete the task. For example,some children with ASDs may grab theexaminer’s hand to reach for more objectsrather than making a verbal request2. Response to name: Evaluates how readily thechild responds to his/her name as said bythe examiner or a parent or other stimulidirected to the child (e.g., phrases, touching).Children with ASDs may require moredirect stimuli or prompting to response3. Make-believe play: Assesses whether child’s playincludes imaginative use of objects beyondtheir usual purpose and the extent to whichdolls are used to depict social interaction4. Joint interactive play: Assesses the extent towhich the child initiates interaction with theexaminer during joint play above and beyondresponding to the examiner’s statements orrequests5. Conversation: Evaluates extent to which thechild responds to the examiner’s statementsor questions in a way that leads to furtherback-and-forth conversation6. Response to joint attention: The examinershifts his/her gaze and observes the extentto which the child follows that shift (withor without pointing). Children with ASDsmay have difficulty engaging injoint attention in this manner, particularlywithout other prompts such as pointing7. Demonstration task: Designed to assesshow well the child demonstrates commonactivities (e.g., brushing teeth, gettingdressed) without the use of objects.8. Description of a picture: For this task, theexaminer attempts to elicit spontaneouslanguage from the child and to observewhat kinds of stimuli (from pictures) areof interest to the child9. Telling a story from a book: In addition toassessing the child’s spontaneous languageand interest in a story, this task, inrequiring the child to tell a story from apicture book, evaluates whether the childcan provide continuity in the story bysequencing events in the story in a sensiblemanner10. Free play: This task is designed to assesswhether or not the child seeks interactionor involvement from the examiner duringfree play. In addition, the extent to which thechild maintains attention to one task for anappropriate interval is observed. Repetitivebehaviors – a symptom of ASDs – may also beevident during this task (Lord et al., 2002)11. Birthday party: This task evaluates thechild’s ability to participate in a common-scriptedevent (i.e., birthday partfor a doll). Children with ASDs may notengage in the make-believe of imaginingthe doll as a child or may have troubleparticipating, including with prompting12. Snack: Assesses how readily the childcommunicates a preference from a choiceof snacks. Other social interaction skills(e.g., gaze, facial expression) are alsoobserved during this task.13. Anticipation of a routine with objects: Thistask is designed to assess how well a childanticipates a routine (e.g., the response ofa balloon when it is deflated or let go) andwhether or not the child participates inthe routine(Continues)

chapter 19 Assessment of autism spectrum443Box 19.1 (Continued)14. Bubble play: This task allows the examinerto observe the child’s enjoyment of aplay activity, social interaction skills, andmotor skills in another play contextOn an achievement test, Corin’s performanceon all areas of achievementwere in the average range, indicating thather academic achievement is consistentwith what would be expected for her age.Several measures were given to assessCorin’s behavioral and emotional functioning.Rating scales were completedby Mrs. Jacobs, two of Corin’s teachers,and Corin. Corin’s mother and a teachercompleted the Achenbach Child BehaviorChecklist (CBCL) and Teacher RatingForm (TRF), respectively. This teacherrated Corin as having significant attentionproblems, delinquent behaviors, andsomatic complaints. Mrs. Jacobs did notindicate any significant problems on theCBCL. Despite the reports of attentionproblems, Corin’s teachers stated that theydo not believe that Corin has ADHD.Corin completed several self-reportmeasures that were used to determine herself-perceptions. She was given the Piers-Harris Self-Concept Scale (PHCSCS),and the YSR. On the PHCSCS, Corinrated herself high on anxiety, depression,somatization, and (external) locus of control.On the YSR, Corin’s responses weresimilar to the other self-report measures.Corin also endorsed items that indicateproblems with impulse control and compliance.On the self-report measures, Corinreported some positive qualities for herself.She described herself as being goodlooking and has having good peer relationships.She reported that all of her friendslike her because she is funny. Corin alsoindicated that her peer relationships andsocial life at school are the areas in her lifethat are not currently stressful for her.As a follow-up measure, Corin wasgiven a structured psychiatric interviewon which Corin endorsed all of the itemsin the area of major depression with theexception of suicidal ideation. Corindescribed herself as being anxious, andshe told the examiner that she often hasstomachaches and dizzy spells that sheassociates with stress. Also, accordingthe structured interview, Corin meetsthe DSM criteria for alcohol abuse. Shestated that she frequently drinks largequantities of alcohol, and occasionally,she has blackouts. Corin also admittedto the examiner that she smokes marijuanaquite a bit.From the information gathered a diagnosisof ADHD does not seem appropriate.Corin’s acting-out behaviors are more thanlikely caused by the stress and depressionthat she currently feels. Current diagnosticimpressions are major depression, singleepisode, moderate, and alcohol abuse.Recommendations for Corin include:1. Consultation with a psychiatrist to helpdetermine the most appropriate treatmentstrategies for Corin’s depression.2. It is also recommended that Corin continuereceiving counseling services.3. Corin is referred to an addictionscounselor who works with adolescentswho are abusing alcohol.4. Corin may benefit from group counselingfor her alcohol abuse. The bestgroup would be one made up of otherteenagers.5. The Jacobs family should seek familycounseling to improve the home situation,particularly in regards to communicationand parenting monitoring.Mrs. Jacobs may also benefit from(Continues)

444 chapter 19 Assessment of autism spectrumBox 19.1 (Continued)individual counseling to help alleviatesome of the stress in her life.6. Corin’s teachers reportedly enjoy havingCorin in their classroom and seemeager to do what they can to help her.The school psychologist can devisebehavior modification plans to beimplemented by Corin’s teachers toincrease Corin’s appropriate classroombehaviors. Corin should take part inthe development and implementationof the plans.7. Communication between the schooland home is necessary. Weekly reportsof Corin’s behaviors should be provided.Mrs. Jacobs is encouraged to work withCorin’s teachers to devise solutions toany problems that may occur.8. After six months to a year of intervention,Corin needs to be reevaluated tosee if there are still issues of attentionproblems in the classroom.9. Corin would benefit from tutoring inalgebra.than age 8. However, the agreement waslower for older individuals (de Bildt et al.,2004). Still, these authors suggested thatan assessment battery that includes boththe ADOS and ADI-R is suitable, and perhapsthe best available, for the assessmentof autism.Gilliam Autism Rating Scale –2nd Edition (GARS-2)The GARS-2 (Gilliam, 2008) is a normreferencedrating scale that can be completedin approximately 5–10 min byparents. The GARS-2 includes 42 items,three scales (i.e., Stereotyped Behaviors,Communication, and Social Interaction)and a total score (Autism Index).norm-referenced scores are available for theGARS-2, and initial research indicates goodinternal consistency and differential validity(Gilliam, 2008).The previous version of the GARS (Gilliam,1995) has been the subject of muchmore research. A consistent finding fromthis research is that many youth otherwiseindicated as having autism scored loweron the GARS than on other measures ofautism symptoms (Mazefsky & Oswald,2006; Sikora, Hall, Hartley, Gerrard-Morris, & Cagle, 2008; South et al., 2002).This set of findings may indicate some seriousdrawbacks of the GARS, although it isunclear whether or not this issue applies tothe current version of the GARS. Becausethe GARS-2 is geared toward screeningfor symptoms of autism, the clinician maywant to use a low threshold of reporting ofsymptoms on the GARS-2 to initiate furtherfollow-up.Functional AnalysisIn the context of the assessment of ASDs,functional analysis helps the cliniciandetermine “what the child is trying to communicatethrough the behavior” (Ozonoffet al., 2005, p. 534), which takes on addedimportance in light of the behavioralproblems and communication difficultiesthat present with ASDs. It is difficultto imagine a comprehensive assessmentof ASDs that does not include some formof behavioral observation and functionalanalysis as a means to determine targetsof intervention. Direct observation, coupledwith functional analysis, allows theclinician to also confirm or disconfirmthe notions of parents, teachers, or othercaretakers as to the communicative intent

chapter 19 Assessment of autism spectrum445of the child’s behavior and on appropriateways to reduce inappropriate communicativebehaviors (e.g., using another’s handto obtain a desired object; biting oneselffrom frustration). Functional analysis,then, is conducted squarely with behavioralinterventions in mind. It is not a toolto differentially diagnose ASDs from otherproblems.An Assessment Strategy forAutismASDs, including autism specifically, arerecognized as having heterogeneous etiologiesand highly variable symptomatologyamong those affected (Campbell et al.,1991). Such within-syndrome variabilitysuggests that children who are eithersuspected of or diagnosed with autism orother ASDs should have access to multidisciplinaryassessment procedures. Theseverity of autistic symptoms, as demonstratedby the overlap with mental retardation,further emphasizes the need fora range of professionals to be involved inassessing the affected child. Such interdisciplinarywork, however, can be challengingif there are no adequate structuresfor encouraging interdisciplinary practice.Consequently, the psychologist may haveto make a concerted effort to go beyondthe typical practice regimen and communicatemore frequently with colleagueswho practice related specialties.Marcus, Lansing, and Schopler (1993)described one of the better methodologiesfor fostering multidisciplinary assessmentof autism. They described the Stateof North Carolina TEACCH evaluationsystem as follows:This process includes a developmental/psychoeducationalassessment of the child, psychologicalassessment, detailed parent interviewingincluding assessment of adaptivefunctioning, and a medical screening. Thisintegrated and multiperspective approachgenerates comprehensive data on diagnosis,intellectual and adaptive functioning, motor,language, and social functioning, behaviorproblems, medical factors, family functioningand school or community factors. The data aresufficiently detailed and objective to be organizedinto dimensions or axes that both overlapwith and extend beyond the DSM-III-R system(p. 350).In situations in which interdisciplinaryassessment and intervention teams are notas readily available, it is still incumbenton the clinician to utilize methods thatassess ASDs comprehensively and specifically,rather than rely on subjective clinicalopinion or measures (e.g., IQ tests) thatdo not differentially diagnose ASDs fromother developmental delays. Having saidthat, the incremental validity and clinicalutility of specific ASD measures arelargely unknown (Ozonoff et al., 2005). Aswith other areas of child assessment, moreresearch is needed as to the most sound,parsimonious battery that would addressreferrals for ASDs and lend itself to usefulrecommendations.Nevertheless, the severity and franknessof the symptomatology of ASDsrequire us to emphasize different aspectsof our five-stage paradigm: screening,classification, comorbidities, alternativecauses, and treatment considerations (seeTable 19.1).The screening process for ASDs isaimed primarily at the population ofchildren with developmental delays asopposed to the general population. Weadvise that clinicians consider the possibilityof ASDs in all samples of youngchildren with developmental disabilities.Because the nature of ASDs usually doesnot allow the child to serve as a sourceof screening information, parent andteacher ratings are likely to be the mostfruitful. The time-efficiency of such ratings

446 chapter 19 Assessment of autism spectrumTable 19.1 Assessment Questions Related to Childhood Autism Spectrum Disorders andImplications for AssessmentCharacteristicScreeningClassificationComorbiditiesAlternative causesTreatmentconsiderationsImplications for AssessmentAdminister screening measures; target young children with signs of developmentaldelayDetermine need for further assessmentAssess for presence of symptoms that meet DSM-IV criteriaDetermine stability and duration of symptomsDetermine age of onsetAssess for presence of mental retardationDetermine the influence of autism on school performance, adaptive behavior,development, etc.Assess social relationships and peer social statusObtain developmental and medical historiesDifferential diagnosis/rule out of Rhett’s Disorder, Asperger’s Syndrome,language disorders, mental retardation, and hearing impairmentAssess adaptive behavior assets and deficitsIdentify behavior problems that require interventionAssess caregiver stress and caregiver ability to provide adaptive behaviorinstruction and behavioral problem managementEvaluate response to previous interventionsargues for their routine use with thesepopulations.Classification of the child as havingASDs requires meeting the DSM-IV-TRcriteria outlined earlier. The classificationprocess will depend heavily on caretakerinformation such as detailed histories/interviews,and possibly child behavior ratings.Furthermore, medical, language, and otherassessment data will need to be integrated inorder to make differential diagnoses.The most likely comorbid conditionfor autism is mental retardation. An indepthassessment of intelligence and adaptivebehavior development is, therefore,required in every comprehensive studyof a child suspected of autism or a relateddisorder. Sensory impairments must alsobe ruled out by other clinicians.The process of ruling out alternativecauses is a lengthy one requiringfull participation by medical and otherprofessionals in the assessment process.Neurological problems, in particular,should be ruled out as primary etiologiesfor the child’s behavior. As noted above,the differentiation within types of ASDsis also often quite difficult (see Box 19.2for a specific discussion of the assessmentof Asperger’s Syndrome).A central aspect of psychological interventionfor children with ASDs is skilldevelopment in a variety of domains. Astructured behavioral observation (e.g.,ADOS, functional analysis) and a comprehensivemeasure of adaptive behavior aretypically very useful for defining neededskills and evaluating the effectiveness ofintervention.In Box 19.1, we provide a case exampleof this assessment approach for a child withautism.

chapter 19 Assessment of autism spectrum447Box 19.2Issues in the Assessment of Asperger’s DisorderThe most straightforward distinction betweenAsperger’s Disorder and autism is the lack oflanguage difficulties seen in the former (AmericanPsychiatric Association, 2000). However,this difference is hard to pinpoint for youth whohave average- to above-average intellectualfunctioning (and thus, typically demonstrateno impairments in verbal communication).A careful, detailed history of developmentaldelays is critical in differentiating ASDs fromother forms of delays and for differentiatingamong ASDs. For example, past and presentimpairments in social skills and the presence ofstereotypical behaviors in the absence of a historyof language delays would be more indicativeof Asperger’s Disorder than autism. Therehas been an increased availability of specificscreening tools for Asperger’s Disorder andother ASDs. The Gilliam Asperger’s DisorderScale (GADS; Gilliam, 2001) is a 32-itemscreener that is meant to be completed by aparent, teacher, or clinician. Items load onto four scales: Social Interaction, CognitivePatterns, Restricted Behavioral Patterns, andPragmatic Skills. The Asperger SyndromeDiagnostic Scale (ASDS; Myles, Bock, &Simpson, 2001) consists of 50 items to be completedby a caretaker, teacher, or clinician withfive scales: Cognitive, Language, Social, Maladaptive,and Sensorimotor. Another exampleis the Krug Asperger’s Disorder Index (KADI;Krug & Arick, 2003) which consists of 32items and yields a total score. Research on thedifferential validity is limited.An exception is Campbell’s (2005) studythat compared these rating scales and twoothers that had been used in research. Campbellcited concerns regarding each of thesemeasures. For the ASDS, Campbell notedthat the standardization sample was weak,particularly due to the lack of there being noindependent diagnosis of Asperger’s Disorderin that sample. The GADS was describedas having a stronger standardization sample,although the Asperger’s Disorder diagnosesin this sample were unconfirmed. In addition,internal consistency was less-than-desirable(Campbell, 2005). The KADI was describedas the best of these measures overall in termsof reliability (i.e., internal consistency, stability,and interrater agreement) as well as incontent validity. However, the standardizationof the KADI also lacked informationon independent diagnoses. Additionally, forall three of these scales, the cognitive functioningof individuals in the standardizationsample with autism was not provided, whichis important in determining how well a measureof Aspeger’s Disorder helps distinguishbetween children with Asperger’s Disordervs. high functioning children with autism(Campbell, 2005). Therefore, at present,such measures should be used as one potentialcomponent of assessments for ASDs butin no way replace detailed histories obtainedthrough interviews and behavioral observationsin determining the child’s specificimpairments.A Sample Case of Autism ina Child with NeurologicalImpairmentJulie was a 5-year, 1-month-old girl whowas referred by her parents.Referral InformationJulie was referred for a psychologicalevaluation by her parents. Original referralquestions included concern about delayedlanguage development, short attention span,difficulty in her kindergarten class, and apossible learning disability. Subsequently,

448 chapter 19 Assessment of autism spectrumhowever, Julie had two seizures and wasdiagnosed as having tuberous sclerosis andseizure disorder. On learning that tuberoussclerosis often affects a child’s intelligence,her parents requested a thorough evaluationto determine Julie’s current level offunctioning.Background InformationJulie is a 5-year-old girl who lives withboth parents. Her mother reported thatshe had no serious difficulties when shewas pregnant with Julie. Julie was reportedlyborn full-term and weighed 7 pounds,0 ounces. Both Julie and her mother werereportedly in good condition at the time ofbirth. Julie did, however, have an open soreon her back that was surrounded by a purplemark. The doctor remarked at the timethat this type of wound was very unusual.According to her mother, Julie has hadtwo seizures. The first was reportedlyabout 6 weeks prior to this evaluation. Hermother describes these seizures as lastingabout a minute and not involving severeconvulsions. Julie was seen by a neurologist.According to Julie’s medical reports,she was somewhat difficult to understandand hyperactive during that medical examination.Her doctor had difficulty obtainingher cooperation. Her physical conditionappeared to be normal, except for a largescar on her right hip and skin lesions on herface and extremities. Based on Julie’s physicalcondition, recent seizures, EEG results,and CAT scan results, she was diagnosedas having seizure disorder and tuberoussclerosis, a genetic disorder associated withtumors in the brain and internal organs andskin lesions on the face and body. Her doctorprescribed Tegretol for seizure control.According to her mother, Julie has alsohad frequent ear infections, beginning at8 months of age and continuing until 3months ago. At that time, she had tubesplaced in her ears. She has reportedly hadno difficulties since then. She had a hearingexamination 2 months ago that indicated aslight hearing loss in the right ear. Julie hasalso had frequent colds, according to hermother. Julie’s vision is reportedly normal.Julie’s mother described Julie’s developmentas inconsistent. She seemed to learngross motor skills at a normal rate, buther fine motor and language developmentappeared to be delayed. She reportedlyspoke her first words at 9 months, but by 18months, she could still speak only four words.Her mother states that Julie did not speak insentences until she was 2½ years old. Herspeech articulation is reportedly clear, buther language is still delayed. She was reportedlyin speech therapy until recently.Her mother stated that Julie is sweet, hasa good sense of humor, and enjoys singing.Julie, however, can also be stubborn andsometimes does not seem to listen to or obeyher mother. Her mother noted that Julielikes to play by herself rather than with otherchildren. She also reported that Julie canbe easily overstimulated in that she seemsuncomfortable in crowded or noisy places, isvery energetic and active, has a short attentionspan, and is impulsive (e.g., not waitingfor her turn, grabbing objects at stores withoutpermission). She added that Julie seemsto require a great deal of parental attention.Julie has reportedly been in a preschoolprogram since she was 2 years old. Hermother stated that Julie’s teachers have beenconcerned because Julie did not communicatein school, had difficulty doing ageappropriatework, required much teacherattention, and seemed to lack self-esteemand confidence. Julie is now enrolled in akindergarten class. Her current teacherreported that Julie is a sweet child and seemsto want to please others. Julie reportedlymemorizes well and knows many songs.Julie’s teacher also stated, however, thatJulie shows many behaviors that make itdifficult to teach her. She reportedly spendsmost of the day by herself, talking only toherself or inanimate objects. She seems torepeat, over and over, phrases that she hasheard adults say. According to her teacher,

chapter 19 Assessment of autism spectrum449Julie does not interact with the other children.When called on in class, she seemsembarrassed and then blurts out anyanswer. She reportedly has a very shortattention span and also sometimes leavesthe room without teacher permission. Herteacher stated that she is teased by theother children because of her behavior,which includes chewing on her mat and herclothes. According to her teacher, Julie hasnot learned many academic skills. She seemsto have learned a few basic concepts, such asshapes. She can also make a capital “A,” butit is frequently upside down. She reportedlywrites and works with other objects upsidedown frequently and does not seem to recognizethat she is doing so. She reportedlycannot complete worksheets, as she justdraws lines up and down the page or colorsthe whole page one color. According to herteacher, Julie sometimes does better withone-on-one assistance, but at other times,she will not accept help from her teacher.Julie was briefly observed at school duringlunch. She was sitting at the end of thetable by herself and did not interact withthe other children. Although she had talkedwith the observer previously, she showedlittle response to her at this time. Whenasked if she remembered the observer, shesaid that she did, but she still did not interactwith her. She just continued eating.Behavioral ObservationsJulie was tested over a 2-day period. Threeexaminers participated in her evaluation.She was brought to the first testing sessionby her mother. She seemed very uncomfortablemeeting strangers and clung toher mother. Her mother carried her to thetesting room and stayed for a while untilJulie was more comfortable. Julie did notbecome relaxed with the examiners forsome time. Adequate rapport for testingpurposes was finally established.During testing, Julie was very active anddistractible. The examiners had difficultykeeping her attention on the test materials.She required much direction and attentionfrom the examiners to stay on task. She frequentlywanted to play with the other toys.At one point, she sat on the floor, facing thewall. When asked what she was doing, shesaid she was in time-out. On further questioning,she responded that she was put therebecause she did not pay attention. Later,she again sat on the floor, this time repeating“Stephanie, it’s time to go to school.”The examiner had difficulty redirecting herfrom this behavior. This type of behaviorseemed to occur most when Julie was beingasked to perform a task that was difficult forher. After this last time, Julie could not beenticed to pay any more attention to thetest materials. She got up and left the roomwithout the examiner’s permission.Julie was more cooperative during thesecond testing session. She seemed muchmore comfortable with the examiners andaccompanied them readily to the testingroom. She was, however, still very distractibleand active, requiring much examinerattention for her to stay on task.Test Results and InterpretationsIntellectual FunctioningJulie’s intellectual functioning was far, towell below, that of her same-aged peers.This level of functioning is consistent withher teacher’s and mother’s reports of herability and with her current school performance.Her performance across subtestswas consistent, showing no relativestrengths or weaknesses.Visual Motor IntegrationOn a test of visual motor integration, Julie’sperformance was at or above that of 8%of children of her age, suggesting that hervisual motor integration and fine motorskills are well below average. Her motherand teacher also reported difficulty with finemotor tasks, such as drawing and writing.

450 chapter 19 Assessment of autism spectrumAcademic AchievementOn a standardized test of academic achievement,only two subtests were able to beadministered. Her performance on thesubtests administered was well to far belowaverage. On these subtests and informalassessment, Julie showed the ability to singthe alphabet and count objects with oneto-onecorrespondence.Adaptive BehaviorThe Vineland-2 is a measure of Julie’sability to take care of herself, get alongwith others, and live in the communityappropriately for her age. It includes fourdomains: Communication, Daily LivingSkills, Socialization, and Motor Skills. Onthe Vineland-2, Julie’s mother indicatedthat Julie’s overall ability in these domainsis slightly below average for her age. Mostof her scores were consistent with heroverall Adaptive Composite score. Herscore in the Socialization domain wasslightly higher than the others. This score,however, is not consistent with teacher andparent reports or behavioral observations.Julie’s teacher completed the Vineland-2,Classroom Edition. Her teacher indicatedthat Julie’s adaptive behavior is variable.Julie’s score on the Daily Living Skillsdomain from her teacher’s report is similarto that given by her mother and isin the Below-Average range. However,her other scores are in the Well Below-Average range, more consistent with herintelligence test scores. Her Socializationdomain score is inconsistent with that ofher mother’s rating but more consistentwith other assessment findings. It indicatesfar below- average socialization skills forJulie at school relative to same-aged peers.Overall, the report of Julie’s mother andteacher on the Vineland-2 indicates thatJulie’s adaptive behavior is below what wouldbe expected for her age but is somewhat aboveher intellectual ability as assessed in this evaluation.However, it does not appear that a diagnosisof Mental Retardation is appropriate atthis point, as Julie’s adaptive functioning is notfar enough below what would be expected forher age to warrant such a diagnosis.Rating ScalesJulie’s mother and father completed theAchenbach CBCL, whereas her teachercompleted the TRF. This rating scaleassesses behavioral and emotional problemsof children. Although the ratersreported similar behavioral patterns, onlyJulie’s father indicated that Julie has anyserious problems. He rated Julie as havingproblems in the areas of social withdrawaland hyperactivity. Julie’s motherand teacher also rated her high in theseareas, but not significantly so. These findingsare consistent with teacher and parentreport and behavioral observations inthat Julie has some incidents of apparentoveractivity in the classroom and consistentlyseems uncomfortable interactingwith others. Based on Julie’s current emotional,behavioral, social, intellectual, andadaptive functioning, as well as her historyof communicative delays, a diagnosis ofPervasive Developmental Disorder, NotOtherwise Specified seems appropriate.Summary and ConclusionsJulie is a 5-year-old girl referred by her parentsto determine her current level of functioning.She has been diagnosed as havingseizure disorder and tuberous sclerosis.Her mother reported that her languageand fine motor development were delayed.Her teacher reported that she made littleprogress in adjusting to kindergarten orlearning basic academic skills. Parent andteacher reports indicate that she is activeand inattentive and requires much adultattention and supervision. These behaviorswere also observed in the testing situation,in which examiners had much difficultygetting Julie to cooperate and attend tothe tasks. She was very active and distractible.Parent and teacher reports, as well asobservation during testing, also suggest

chapter 19 Assessment of autism spectrum451that Julie displays several unusual behaviors,such as talking to herself or inanimateobjects, repeating the same phrases ormotions, panting heavily at times, chewingon clothes, preferring to stay by herself,and not interacting with other children.Most of Julie’s test results, includingintellectual, achievement, visual motor, andsome adaptive behavior areas, are in theWell Below-Average range. Other adaptivebehavior scores, however, were onlyslightly below average for her age. Informationfrom rating scales indicates that Julietends to be socially withdrawn, hyperactive,distractible, slow to learn or adapt to newsituations, and sometimes fussy too.Taking all of the assessment informationtogether, the diagnosis of PervasiveDevelopmental Disorder–Not OtherwiseSpecified seems appropriate. At this time,a diagnosis of mental retardation does notseem applicable. Although her intelligencetest scores are within the mild mentallyretarded range, her adaptive behaviorscores lie outside the mental retardationrange. Both of these areas of functioningmust be significantly below average to warranta diagnosis of Mental Retardation.RecommendationsJulie should be evaluated again at the end ofthis academic year to assess her progress. Atthis time the following recommendations aremade for Julie:1. Julie would benefit from a highly structurededucational environment whereshe can get the attention and supervision,she needs, to learn. Her parentsshould consult school personnel in orderto determine the most appropriate placementfor her2. Julie will need much individual instruction,gradually moving to small-groupinstruction3. Julie’s classroom assignments shouldbe within the range appropriate for herintellectual and academic levels4. Julie should be allowed to learn andwork at her own pace5. Instructional tasks should be organizedinto short, structured units6. Julie will need to be taught at a veryconcrete and practical level, usingmany manipulatives and teaching aids7. Skills and concepts will need to berepeated and reviewed often for Julieto master them8. Julie may be helped by being tutored bysomeone who has been trained in tutoringchildren with her set of difficulties9. Julie’s parents and teachers shouldcontinue to search for activities thatshe does well and enjoys and encourageher in these activities10. Julie should be provided withincreased opportunities to learn andpractice age-appropriate self-careand socialization skills11. Julie’s parents and teachers should continueto use behavioral managementstrategies, such as positive reinforcementand time-out, to help her learnappropriate behavior. Her parents maywish to attend sessions with a mentalhealth specialist to become more facilewith these techniques12. Inappropriate behavior should beignored as much as possible, whileappropriate behavior is rewarded withpraise and other appropriate reinforcers13. Communication between school andhome should be maintained. A homeschoolnote, in which Julie is rewardedat home and for appropriate behavior atschool, may be helpful14. Julie should be allowed as much as possibleto be around other children herown age, so that she can learn socializationand develop behavioral skills byobservation and interaction with them15. Julie’s parents should closely monitorher physical functioning and symptomsand maintain regular contact with herphysicians regarding her seizures16. Julie should be evaluated in approximately1 year to determine her level

452 chapter 19 Assessment of autism spectrumof functioning, particularly after theabove strategies have been implemented.ConclusionsAutism and its related disorders are amongthe most difficult of the developmentaldisorders to reliably classify, yet classificationis central to research efforts aimed atprevention and intervention. Considerableprogress has been made in defining thethree core dimensions of impairment thatseparate this condition from other developmentaldisabilities and in developingassessment tools geared toward evaluatingthese dimensions. Many of these toolsappear to be useful in devising interventionstrategies for individuals with ASDs.Many of the available instruments alsodifferentiate ASDs from other developmentaldisabilities, but this differentiatingis particularly difficult for very youngchildren and those with more severe developmentaldelays. While new methods ofassessment have made strides in terms ofdifferential validity, considerably moreprogress is needed. The most promisingmethod for assessing and classifying autismat this time is a thorough and standardinterview of the child’s primary caregiveralong with either a structured observationthat systematically evaluates symptomsof ASDs and/or a functional analysisthat evaluates the communicative intentof a child’s behavioral problems. Complementingthe diagnostic process is the useof intelligence and adaptive behavior scalesuch as the Vineland-2.In summary, current recommendations(e.g., de Bildt et al., 2004; Ozonoffet al., 2005) call for a combination of adetailed developmental history (e.g., ADI-R), behavioral observation (e.g., ADOS),assessment of cognitive functioning, andassessment of adaptive functioning.Chapter Summary1. Autistic Disorder is part of a class ofproblems known as Pervasive DevelopmentalDisorders (PDD) or AutismSpectrum Disorders (ASDs)2. The DSM emphasizes three classesof symptoms that are central to thedisorder: social interaction problems,communication problems, and repetitiveand stereotyped behaviors3. Much remains to be learned aboutrisk factors that may be etiologicallyinvolved in autism4. Differentiation within ASDs and of autismfrom Pervasive Developmental Disorder– Not Otherwise Specified (PDD-NOS)and Mental Retardation is difficult5. The most frequent comorbid problemfor autism is mental retardation6. The Autism Diagnostic Interview –Revised (ADI-R) is a semi-structureddiagnostic interview that is designed tobe used with a child’s primary caregiver.It provides for a detailed history of thedomains of functioning relevant to ASDs7. The Autism Diagnostic ObservationSchedule (ADOS) is designed for theassessment of autism through structuredobservation in a classroom or clinical setting.Four modules are available basedon the individual’s expressive communicationskills and developmental level8. Assessment strategies such as functionalanalysis can be useful in determiningthe communicative intent of problembehaviors in children with ASDs and maydirectly translate to intervention plans9. There has been an increase in ratingscales aimed at screening for Asperger’sDisorder. However, more research onthe psychometrics and utility of thesemeasures is needed10. The within-syndrome variability associatedwith autism suggests that childrenwho are either suspected of, or diagnosedwith, the disorder should have access tomultidisciplinary assessment procedures

R e f e r e n c e sAAMR. (1992). Mental retardation: Definition, classificationand systems of supports (9th ed.). Annapolis, MD:Author.Abad, J., Forns, M., Amador, J. A., & Martorell, B.(2000). Youth self-report reliability and validity in anadolescent sample. Psicothema, 12(1), 49–54.Abidin, R. R. (1986). Parenting stress index manual (2nded.). Charlottesville, VA: Pediatric Psychology Press.Abidin, R. R. (1995). Parenting Stress Index: professionalmanual (3rd ed.). Odessa, FL: PsychologicalAssessment.Abidin, R., Flens, J.R., & Austin, W.G. (1995). The ParentingStress Index. In R.P. Archer (Ed.), Forensic usesof clinical assessment instruments (pp. 297-328). Mahwah,NJ: Erlbaum.Abikoff, H., & Klein, R. G. (1992). Attention-deficithyperactivity and conduct disorder: Comorbidity andimplications for treatment. Journal of Consulting andClinical Psychology, 60, 881–892.Ablow, J. C., Measelle, J. R., Kraemer, H. C., Harrington,R., Luby, J., Smider, N., et al. (1999). TheMacArthur three-city outcome study: Evaluatingmulti-informant measures of young children’s symptomatology.Journal of the American Academy of Child& Adolescent Psychiatry, 38, 1580–1590.Abramowitz, A. J., & O’Leary, S. G. (1991). Behavioralinterventions for the classroom: Implications forstudents with ADHD. School Psychology Review, 20,220–234.Achenbach, T. M. (1982). Assessment and taxonomyof children’s behavior disorders. In B. B. Lahey & A.E. Kazdin (Eds.), Advances in clinical child psychology(Vol. 5, pp. 1–38). New York: Plenum.Achenbach, T. M. (1986). Child behavior checklist-directobservation form (rev. ed.). Burlington: University ofVermont.Achenbach, T. M. (1991a). Integrative guide for the 1991CBCL/4–18, YSR, and TRF profiles. Burlington: Universityof Vermont, Department of Psychiatry.Achenbach, T. M. (1991b) Manual for the child behaviorchecklist/4–18 and 1991 profile. Burlington: Universityof Vermont, Department of Psychiatry.Achenbach, T. M. (1991c). Manual for the teacher’s reportform and 1991 profile. Burlington: University of Vermont,Department of Psychiatry.Achenbach, T. M. (1991d). Manual for the youth selfreport& 1991 profile. Burlington: University of Vermont,Department of Psychiatry.Achenbach, T. M. (1993). Empirically based taxonomy:How to use syndromes and profile types derived from theCBCL/4–18, TRT & YSR. Burlington: University ofVermont, Department of Psychiatry.Achenbach, T. M. (1995). Empirically based assessment& taxonomy: Applications to clinical research. PsychologicalAssessment, 7, 261–274.Achenbach, T. M. (1997). Manual for the young adult selfreportand young adult behavior checklist. Burlington:University of Vermont, Department of PsychiatryAchenbach, T. M., & Rescorla, L. A. (2000). Manual forthe ASEBA preschool forms and profiles. Burlington, VT:Research Center for Children, Youth, and Families.Achenbach, T. M. (2001). Manual for the ASEBA schoolageforms and profiles. Burlington, VT: Research Centerfor Children, Youth, and Families.Achenbach, T. M., Conners, C. K., Quay, H. C., Verhulst,F. C., & Howell, C. T. (1989). Replication of empiricallyderived syndromes as a basis for taxonomy ofchild/adolescent psychopathology. Journal of AbnormalChild Psychology, 17, 299–323.Achenbach, T. M., & Edelbrock, C. (1978). The classificationof child psychopathology: A review and analysis ofempirical efforts. Psychological Bulletin, 85, 1275–1301.Achenbach, T. M., & Edelbrock, C. (1983). Manual forthe child behavior checklist and revised child behavior profile.Burlington: University of Vermont, Departmentof Psychiatry.Achenbach, T. M., & Edelbrock, C. (1986). Manual forthe teacher’s report form and teacher- version of the child453P.J. Frick et al., Clinical Assessment of Child and Adolescent Personality and Behavior,DOI 10.1007/978-1-4419-0641-0_5, © Springer Science+Business Media, LLC 2009

454 referencesbehavior profile. Burlington: University of Vermont,Department of Psychiatry.Achenbach, T. M., & Edelbrock, C. (1991). Manual forthe youth self report and profile. Burlington: Universityof Vermont Department of Psychiatry.Achenbach, T. M., & Howell, C. T. (1993). Are Americanchildren’s problems getting worse? A 13-yearcomparison. Journal of the American Academy of Childand Adolescent Psychiatry, 32, 1145–1154.Achenbach, T. M., Howell, C. T., McConaughy, S. H.,& Stanger, C. (1995a). Six-year predictors of problemsin a national sample of children & youth: I. Crossinformantsyndromes. Journal of the American Academyof Child and Adolescent Psychiatry, 34, 336–347.Achenbach, T. M., Howell, C. T., McConaughy, S. H.,& Stanger, C. (1995b). Six-year predictors of problemsin a national sample of children & youth: II.Signs of disturbance. Journal of the American Academyof Child and Adolescent Psychiatry, 34, 488–498.Achenbach, T. M., Howell, C. T., McConaughy, S. H.,& Stanger, C. (1998). Six-year predictors of problemsin a national sample: IV. Young adult signs of disturbance.Journal of the American Academy of Child andAdolescent Psychiatry, 37, 718–727.Achenbach, T. M., Howell, C. T., Quay, H. C., &Conners, C. K. (1991). National survey of problemsand competencies among four-to sixteen-year-olds:Parents’ reports for normative and clinical samples.Monographs of the Society for Research in Child Development,56(3), 1–131.Achenbach, T. M., & McConaughy, S. H. (1987).Empirically based assessment of child and adolescentpsychopathology: Practical applications. Newbury Park,CA: Sage.Achenbach, T. M., & McConaughy, S. H. (1996). Relationsbetween DSM-1V and empirically based assessment.School Psychology Review, 25, 330–342.Achenbach, T. M., & McConaughy, S. H. (1997).Empirically based assessment of child & adolescent psychopathology(2nd ed.). Thousand Oaks, CA: Sage.Achenbach, T. M., McConaughy, S. H., & Howell, C.T. (1987). Child/adolescent behavioral and emotionalproblems: Implications of cross-informant correlationsfor situational specificity. Psychological Bulletin,101, 213–232.Achenbach, T. M., & Rescorla, L. A. (2000). Manual forthe ASEBA preschool forms & profiles. Burlington: Universityof Vermont, Department of Psychiatry.Adrien, J. L., Barthelemy, C., Perrot, A., & Roux, S.(1992). Validity and reliability of the infant behavioralsummarized evaluation (IBSE): A rating scalefor the assessment of young children with autism anddevelopmental disorders. Journal of Autism and DevelopmentalDisorders, 22, 371–394.AERA, APA, & NCME. (1999). Standards for educationaland psychological testing. Washington, DC:AERA.Agras, W. S., Chapin, H. N., & Oliveau, D. C. (1972).The natural history of phobias: Course and prognosis.Archives of General Psychiatry, 26, 315–317.Albano, A. M., & Silverman, W. K. (1996). Anxiety disordersinterview schedule for DSM-IV: Clinician manual.New York: The Psychological Corporation.Albert-Corush, J., Firestone, P., & Goodman, J. T.(1986). Attention and impulsivity characteristics ofthe biological and adoptive parents of hyperactive andnormal children. American Journal of Orthopsychiatry,56, 413–423.Alderman, E., & Kennedy, C. (1995). The right to privacy.New York: Alfred A. Knopf.Allen, J. C., & Hollifield, J. (1990). Using the Rorschachwith children and adolescents: The Exner ComprehensiveSystem. In C. R. Reynolds & R. W. Kamphaus(Eds.), Handbook of psychological and educationalassessment of children: Personality, behavior, and context(pp. 168–185). New York: Guilford.Allen, J. C., & Hollifield, J. (2003). Using the Rorschachwith children and adolescents: The Exner ComprehensiveSystem. In C. R. Reynolds & R. W. Kamphaus(Eds.), Handbook of psychological and educationalassessment of children: Personality, behavior, and context– 2nd edition (pp. 182–197). New York: Guilford.Allen, M. J., & Yen, W. M. (1979). Introduction to measurementtheory. Long Grove, IL: Waveland.Altepeter, T. S., & Breen, M. J. (1989). The home situationsquestionnaire (HSQ) and the school situationsquestionnaire (SSQ): Normative data and an evaluationof psychometric properties. Journal of PsychoeducationalAssessment, 7, 312–322.Aman, M. G., Werry, J. S., Fitzpatrick, J., Lowe, M., &Waters, J. (1983). Factor structure and norms for therevised behavior problem checklist on New Zealandchildren. Australian and New Zealand Journal of Psychiatry,17, 354–360.Aman, M. G. (1994). Instruments for assessing treatmenteffects in developmentally delayed populations.Assessment in Rehabilitation and Exceptionality,1, 1–19.Amanat, E., & Butler, C. (1984). Oppressive behaviorsin the families of depressed children. Family Therapy,11, 65–77.Amato, P. R., & Keith, B. (1991). Parental divorce andthe well-being of children: A meta-analysis. PsychologicalBulletin, 110, 26–46.Ambrosini, P. J. (2000). Historical development andpresent status of the schedule for affective disordersand schizophrenia for school-age children(K-SADS). Journal of the American Academy of Childand Adolescent Psychiatry, 39, 39–48.American Educational Research Association. AmericanPsychological Association, and National Councilon Measurement in Education. (1999). Standards foreducational and psychological testing. Washington, DC:Author.

eferenes455American Psychiatric Association. (1968). The diagnosticand statistical manual of mental disorders (2nd ed.).Washington, DC: Author.American Psychiatric Association. (1980). The diagnosticand statistical manual of mental disorders (3rd ed.).Washington, DC: Author.American Psychiatric Association. (1987). 7 he diagnosticand statistical manual of mental disorders (3rd ed., rev.).Washington, DC: Author.American Psychiatric Association. (1994). The diagnosticand statistical manual of mental disorders (4th ed.).Washington, DC: Author.American Psychiatric Association. (2000). The diagnosticand statistical manual of mental disorders (4th ed.,text rev.). Washington, DC: Author.American Psychological Association. (1985). Standardsfor educational and psychological testing. Washington,DC: Author.American Psychological Association. (1992). Ethicalprinciples of psychologists and code of conduct.American Psychologist, 47, 1612–1628.American Psychological Association. (1994). Guidelinesfor child custody evaluations in divorce proceedings.American Psychologist, 49, 677–680.Ames, L. B., Metraux, R., Rodell, J. L., & Walker, R.(1974). Child Rorschach responses (rev. ed.). New York:Bruner/Mazel.Anastasi, A. (1992). Introductory remarks. In K. F. Geisinger(Ed.), Psychological testing of Hispanics (pp. 1–7).Washington, DC: American Psychological Association.Anastasi, A. (1988). Psychological testing (16 th ed). NewYork: Macmillan.Anastasi, A., & Urbina, S. (1998). Psychological testing(7th ed.). New York: Prentice-Hall.Anderson, J. C., Williams, S., McGee, R., & Silva, P. A.(1987). DSM-III disorders in preadolescent children.Archives of General Psychiatry, 44, 69–76.Angold, A., & Costello, J. (2000). The child and adolescentpsychiatric assessment (CAPA). Journal of theAmerican Academy of Child and Adolescent Psychiatry,39, 49–58.Angold, A., Costello, E. J., & Erkanli, A. (1999).Comorbidity. Journal of Child Psychology and Psychiatry& Allied Disciplines, 40, 57–87.Angold, A., Erkanli, A., Egger, H. I., & Costello, F. J.(2000). Stimulant treatment for children: A communityperspective. Journal of the American Academy ofChild and Adolescent Psychiatry, 39, 975–983.Applegate, B., Lahey, B. B., Hart, E. L., Waldman, I.,Biederman, J., Hynd, G. W., et al. (1997). The age ofonset criterion for DSM-W attention-deficit hyperactivitydisorder: A report of the DSM-IV field trials.Journal of the American Academy of Child and AdolescentPsychiatry, 36, 1211–1221.Archer, R. P. (1992). MMPI-A: Assessing adolescent psychopathology.Hillsdale, NJ: Erlbaum.Archer, R. P. (1997). MMPI-A: Assessing adolescent psychopathology(2nd ed.). Mahwah: Lawrence Erlbaum.Archer, R. P. (2005). Implications of MMPI/MMPI-Afindings for understanding adolescent developmentand psychopathology. Journal of Personality Assessment,85, 257–270.Archer, R. P., Belevich, J. K. S., & Elkins, D. E. (1994).Item-level and scale-level factor structures of theMMPI-A. Journal of Personality Assessment, 62, 332–345.Archer, R. P., & Elkins, D. E. (1999). Identification ofrandom responding on the MMPI-A. Journal of PersonalityAssessment, 73, 407–421.Archer, R. P., Handel, R. W., & Lynch, K. D. (2001).The effectiveness of MMPI-A items in discriminatingbetween normative and clinical samples. Journal of PersonalityAssessment, 77, 420–435.Archer, R. P., & Krishnamurthy, R. (1994). A structuralsummary approach for the MMPI-A: Developmentand empirical correlates. Journal of Personality Assessment,63, 554–573.Archer, R., Krishnamurthy, R., & Stredny, R. (2007).The Minnesota Multiphasic Personality Inventory-Adolescent.In S. R. Smith (Ed.), The clinicalassessment of children and adolescents: A practitioner’shandbook (pp. 237–266). Mahwah, NJ: Lawrence ErlbaumAssociates.Archer, R. P., & Krishnamurthy, R. (1997). MMPI-Aand Rorschach indices related to depression and conductdisorder: An evaluation of the incremental validityhypothesis. Journal of Personality Assessment, 69,517–533.Archer, R. P., & Newsom, C. R. (2000). Psychological testusage with adolescent clients: Survey update. Assessment,7, 227–235.Archer, R. P., & Slesinger, D. (1999). MMPI-A patternsrelated to the endorsement of suicidal ideation. Assessment,6(1), 51–59.Arita, A. A., & Baer, R. A. (1998). Validity of selectedMMPI-A content scales. Psychological Assessment, 10,59–63.Arney, F., Rogers, H., Baghurst, P., Sawyer, M., & Prior,M. (2008). The reliability and validity of the parentingscale for Australian mothers of pre-school agedchildren. Australian Journal of Psychology, 60, 44–52.Arnold, D. S., O’Leary, S. G., Wolff, L. S., & Acker, M.M. (1993). The Parenting Scale: A measure of dysfunctionalparenting in discipline situations. PsychologicalAssessment, 5, 137–144.Aronen, E. T., Teerikangas, O. M., & Kurkela, S. A. (1999).The continuity of psychiatric symptoms from adolescenceinto young adulthood. Nordic Journal of Psychiatry,53(5), 333–338.Arseneault, L., Kim-Cohen, J., Taylor, A., Caspi, A., &Moffit, T. (2005). Psychometric evaluation of 5- and7-year-old children’s self-reports of conduct problems.Journal of Abnormal Child Psychology, 33(5), 537–550.

456 referencesAsher, S. R., & Hymel, S. (1981). Children’s social competencein peer relations: Sociometric and behavioralassessment. In J. D. Wine & M. D. Smye (Eds.), Socialcompetence (pp. 125–157). New York: Guilford Press.Asher, S. R., Singleton, L. C., Tinsley, B. R., & Hymel,S. (1979). A reliable sociometric measure for preschoolchildren. Developmental Psychology, 15, 443–444.Assing, R. (1998, August). BASC parent rating scale(Spanish translation): Validity study. Paper presentedat the annual meeting of the American PsychologicalAssociation, San Francisco.Aten, J. D., Madson, M. B., & Kruse, S. J. (2008). Thesupervision genogram: A tool for preparing supervisors-in-training.Psychotherapy: Theory, Research, Practice,Training, 45, 111–116.Atkins, M. S., Pelham, W. E., & Licht, M. (1985). Acomparison of objective classroom measures andteacher ratings of attention deficit disorder. Journal ofAbnormal Child Psychology, 13, 155–167.Atkinson, L. (1990). Standard errors of prediction forthe Vineland adaptive behavior scales. Journal of SchoolPsychology, 28, 255–259.August, G. J., McDonald, A. W., Realmuto, G. M., &Skare, S. S. (1996). Hyperactive and aggressive pathways:Effect of demographic, family, and child characteristicson children’s adaptive functioning. Journal ofClinical Child Psychology, 25, 341–351.Baer, R. A., Ballenger, J., & Kroll, L. S. (1998). Detectionof underreporting on the MMPI-A in clinicaland community samples. Journal of Personality Assessment,71, 98–113.Bailey, A., LeCouteur, A., Gottesman, I., Bolton, P.,Simonoff, E., Yuzda, E., et al. (1995). Autism as astrongly genetic disorder: Evidence from a Britishtwin study. Psychological Medicine, 25, 63–77.Baldwin, J. A., & Bell, Y. R. (1985). The African-AmericanSelf-Consciousness Scale: An Africentric personalityquestionnaire. Western Journal of Black Studies,9, 61–68.Baldwin, J. S., & Dadds, M. R. (2008). Anxiety disorders.In M. Hersen (Series Ed.) & D. Reitman (Vol.Ed.), Handbook of Psychological Assessment, Case Conceptualization,and Treatment (pp. 231–263). Hoboken,NJ: John Wiley & Sons.Ball, J. D., Archer, R. P., Gordon, R. A., & French, J. (1991).Rorschach depression indices with children and adolescents:Concurrent validity findings. Journal of PersonalityAssessment, 57, 465–476.Bao-Yu, W., & Lin-Yan, S. (2004). Family environment andthe development of behavior in children with ADHD.Chinese Journal of Clinical Psychology, 12, 145–146.Barker, P. (1990). Clinical interviews with children and adolescents.New York: Norton.Barkley, R. A. (1981). Hyperactive children: A handbookfor diagnosis and treatment. New York: GuilfordPress.Barkley, R. A. (1988). Attention-deficit disorder withhyperactivity. In E. J. Mash & L. G. Terdal (Eds.), Behavioralassessment of childhood disorders (2nd ed., pp. 69–104).New York: Guilford Press.Barkley, R. A. (1990). Attention-deficit hyperactivity disorder:A handbook to diagnosis and treatment (2nd ed.).New York: Guilford Press.Barkley, R. A. (1991). The ecological validity of laboratoryand analogue assessment methods of ADHD.Journal of Abnormal Child Psychology, 19, 149–178.Barkley, R. A. (1997a). Attention deficit hyperactivitydisorder. In E. J. Mash & L. G. Terdel (Eds.), Assessmentof childhood disorders (3rd ed., pp. 71–129). NewYork: Guilford Press.Barkley, R. A. (1997b). ADHD and the nature of selfcontrol.New York: Guilford Press.Barkley, R. A., & Edelbrock, C. (1987). Assessing situationalvariation in children’s problem behaviors: Thehome and school situations questionnaires. In R. J.Prinz (Ed.), Advances in behavioral assessment of childrenand families (Vol. 3, pp. 157–176). New York:JAI.Barkley, R. A., Fischer, M., Smallish, L., Fletcher, K.(2002). The persistence of attention deficit hyperactivitydisorder into adulthood as a function of reportingsource and definition of disorder. Journal of AbnormalPsychology, 111, 279–289.Barkley, R. A., Guevremont, D. G., Anastopoulos, A.D., & Fletcher, K. F. (1993). Driving related risks andoutcomes of attention-deficit hyperactivity disorder inadolescents and young adults. Pediatrics, 92, 212–218.Barlow, D. H. (2002). Anxiety and its disorders: The natureand treatment of anxiety and panic (2nd ed.). NewYork:Guilford.Baron-Cohen, S., Allen, J., & Gillberg, C. (1992). Canautism he detected at 18 months? The needle, thehaystack, and the CHAT. British Journal of Psychiatry,161, 839–843.Barringer, M. S. (2000). Assessment of ADHD-hyperactivesub-type using visual continuous performance,psychophysiological, and behavior measures. A dissertationsubmitted to the Office of Graduate Studies ofTexas A & M University.Barrios, B. A. (1988). Fears and anxiety. In E. J. Mash & L.G. Terdal (Eds.), Behavioral assessment of childhood disorders(2nd ed., pp. 196–264). New York: Guilford Press.Barrios, B. A. (1993). Direct observation. In F. H. Ollendick& M. Hersen (Eds.), Handbook of child and adolescentassessment (pp. 140–164). Boston, MA: Allyn& Bacon.Barrios, B. A., & Hartmann, D. P. (1988). Fears andanxiety. In E. J. Mash & L. G. Terdal (Eds.), Behavioralassessment of child-hood disorders (pp. 167–221).New York: Guilford Press.Barry, C. T., Frick, P. J., & Grafeman, S. J. (in press).Child versus parent report of parenting practices:Implications for the conceptualization of childbehavioral and emotional problems. Assessment.Barry, T., Dunlap, S., Cotten, S., Lochman, J., & Wells,K. (2005). The influence of maternal stress and dis-

eferenes457tress on disruptive behavior problems in boys. Journalof the American Academy of Child & Adolescent Psychiatry,44, 265–273.Bauer, D. H. (1976). An exploratory study of developmentalchanges in children’s fear. Journal of Child Psychologyand Psychiatry, 17, 69–74.Baumrind, D. (1971). Harmonious parents and their preschoolchildren. Developmental Psychology, 4, 99–102.Beck, S. T., Steer, R. A., & Garbin, M. G. (1988). Psychometricproperties of the Beck depression inventory:Twenty-five years of evaluation. Clinical PsychologyReview, 8, 77–100.Befera, M. S., & Barkley, R. A. (1985). Hyperactive andnormal girls and boys: Mother-child interactions,parent psychiatric status, and child psychopathology.Journal of Child Psychology and Psychiatry, 26,439–452.Behar, L. B., & Stringfield, S. A. (1974). A behavior ratingscale for the preschool child. Child Development,10, 601–610.Bell, N. L., & Nangle, R. J. (1999). Interpretive issueswith the Roberts apperception test for children:Limitations of the standardization group. Psychologyin the Schools, 36, 277–283.Bellak, L., & Bellak, S. (1949a). A manual for the children’sapperception test. Larchmont, NY: CPS.Bellak, L., & Bellak, S. S. (1949b). The children’s apperceptiontest. New York: CPS.Belter, R. W., & Piotrowski, C. (1999). Current statusof master’s-level training in psychological assessment.Journal of Psychological Practice, 5, 1–5.Belardinelli, C., Hatch, J. P., Olvera, R. L., Fonseca, M.,Caetano, S. C., Nicoletti, M., et al. (2008). Familyenvironment patterns in families with bipolar children.Journal of Affective Disorders, 107, 299–305.Bender, W. N. (1997). Medical interventions and schoolmonitoring. In W. N. Bender (Ed.), UnderstandingADHD: A practical guide for teachers and parents (pp.107–122). Upper Saddle River, NJ: Merrill.Benson, J. (1998). Developing a strong program of constructvalidation: A test anxiety example. EducationalMeasurement: Issues and Practice, 17, 10–22.Berg, I. (1993). Aspects of school phobia. In C. G. Last(Ed.), Anxiety across the lifespan: A developmental perspective(pp. 78–93). New York: Springer.Bergman, L. R., & Magnusson, D. (1997). A personorientedapproach in research on developmental psychopathology.Development and Psychopathology, 9,291–319.Bertrand, J., Mars, A., Boyle, C., Bove, F., Yeargin-Allsopp, M., & Decoufle, P. (2001) Prevalence ofautism in the United States population: The bricktownship New Jersey investigation. Pediatrics, 108,1155–1161.Berument, S. K., Rutter, M., Lord, C., Pickles, A., &Bailey, A. (1999). Autism screening questionnaire:Diagnostic validity. British Journal of Psychiatry, 175,441–451.Beutler, L. E., & Malik, M. (Eds.). (2002). Rethinkingthe DSM: Psychological perspectives. Washington, DC:American Psychological Association.Bickman, L. (2002). The death of treatment as usual: Anexcellent first step on a long road. Clinical PsychologyScience and Practice, 9, 195–198.Biederman, J., Gao, H., Rogers, A., & Spencer, T. (2006).Comparison of parent and teacher reports of attentiondeficit/hyperactivitydisorder symptoms from twoplacebo-controlled studies of atomoxetine in children.Biological Psychiatry, 60, 1106–1110.Biederman, J., Keenan, K., & Faraone, S. V. (1990).Parent-based diagnosis of attention deficit disorderpredicts a diagnosis based on teacher report. Journal ofthe American Academy of Child and Adolescent Psychiatry,29, 698–701.Bigras, M., LaFrenier, P. J., & Dumas, J. E. (1996). Discriminantvalidity of the parent-child scales of theparenting stress index. Early Education and Development,7, 167–178.Bird, H. R., Gould, M. S., & Stagheeza, B. (1992). Aggregatingdata from multiple informants in child psychiatryepidemiological research. Journal of the AmericanAcademy of Child and Adolescent Psychiatry, 31, 78–85.Bird, H. R., Gould, M. S., & Staghezza, B. M. (1993)Patterns of diagnostic comorbidity in a communitysample of children aged 9 through 16 years. Journal ofthe American Academy. of Child and Adolescent Psychiatry,32, 361–368.Bird, H. R., Yager, T. J., Staghezza, B., Gould, M. S.,Canino, G., & Rubio-Stipec, M. A. (1990). Impairmentin the epidemiological measurement of childhoodpsychopathology in the community. Journal ofthe American Academy of Child and Adolescent Psychiatry,29, 796–803.Birmaher, B., Arbelaez, C., & Brent, D. (2002). Courseand outcome of child and adolescent major depressivedisorder. Child and Adolescent Clinics of North America,11, 619–638.Black, K. (1994). A critical review of the MMPI-A.Child Assessment News, 4(1), 9–12.Blader, J. C. (2004). Symptom, family, and service predictorsof children’s psychiatric rehospitalization withinone year of discharge. Journal of the American Academyof Child & Adolescent Psychiatry, 43, 440–451.Blair, R. J. R. (1999). Responsiveness to distress cues inthe child with psychopathic tendencies. Personalityand Individual Differences, 27, 135–145.Blashfield, R. K. (1984). The classification of psychopathology:Neo-Kraepelinian and quantitative approaches.New York: Plenum.Block, J. H. (1965). The child rearing practices report.Berkeley: Institute of Human Development, Universityof California-Berkeley.Block, J. H., Block, J., & Morrison, A. (1981). Parentalagreement-disagreement on child-rearing orientationsand gender-related personality correlates in children.Child Development, 52, 965–974.

458 referencesBloomquist, M. L., August, G. J., Cohen, C., Doyle,A., & Ever-hart, K. (1997). Social problem solvingin hyperactive-aggressive children: How and whatthey think in conditions of automatic and controlledprocessing. Journal of Clinical Child Psychology, 26,172–180.Blum, G. S. (1950). The blacky pictures: Manual of instructions.New York: The Psychological Corporation.Bölte, S., & Poustka, F. (2002). The relation betweengeneral cognitive level and adaptive behavior domainsin individuals with autism with and without co-morbidmental retardation. Child Psychiatry & HumanDevelopment, 33, 165–172.Borgen, F. H., & Barnett, D. C. (1987). Applying clusteranalysis in counseling psychology research. Journal ofCounseling Psychology, 34, 456–468.Bornstein, M. R., Bellack, A. S., & Hersen, M. (1977).Social-skills training for unassertive children: A multibaselineanalysis. Journal of Applied Behavioral Analysis,10, 183–195.Bower, E. M. (1982). Severe emotional disturbance:Public policy and research. Psychology in the Schools,19, 55–60.Bowlby, J. (1969). Attachment and loss: I. Attachment.New York: Basic Books.Boyd, C. P., Gullone, E., Needleman, G. L., & Burt, T.(1997). The family environment scale: Reliability andnormative data for an adolescent sample. Family Process,36, 369–373.Braaten, E. (2007). The child clinician’s report writinghandbook. New York: Guilford.Bracken, B. A., Keith, L. K., & Walker, K. C. (1994).Assessment of preschool behavior and social-emotionalfunctioning: A review of thirteen third-partyinstruments. Assessment in Rehabilitation and Exceptionality,1, 331–346.Bracken, B. A., Keith, L. K., & Walker, K. C. (1998).Assessment of preschool behavior and social-emotionalfunctioning: A review of thirteen third-party instruments.Journal of Psychoeducational Assessment, 16, 153–169.Brandt, H. M., & Giebink, J. W. (1968). Concretenessand congruence in psychologists’ reports to teachers.Psychology in the Schools, 5, 87–89.Briggs-Gowan, M. J., Carter, A. S., & Schwab-Stone,M. (1996). Discrepancies among mother, child, andteacher reports: Examining the contributions ofmaternal depression and anxiety. Journal of AbnormalChild Psychology, 24, 749–765.Brody, G., Stoneman, Z., & Burke, M. (1987). Familysystem and individual child correlates of sibling behavior.American Journal of Orthopsychiatry, 57, 561–569.Brody, G., Stoneman, Z., & Burke, M. (1988). Childtemperament and parental perceptions of individualchild adjustment: An intrafamilial analysis. AmericanJournal of Orthopsychiatry, 58, 532–542.Brown, R. A. (1999). Assessing attitudes and behaviorsof high-risk adolescents: An evaluation of the selfreportmethod. Adolescence, 34, 25–32.Brown, T. E. (2001). Brown attention-deficit disorder scalesfor children and adolescents. San Antonio, TX: PsychologicalCorporation.Brubaker, R. G. & Szakowski, A. (2000). Parentingpractices and behavior problems among deaf children.Child & Family Behavior Therapy, 22, 13–28.Bruininks, R. J., Hill, B. K., Weatherman, R. F., &Woodcock, R. N. (1986). Inventory for client and agencyplanning. Allen, TX: DLM Teaching Resources.Bruininks, R. J., Woodcock, R. W., Weatherman, R. F.,& Hill, B. K. (1986). Scales of independent behavior.Allen, TX: DLM Teaching Resources.Bruininks, R. H., Woodcock, R. W., Weatherman, R.E., & Hill, B. K. (1996). Scales of independent behavior– revised. Itasca, IL: Riverside.Budd, K. S., Heilman, N. E., & Kane, D. (2000). Psychosocialcorrelates of child abuse potential in multiplydisadvantaged adolescent mothers. Child Abuse &Neglect, 24, 611–625.Burns, G. L., & Patterson, D. R. (1990). Conductproblem behaviors in a stratified random sample ofchildren and adolescents: New standardization dataon the Eyberg child behavior inventory. PsychologicalAssessment: A Journal of Consulting and Clinical Psychology,2, 391–397.Burns, R. C., & Kaufman, S. H. (1970). Kinetic familydrawings (KF-D): An introduction to understandingchildren through kinetic drawings. New York: Bruner/Mazel.Buss, A. H. (1995). Personality: Temperament, socialbehavior, and the self. Boston: Allyn and Bacon.Bussing, R., Schumann, E., Belin, T. R., Widawski, M.,& Perwien, A. R. (1998). Diagnostic utility of twocommonly used ADHD screening measures amongspecial education students. Journal of the AmericanAcademy of Child and Adolescent Psychiatry, 37, 74–82.Butcher, J. N., Cabiyo, J., Lucio, E., & Garrido, M.(2007). Assessing Hispanic clients using the MMPI-2and MMPI-A. Washington, DC: American PsychologicalAssociation.Butcher, J. N., Williams, C. L., Graham, J. R., Archer, R. P.,Tellegen, A., Ben-Porath, Y. S., et al. (1992). MIJPI-A,Minnesota multiphasic personality inventory-adolescent:Manual for administration scoring, and interpretation.Minneapolis: University of Minnesota Press.Cai, X., Kaiser, A., & Hancock, T. (2004). Parent andteacher agreement on child behavior checklist itemsin a sample of preschoolers from low-income andpredominantly African American families. Journal ofClinical Child and Adolescent Psychology, 33, 303–312.Campbell, J. (2005). Diagnostic assessment of Asperger’sdisorder: A review of five third-party rating scales.Journal of Autism and Developmental Disorders, 35,25–35.Campbell, M., Kafantaris, V., Malone, R. P., & Kowalik,S. C. (1991). Diagnostic and assessment issuesrelated to pharmacotherapy for children and adolescentswith autism. Special issue: Current perspectives

eferenes459in the diagnosis, assessment, and treatment of childand adolescent disorders. Behavior Modification, 15,326–354.Campbell, S. B. (1986). Developmental issues in childhoodanxiety. In R. Gittelman (Ed.), Anxiety disordersof childhood (pp. 24–57). New York: Guilford Press.Campbell, S. B. (1990). Behavior problems in preschoolchildren. New York: Guilford Press.Canino, G., & Bravo, M. (1999). The translation andadaptation of diagnostic instruments for cross-culturaluse. In D. Shaffer and C. P. Lucas (Eds.), Diagnosticassessment in child and adolescent psychopathology (pp.285–298). New York: Guilford Press.Cantwell, D. P. (1996). Classification of child and adolescentpsychopathology. Journal of Child Psychology andPsychiatry, 37, 3–12.Cantwell, D. P., Lewinsohn, P. M., Rohds, P., & Seeley,J. R. (1997). Correspondence between adolescentreport and parent report of psychiatric diagnostic data.Journal of the American Academy of Child and AdolescentPsychiatry, 36, 610–619.Capaldi, D. M. (1992). Co-occurrence of conduct problemsand depressive symptoms in early adolescent boys: II. A2-year follow-up at Grade 8. Development and Psychopathology,4, 125–144.Cardell, C. P., & Parmer, R. S. (1988). Teacher perceptionof temperament characteristics of childrenclassified as learning disabled. Journal of Learning Disabilities,21, 497–502.Carlson, C. L., Lahey, B. B., & Neeper, R. (1984). Peerassessment of the social behavior of accepted, rejected,and neglected children. Journal of Abnormal Child Psychology,12, 189–198.Caron, C., & Rutter, M. (1991). Comorbidity in childpsychopathology: Concepts, issues, and researchstrategies. Journal of Child Psychology and Psychiatry,32, 1063–1080.Carrion, V. G., Weems, C. F., Ray, R., & Reiss, A. L.(2002). Toward an empirical definition of pediatricPTSD: The phenomenology of PTSD symptoms inyouth. Journal of the American Academy of Child & AdolescentPsychiatry, 41, 166–173.Carroll, A., Houghton, S., Taylor, M., West, J., & List-Kerz, M. L. (2006). Responses to interpersonal andphysically provoking situations: The utility and applicationof an observation schedule for school-aged studentswith and without attention deficit/hyperactivitydisorder. Educational Psychology, 26, 483–498.Carter, C. L., & Dacey, C. M. (1996). Validity of theBeck’s depression inventory, MMPI, and Rorschachin assessing adolescent depression. Journal of Adolescence,19, 223–231.Casat, C., Norton, H., & Boyle-Whitesel, M. (1999).Identification of elementary school children at riskfor disruptive behavioral disturbance: Validation of acombined screening method. Journal of the AmericanAcademy of Child & Adolescent Psychiatry, 38, 1246–1253.Cassen, E. H. (1990). Depression and anxiety secondaryto medical illness. In G. Winokur (Ed.), The psychiatricclinics of North America: Anxiety and depression assecondary phenomena (pp. 597–612). Philadelphia, PA:W. B. Saunders.Castellanos, F. X., Lee, P. P., Sharp, W., Jeffries, N. O.,Greenstein, D. K., Clasen, L. S., et al. (2002). Developmentaltrajectories of brain volume abnormalitiesin children and adolescents with attention deficithyperactivity disorder. Journal of the American MedicalAssociation, 288, 1740–1748.Cavell, T. A., Meehan, B. T., & Fiala, S. E. (2003).Assessing social competence in children and adolescents.In C. R. Reynolds and R. W. Kamphaus (Eds.),Handbook of psychological and educational assessment ofchildren (2nd ed., pp. 433–454). New York: Guilford.Chambers, W. J., Puig-Antich, J., Hirsch, M., Paez, P.,Ambrosini, P. J., Tabrizi, M. A., et al. (1985). Theassessment of affective disorders in children andadolescents by semi-structured interview: Test-retestreliability of the schedule for affective disorders andschizophrenia for school-aged children, present episode.Archives of General Psychiatry, 42, 696–702.Chandler, L. A. (1990). The projective hypothesis andthe development of projective techniques for children.In C. R. Reynolds & R. W. Kamphaus (Eds.), Handbookof psychological and educational assessment of children:Personality, behavior, & context (pp. 55–69). NewYork: Guilford Press.Chapieski, I., Friedman, A., & Lachar, D. (2000). Psychologicalfunctioning in children and adolescentswith Sturge-Weber syndrome. Journal of Child Neurology,15(10), 660–665.Chilcoat, H. D., & Breslau, N. (1997). Does psychiatrichistory bias mothers’ reports? An application of a newanalytic approach. Journal of the American Academy ofChild and Adolescent Psychiatry, 36, 971–979.Christ, M. G. (1990). A four-year longitudinal study ofdepression in boys: Phenomenology and comorbidity ofconduct disorder. Doctoral dissertation: University ofGeorgia, Athens.Christensen, A., Margolin, G., & Sullaway, M. (1992).Inter-parental agreement on child behavior problems.Psychological Assessment, 4, 419–425.Cicchetti, D., Lord, C., Koenig, K., Klin, A., & Volkmar,F. (2008). Reliability of the ADI-R: Multiple examinersevaluate a single case. Journal of Autism and DevelopmentalDisorders, 38, 764–770.Cicchetti, D., & Toth, S. L. (1998). The development ofdepression in children and adolescents. American Psychologist,53, 221–241.Cicchetti, D. & Walker, E. F. (2001). Stress and development:Biological and psychological consequences.Development and Psychopathology, 13, 413–418.Claiborn, C. D. (1995). Review of the Minnesota multiphasicpersonality inventory – Adolescent. In: J. C.Conoley & J. Impara (Eds.), Twelfth mental measurementsyearbook. Lincoln, NE: Burns Institute.

460 referencesClerkin, S. M., Marks, D. J., Policaro, K., & Halperin, J.M. (2007). Psychometric properties of the Alabamaparenting questionnaire-preschool revision. Journal ofClinical Child and Adolescent Psychology, 36, 19–28.Cohen, N. J., Gotlieb, H., Kershner, J., & Wehrspann,W. (1985). Concurrent validity of the internalizingand externalizing profile patterns of the Achenbachchild behavior check-list. Journal of Consulting andClinical Psychology, 53, 724–728.Cohen, P., & Kasen, S. (1999). The context of assessment:Culture, race, and socioeconomic status asinfluences on the assessment of children. In D. Shaffer,C. P. Lucas, & J. E. Richters (Eds.), Diagnosticassessment in child and adolescent psychopathology (pp.299–318). New York: Guilford Press.Coie, J.D. & Dodge, K.A. (1983). Cotinuities andchanges in children’s social status: A five yearlongitudinal study. Merrill-Palmer Quarterly, 29,261–282.Coie, J. D., Dodge, K. A., & Coppotelli, H. (1982).Dimensions and types of social status: A cross-ageperspective. Developmental Psychology, 18, 557–570.Coie, J.D. & Kupersmidt, J.B. (1983). A behavioral analysisof emerging social status in boys’ groups. ChildDevelopment, 54, 1400–1416.Collett, B., Ohan, J., & Myers, K. (2003). Ten-yearreview of rating scales. VI: Scales assessing externalizingbehaviors. Journal of the American Academy of Child& Adolescent Psychiatry, 42, 1143–1170.Conduct Problems Prevention Research Group. (1999).Initial impact of the fast track prevention trial forconduct problems: I. The high-risk sample. Journal ofConsulting and Clinical Psychology, 67, 631–647.Connell, J. P. (1985). A new multidimensional measure ofchildren’s perception of control. Child Development, S6,1018–1041.Conners, C. K. (1970). Symptom patterns in hyperkinetic,neurotic, and normal children. Child Development,41, 667–682.Conners, C. K. (1995). Conners continuous performancetest. Toronto: Multi-Health Systems.Conners, C. K. (1997). Conners rating scales -revised.Toronto: Multi-Health Systems.Conners, C. K. (2008a). Conners (3rd ed.). Toronto:Multi-Health Systems.Conners, C. K. (2008b). Conners comprehensive behaviorrating scales. Toronto: Multi-Health Systems.Conners, C. K. (1999). Conners Rating Scales-Revised.In M. E. Maruish (Ed.), The use of psychological testingfor treatment planning and outcomes assessment (2nded., pp. 467–495). Mahwah, NJ: Lawrence ErlbaumAssociates, Inc.Conoley, J. C., & Conoley, C. W. (1991). Collaborationfor child adjustment: Issues for school and clinicbasedchild psychologists. Journal of Consulting andClinical Psychology, S9, 821–829.Constantino, G., & Malgady, R. G. (1983). Verbal fluencyof Hispanic, Black, & White children on TATand TEMAS, a new thematic apperception test. HispanicJournal of Behavioral Sciences, 5, 199–206.Constantino, G., Malgady, R. G., Rogler, L. H. (1988).Technical manual: TEMAS thematic apperception test.Los Angeles, CA: Western Psychological Services.Cooper, D. K. (1990). Developmental trends in children’sdepression inventory responses across age and gender.Master’s thesis, University of Georgia, Athens.Corbett, J. A., & Trimble, M. R. (1983). Epilepsy andanticonvulsant medication. In M. Rutter (Ed.),Developmental neuropsychiatry (pp. 112–132). NewYork: Guilford Press.Costa, P. T., & McCrae, R. R. (1985). The NEO personalityinventory manual. New York: Psychological AssessmentResources.Costello, A. J. (1983). The NIMH diagnostic interviewschedule for children. Pittsburgh, PA: University ofPittsburgh.Costa, N. M., & Weems, C. W. (2005). Maternal andchild anxiety: Do attachment beliefs or children’sperceptions of maternal control mediate their association?Social Development, 14, 574–590Costello, A. J., Edelbrock, C. S., Dulcan, M. D., Kalas,R., & Marie, S. H. (1984). Report of the NIMH diagnosticinterview schedule for children (DISC). Washington,DC: National Institute of Mental Health.Costello, E. J. (1989). Child psychiatric disorders andtheir correlates: A primary care pediatric sample.Journal of the American Academy of Child and AdolescentPsychiatry, 28, 851–888.Costello, E. J., Edelbrock, C. S., & Costello, A. J. (1985).Validity of the NIMH diagnostic interview schedulefor children: A comparison between psychiatric andpediatric referrals. Journal of Abnormal Child Psychology,13, 579–595.Costello, E. J., Egger, H., & Angold, A. (2005). 10-yearresearch update review: The epidemiology of childand adolescent psychiatric disorders: I. Methods andpublic health burden. Journal of the American Academyof Child and Adolescent Psychiatry, 44, 972–986.Courchesne, E., Karns, C. M., Davis, B. S., Ziccardi,R., Carper, R. A., Tigue, Z. D., et al. (2001). Unusualbrain growth patterns in early life in patients withautistic disorder: An MRI study. Neurology, 57, 245–254.Crick, N. R., & Dodge, K. A. (1996). Social informationprocessing mechanisms in reactive and proactiveaggression. Child Development, 67, 993–1002.Crijnen, A. A. M., Achenbach, T. M., & Verhulst, F. C.(1997). Comparisons of problems reported by parentsof children in 12 cultures: Total problems, externalizing,and internalizing. Journal of the American Academyof Child and Adolescent Psychiatry, 36, 1269–1277.Crijnen, A. A. M., Achenbach, T. M., & Verhulst, F. C.(1999). Problems reported by parents of children inmultiple cultures: The child behavior checklist syndromeconstructs. American Journal of Psychiatry, 156,569–574.

eferenes461Crnic, K. A., & Reid, M. (1989). Mental retardation.In E. J. Mash & R. A. Barkley (Eds.), Treatment ofchildhood disorders (pp. 247–346). New York: GuilfordPress.Crooks, C. V. & Wolfe, D. A. (2004). Child abuseand neglect. In E. J. Mash & R. A. Barkleyl (Eds.),Assessment of childhood disorders (4th ed., pp. 639–684). New York: Guilford Press.Cuellar, L., Harris, I. C., & Jasso, R. (1980). An acculturationscale for Mexican American normal andclinical population. Hispanic Journal of Behavioral Science,2, 199–217.Cumella, E. J., Wall, A. D., & Kerr-Almcida, N. (1999).MMPI-A in the inpatient assessment of adolescentswith eating disorders. Journal of Personality Assessment,73, 31–44.Current Population Survey. (2001). Washington, DC:U.S. Census Bureau.Cummings, E. M., Keller, P. S., & Davies, P. T. (2005).Towards a family process model of maternal and paternaldepressive symptoms: Exploring multiple relationswith child and family functioning. Journal of Child Psychologyand Psychiatry, 46, 479–489.Cummings, J. A. (1986). Projective drawings. In H. M.Knoll (Ed.), The assessment of child and adolescent personality(pp. 199–244). New York: Guilford Press.Cunningham, C. E. & Boyle, M. H. (2002). Preschoolersat risk for attention-deficit hyperactivity disorderand oppositional defiant disorder: Family, parenting,and behavioral correlates. Journal of Abnormal ChildPsychology, 30, 555–569.Curry, J. F., & Craighead, W. E. (1993). Depression. InT. H. Ollendick & M. Hersen (Eds.), Handbook ofchild and adolescent assessment (pp. 251–268) Boston,MA: Allyn and Bacon.Cytryn, L., & McKnew, D. H. (1974). Factors influencingthe changing clinical expression of the depressiveprocess in children. American Journal of Psychiatry,131, 879–881.Dadds, M. R., Barrett, P. M., Rapee, R. M., & Ryan, S.(1996). Family process and child anxiety and aggression:An observational analysis. Journal of AbnormalChild Psychology, 24, 715–734.Dadds, M. R., Maujean, A., & Fraser, J. A. (2003). Parentingand conduct problems in children: Australiandata and psychometric properties of the Alabamaparenting questionnaire. Australian Psychologist, 38,238–241.Dadds, M. R., & Sanders, M. R. (1992). Family interactionand child psychopathology: A comparison oftwo observation strategies. Journal of Child and FamilyStudies, 1, 371–391.Dahlstrom, W. G. (1993). Small samples, large consequences.American Psychologist, 48, 393–399.Dalby, J. T. (1985). Taxonomic separation of attentiondeficit disorders. Contemporary Educational Psychology,10, 228–234.Dalton, J. E. (1996). Juvenile male sex offenders: Meanscores and the BASC self-report of personality. PsychologicalReports, 79, 634.Dana, R. H. (1993). Multicultural assessment perspectivesfor professional psychology. Boston, MA: Allyn andBacon.Dana, R. H. (2000). Handbook of cross-cultural and multiculturalpersonality assessment. Mahwah, NJ: LawrenceErlbaum.Danforth, J. S., & DuPaul, G. J. (1996). Interraterreliability of teacher rating scales for children withattention-deficit hyperactivity disorder. Journal of Psychopathologyand Behavioral Assessment, 18, 227–238.Daniel, M. H. (1992, August). Item-level covariancestructure analysis for constructing personality/behavioralscales. Paper presented at the annual meeting of theAmerican Psychological Association, Washington,DC.Daniel, M. H. (1993, April). Diagnostic specificity of parents’vs. teachers’ behavior ratings. Paper presented atthe Annual Meeting of the National Association ofSchool Psychologists, Washington, DC.Daniel, M. H. (1993, August). Diagnostic specificity ofparents’ and teachers’ behavior ratings. Paper presentedat the annual meeting of the American PsychologicalAssociation, Washington, DC.Darling, N., & Steinberg, L. (1993). Parenting style ascontext: An integrative model. Psychological Bulletin,113, 487–496.David, C. F., & Kistner, J. A. (2000). Do positive selfperceptionshave a “dark side”? Examination of thelink between perceptual bias and aggression. Journalof Abnormal Child Psychology, 28, 327–337.David-Ferdon, C., & Kaslow, N. J. (2008). Evidencebasedpsychosocial treatments for child and adolescentdepression.Journal of Clinical Child and AdolescentPsychology, 37, 62–104.de Bildt, A., Sytema, S., Ketelaars, C., Kraijer, D., Mulder,E., Volkmar, F., et al. (2004). Interrelationshipsbetween the ADOS, ADI-R, and DSM-IV-TR classificationin children and adolescents with mentalretardation. Journal of Autism and Developmental Disorders,34, 129–137.De Groot, A., Koot, H. M., & Verhulst, F. C. (1996).Cross-cultural generalizability of the youth self-reportand teacher’s report form cross-informant syndromes.Journal of Abnormal Child Psychology, 24, 651–664.De Los Reyes, A. & Kazdin, A. E. (2005). Informantdiscrepancies in the assessment of childhood psychopathology:A critical review, theoretical framework,and recommendations for further study. PsychologicalBulletin, 131, 483–509.Deater-Deckard, K., Reiss, D., & Hetherington, F.M. (1997). Dimensions and disorders of adolescentadjustment: A quantitative genetic analysis of unselectedsamples and selected extremes. Journal of ChildPsychology and Psychiatry, 38, 515–525.

462 referencesDeBaryshe, B. D., Patterson, G. R., & Capaldi, D. M.(1993). A performance model for academic achievementin early adolescent boys. Developmental Psychology,29, 795–804.Department of Health and Human Services. (2000).Standards for privacy of individually identifiable healthinformation. Federal Register, 65(250), 82462–82829.Departments of Psychiatry and Child Psychiatry, TheInstitute of Psychiatry, & The Maudsley HospitalLondon (1987). Psychiatric examination: Notes oneliciting and recording clinical in-formation psychiatricparents. Oxford: Oxford University Press.Desombre, H., Malvy, J., Roux, S., de Villard, R., Sauvage,D., Dalery, J., et al. (2006). Autism and developmentaldelay: A comparative clinical study in veryyoung children using IBSE scale. European Child &Adolescent Psychiatry, 15, 343–351.DeStefano, L. (1998). Review of the Personality Inventoryfor Youth. Thirteenth Mental Measurements Yearbook.Lincoln, NE: Buros Institute.DeStefano, L., & Thompson, D. S. (1990). Adaptivebehavior: The construct and its measurement. In C.R. Reynolds & R. W. Kamphaus (Eds.), Handbookof psychological and educational assessment of children:Personality, behavior, and context (pp. 445–469). NewYork: Guilford Press.Dishion, T. J., Patterson, G. R., & Skinner, M. S. (1989).A process model for the role of peers in adolescent socialadjustment. Paper presented at the biennial meetingof the Society for Research in Child Development,Kansas City, MO.Dishion, T. J., Capaldi, D., Spracklen, K. M., & Li, F.(1995). Peer ecology of male adolescent drug use.Development and Psychopathology, 7, 803–824.DiSibio, M. (1993). Conjoint effects of intelligence andadaptive behavior on achievement in a nonreferredsample. Journal of Psychoeducational Assessment, 11,304–313.Dodge, K. A. (1983). Behavioral antecedents of peersocial crams. Child Development, 54, 1386–1399.Dodge, K. A., & Coie, J. D. (1987). Social-informationprocessing factors in reactive and proactive aggression inchildren’s peer groups. Journal of Personality and Social Psychology,53, 1146–1158.Dodge, K. A., Coie, J., Brakkc, N. (1982). Behavior patternsof socially rejected and neglected adolescents:The roles of social approach and aggression. Journal ofAbnormal Child Psychology, 10, 389–410.Dodge, K. A., McClaskey, C. L., & Feldman, E. (1985).Situational approach to the assessment of social competencein children. Journal of Consulting and ClinicalPsychology, 53, 344–353.Dodge, K. A., & Pettit, G. S. (2003). A biopsychosocialmodel of the development of chronic conduct problemsin adolescence. Developmental Psychology, 39, 349–371.Doleys, D. M. (1977). Behavioral treatments for nocturnalenuresis in children: A review of the recent literature.Psychological Bulletin, 84, 30–54.Doll, B. (1998). Review of the Child Behavior Checklist.Thirteenth Mental Measurements Yearbook. Lincoln,NE: Buros Institute.Doll, E. A. (1940). A social basis of mental diagnosis.Journal of Applied Psychology, 24, 160–169.Doll, E. A. (1953). Measurement of social competence: Amanual for the Vineland Social Maturity Scale. CirclePines, MN: American Guidance Services.Dopke, C. A., Lundahl, B. W., Dunsterville, E., & Lovejoy,M. C. (2003). Interpretations of child compliancein individuals at high- and low-risk for child physicalabuse. Child Abuse & Neglect, 27, 285–302.Downey, G., & Coyne, J. C. (1990). Children ofdepressed parents: An integrative review. PsychologicalBulletin, 108, 50–76.Doyle, A., Ostrander, R., Skare, S., Crosby, R. D., &August, C. (1997). Convergent and criterion-relatedvalidity of the behavior assessment system for children-parentrating scales. Journal of Clinical ChildPsychology, 26, 276–284.Dresden, J. (1994). Gender, temperament, and mathematicsachievement. Doctoral dissertation, University ofGeorgia, Athens.Drotar, D., Stein, R. E. K., & Perrin, E. C. (1995).Methodological issues in using the child behaviorchecklist and its related instruments in clinical childpsychology research. Journal of Clinical Child Psychology,24, 184–192.DuBois, P. H. (1970). A history of psychological testing.Boston, MA: Allyn and Bacon.Duhig, A., Renk, K., Epstein, M., & Phares, V. (2000).Interparental agreement on internalizing, externalizing,and total behavior problems: A meta-analysis.Clinical Psychology: Science and Practice, 7, 435–453.Durbin, C., Hayden, E., Klein, D., & Olino, T. (2007).Stability of laboratory-assessed temperamental emotionalitytraits from ages 3 to 7. Emotion, 7, 388–399.Dunbar, J. L. (1999). Differential item performance bygender on the externalizing scales of the behaviorassessment system for children. Dissertation AbstractsInternational Section A: Humanities and Social Sciences,60(6-A), 1902.DuPaul, G. J., Anastopolous, A. D., Shelton, T. L.,Guevremont, I. C., & Metevia, L. (1992). Multimethodassessment of attention-deficit hyperactivitydisorder: The diagnostic utility of clinic-based tests.Journal of Clinical Child Psychology, 21, 394 402.DuPaul, G. J., Power, T. J., Anastopolous, A. D., & Reid,R. (1998). ADHD-IV rating scale: Checklists, norms,and clinical interpretation. New York: Guilford.DuPaul, G. J., Rapport, M. D., & Perrillo, L. M. (1991).Teacher ratings of academic skills: The developmentof the academic performance rating scale. School PsychologyReview, 20, 284–300.DuPaul, G.J. & Stoner, G. (1994). ADHD in theschools: Assessment and intervention strategies.New York: Guilford.

eferenes463Eaves, L. J., Silberg, J. L., Hewitt, J. K., Rutter, M.,Meyer, J. M., Neale, M. C., et al. (1993). Analyzingtwin resemblance in multisymptom data: Geneticapplications of a latent class model for symptoms ofconduct disorder in juvenile boys. Behavior Genetics,23, 5–19.Eaves, R. C., & Milner, B. (1993). The criterion-relatedvalidity of the childhood autism rating scale and theautism behavior checklist. Journal of Abnormal ChildPsychology, 21, 481–491.Edelbrock, C. (1979). Empirical classification of children’sbehavior disorders: Progress based on parent andteacher ratings. School Psychology Digest, 8, 355–369.Edelbrock, C. (1985). Age differences in the reliabilityof the psychiatric interview of the child. Child Development,56, 265–275.Edelbrock, C., & Costello, A. J. (1988). Convergencebetween statistically derived behavior problem syndromesand child psychiatric diagnoses. Journal ofAbnormal Child Psychology, 16, 219–231.Edelbrock, C., Costello, A. J., Dulcan, M. K., Conover,N. C., & Kalas, R. (1986). Parent child agreement onchild psychiatric symptoms assessed via structuredinterview. Journal of Child Psychology and Psychiatry,27, 181–190.Edelbrock, C., Costello, A. J., Dulcan, M. K., Kalas, R., &Conover, N. C. (1985). Age differences in the reliabilityof the psychiatric interview of the child. Child Development,56, 265–275.Edelbrock, C., Crnic, K., & Bohnert, A. (1999). Interviewingas communication: An alternative way ofadministering the diagnostic interview schedule forchildren. Journal of Abnormal Child Psychology, 27,447–453.Edelsohn, G., Ialongo, N., Werthamer-Larrson, L.Crockett, L., & Kellam, S. (1992). Self-reporteddepressive symptoms in first-grade children: Developmentallytransient phenomena? Journal of theAmerican Academy of Child and Adolescent Psychiatry,31, 282–290.Education for All Handicapped Children’s Act of 1975,P.L. 94–142, 20 U.S.C. Section 1401 (1975).Eiraldi, R. B., Power, T. J., Karustis, J. L., & Goldstein, S.G. (2000). Assessing ADHD and comorbid disordersin children: The Child Behaviour Checklist and theDevereaux Scales of Mental Disorders. Journal of ClinicalChild Psychology, 29, 3–16.Eisenstadt, T. H., Eyberg, S., McNeil, C. B., Newcomb,K., & Funderbunk, B. (1993). Parent-child interactiontherapy with behavior problem children: Relativeeffectiveness of two stages and overall treatment outcome.Journal of Clinical Child Psychology, 22, 42–51.Elgar, F. J., Waschbusch, D. A., Dadds, M. R., & Sivaldason,N. (2007). Development and validation of ashort form of the Alabama parenting questionnaire.Journal of Child and Family Studies, 16, 243–259.Elliott, S., & Bretzing, B. (1980). Using and updatinglocal norms. Psychology in the Schools, 17, 196–201.Elliot, D. S., Huizinga, D., & Ageton, S. S. (1985).Explaining delinquency and drug use. Newbury Park,CA: Sage.Emery, R. E. (1982). Interparental conflict and the childrenof discord and divorce. Psychological Bulletin, 92,310–330.Endicott, J., & Spitzer, R. L. (1978). A diagnostic interview:The schedule for affective disorders and schizophrenia.Archives of General Psychiatry, 35, 57–63.Erickson, M. T. (1998). Etiological factors. In T. H.Ollendick & M. Hersen (Eds.), Handbook of ChildPsychopathology (3rd ed., pp. 37–62). New York: PlenumPress.Erikson, E. H. (1968). Identity, youth, and crisis. New York:Norton.Ernst, M., Cookus, B. A., & Moravec, B. C. (2000). Pictorialinstrument for children and adolescents (PICA-III-R). Journal of the American Academy of Child andAdolescent Psychiatry, 39, 94–99.Essau, C. A., Sasagawa, S., & Frick, P. J. (2006). Psychometricproperties of the Alabama parenting questionnaire.Journal of Child and Family Studies, 15,597–616.Essau, C. A., Sasagawa, S., & Frick, P. J. (2006). Callousunemotionaltraits in a community sample of adolescents.Assessment, 13, 454–469.Eyberg, S. M., Nelson, M. M., Duke, M., & Boggs, S.R. (2005). Manual for the dyadic parent-child interactioncoding system (3rd ed.). Available on-line at www.PCIT.org.Exner, J. E. (1974). The Rorschach: A comprehensive system,I. New York: Wiley.Exner, J. E. (1978). The Rorschach: A comprehensive system,II: Recent research and advanced interpretation.New York: Wiley.Exner, J. E. (1983, June). The depression index. 1983Alumni Newsletter (p. 7). Bayville, NY: RorschachWorkshops.Exner, J. E. (1990, April). The depression index. 1990Alumni Newsletter (pp. 12–13). Asheville, NC: RorschachWorkshops.Exner, J. E., & Martin, L. S. (1983). The Rorschach: Ahistory and description of the comprehensive system.School Psychology Review, 12, 407–413.Exner, J. E., Jr., & Weiner, I. B. (1982). The Rorschach: Acomprehensive system, Vol. 3: Assessment of children andadolescents. New York: Wiley.Exner, J. E., & Weiner, I. B. (1994). The Rorschach: Acomprehensive system, Vol. 3: Assessment of children andadolescents (2nd ed.). New York: John Wiley andSons.Exner, J. E., Thomas, E. A., & Mason, B. (1985). Children’sRorschachs: Description and prediction. Journalof Personality Assessment, 49, 13–20.Eyberg, S. M., & Robinson, E. A. (1982). Conductproblem behavior: Standardization of a behavioralrating scale with adolescents. Journal of Clinical ChildPsychology, 12, 347–354.

464 referencesEyberg, S. M., Bessmer, J., Newcomb, K., Edwards, D.,& Robin-son, E. (1994). Dyadic parent-child interactioncoding system 11: A manual. Unpublished manuscript,University of Florida.Eyberg, S. M., & Robinson, E. A. (1983). Dyadic parent-childinteraction coding system: A manual. PsychologicalDocuments, 13, 24 (Ms. No., 2582).Eyberg, S. M., & Ross, A. W. (1978). Assessment of childbehavior problems: The validation of a new inventory.Journal of Clinical Child Psychology, 7, 113–116.Eyde, L. D., Robertson, G. J., Krug, S. E., Moreland, K.L., Robertson, A. G., Shewan, C. M., et al. (1993).Responsible test use: Case studies for assessing humanbehavior. Washington, DC: American PsychologicalAssociation.Eysenck, H. J. (1965). Review of the Wittenborn PsychiatricRating Scales. In O. K. Buros (Ed.), MentalMeasurements Yearbook (5th ed.). Highland Park, NJ:Gryphon Press.Faraone, S. V., Biederman, J., Keenan, K., & Tsuang, M.T. (1991). Separation of DSM-III attention deficitdisorder and conduct disorder: Evidence from a familygenetic study of American child psychiatric patients.Psychological Medicine, 21, 109–121.Federal Register. (March 12, 1999). Assistance to statesfor the education of children with disabilities and theEarly Intervention Program for infants and toddlerswith disabilities; Final Regulations.Feighner, J. P., Robins, E., Guze, S. B., Woodruff, R. A.,Winokur, G., & Munoz, R. (1972). Diagnostic criteriafor use in psychiatric research. Archives of GeneralPsychiatry, 26, 57–63.Feinberg, T. L., Crouse-Novak, M. A., Paulauskas, S.L., & Finkelstein, R. (1984). Depressive disordersin childhood: I. A longitudinal prospective study ofcharacteristics and recovery. Archives of General Psychiatry,41, 229–237.Feinfield, K. A. & Baker, B. L. (2004). Empirical supportfor a treatment program for families of young childrenwith externalizing problems. Journal of Clinical Child& Adolescent Psychology, 33, 182–195.Ferdinand, R. (2008). Validity of the CBCL/YSRDSM-IV scales anxiety problems and affective problems.Journal of Anxiety Disorders, 22, 126–134.Fergusson, D. M., & Horwood, J. (1995). Predictivevalidity of categorically and dimensionally scoredmeasures of disruptive childhood behaviors. Journalof Clinical Child and Adolescent Psychology, 32,396–407.Fergusson, D. M., Swain, N. R., & Horwood, L. J.(2002). Deviant peer affiliations, crime and substanceuse: A fixed effects regression analysis. Journal ofAbnormal Child Psychology, 30, 419–430.Figueroa, R. A. (1990). Assessment of linguistic minoritygroup children. In C. R. Reynolds & R. W. Kamphaus(Eds.), Handbook of psychological and educationalassessment of children: Intelligence and achievement (pp.671–696). New York: Guilford Press.Filipek, P. A., Accardo, P. J., Ashwal, S., Baranek, G. T.,Cook, E. H., Jr., Dawson, G., et al. (2000). Practiceparameter: Screening and diagnosis of autism. Neurology,55, 468–479.Finch, A. J., & Belter, R. W. (1993). Projective techniques.In T. H. Ollendick & M. Hersen (Eds.),Handbook of child and adolescent assessment (pp. 229–238). Boston, LH : Allyn and Bacon.Finch, A. J., Lipovsky, J. A., & Casat, C. D. (1989).Anxiety and depression in children and adolescents:Negative affectivity or separate constructs? In P. C.Kendall & D. Watson (Eds.), Anxiety and depression:Distinctive and overlapping features (pp. 171–202).New York: Academic Press.First, M. B., Spitzer, R. I., Gibbon, M., & Williams, J. B.W. (1995). Structure clinical interview for Axis I DSM-IV Disorders – Patient edition. New York: New YorkState Psychiatric Institute.Fischer, M., Barkley, R. A., Fletcher, K., & Smallish, L.(1993). The adolescent outcome of hyperactive childrendiagnosed by research criteria: V. Predictors ofoutcome. Journal of the American Academy of Child andAdolescent Psychiatry, 32, 324–332.Fisher, L., & Blair, R. J. R. (1998). Cognitive impairmentand its relationship to psychopathic tendenciesin children with emotional and behavioral difficulties.Journal of Abnormal Child Psychology, 26, 511–519.Fisher, P., Wicks, J., Shaffer, D., Piacentini, J., & Laplun,J. (1992). NIMH diagnostic interview schedule for children:Users manual. New York: New York State PsychiatricInstitute.Flanagan, D. P., Alfonso, V. C., Primavera, L. H., Povall,L., & Higgins, D. (1996). Convergent validity of theBASC and SSRS: Implications for social skills assessment.Psychology in the Schools, 33, 13–23.Flanagan, R. (1995). A review of the behavior assessmentsystem for children (BASC): Assessment consistentwith the requirements of the individuals withdisabilities education act (IDEA). Journal of SchoolPsychology, 33, 177–186.Flanagan, R., & DiGiuseppe, R. (1999). Critical reviewof the TEMAS: A step within the development ofthematic apperception instruments. Psychology in theSchools, 36, 21–30.Fletcher, J. M., Francis, D. J., Morris, R. D., & Lyon, G.R. (2005). Evidence-based assessment of learning disabilitiesin children and adolescents. Journal of ClinicalChild and Adolescent Psychology, 34, 506–522.Foley, D. L., Rutter, M., Agnold, A., Pickles, A., Maes,H. M., Silberg, J. L., et al. (2005). Making sense ofinformant disagreement for overanxious disorder.Journal of Anxiety Disorders, 19, 193–210.Fombonne, F. (1992). Diagnostic assessment in a sampleof autistic and developmentally impaired adolescents.Special issue: Classification and diagnosis. Journal ofAutism & Developmental Disorders, 22, 563–581.Forbey, J. D., Handel, R. W., & Ben-Porath, Y. S.(2000). A real-data simulation of computerized adap-

eferenes465tive administration of the MMPI-A. Computers inHuman Behavior, 16, 83–96.Forehand, R., Long, N., & Hedrick, M. (1987). Familycharacteristics of adolescents who display overt andcovert behavior problems. Journal of Behavior Therapyand Experimental Psychiatry, 18, 325–328.Forness, S. R., & Kritzer, J. (1992). A new proposed definitionand terminology to replace “serious emotionaldisturbance” in the individuals with disabilities act. SchoolPsychology Review, 21, 12–20.Forth, A. E., Kosson, D. S., & Hare, R. D. (2003). Harepsychopathy checklist: Youth version. Toronto: Multi-Health Systems.Fowler, P. (1982). Factor structure of the family environmentscale: Effects of social desirability. Journal ofClinical Psychology, 38, 285–292.Francis, D. J., Fletcher, J. M., Sniebing, K. K., Davidson, K. C.,& Thompson, N. M. (1991). Analysis of change: Modelingindividual growth. Journal of Consulting and Clinical Psychology,59, 27–37.Franco, J. N. (1983). An acculturation scale for Mexican–Americanchildren. Journal of General Psychology,108, 175–181.Frazier, T. W., Youngstrom, E. A., Kubu, C. S., Sinclair, L.,& Rezai, A. (2008). Exploratory and confirmatory factoranalysis of the autism diagnostic interview-revised. Journalof Autism and Developmental Disorders, 38, 474–480.Freeman, B. J., Ritvo, E. R., Yokota, A., & Rirvo, A.(1986). A scale for rating symptoms of patients withthe syndrome of autism in real life settings. Journal ofthe American Academy of Child Psychiatry, 25, 130–136.French, J. L., & Hale, R. L. (1990). A history of thedevelopment of psychological and educational testing.In C. R. Reynolds & R. W. Kamphaus (Eds.),Handbook of psychological and educational assessment ofchildren: Intelligence and achievement (pp. 3–28). NewYork: Guilford Press.Freud, S. (1936). The Problem of anxiety. New York: W.W. Norton & Co.Frick, P. J. (1993). Childhood conduct problems in familycontext. School Psychology Review, 22, 376–385.Frick, P. J. (1994). Family dysfunction and the disruptivebehavior disorders: A review of recent empirical findings.In T. H. Ollendick & R. J. Prinz (Eds.), Advancesin clinical child psychology (Vol. 16, pp. 213–226). NewYork: Plenum.Frick, P. J. (1998). Conduct disorders and severe antisocialbehavior. New York: Plenum.Frick, P. J. (2000). Laboratory and performance-basedmeasures of childhood disorders. Journal of ClinicalChild Psychology, 29, 475–478.Frick, P. J. (2004). Integrating research on temperamentand childhood psychopathology. Its pitfalls and promise.Journal of Clinical Child and Adolescent Psychology,33, 2–7.Frick, P. J. (2006). Developmental pathways to conductdisorder. Child and Adolescent Psychiatric Clinics of NorthAmerica, 15, 311–331.Frick, P. J., Bodin, S. D., & Barry, C. T. (2000). Psychopathictraits and conduct problems in community andclinic-referred samples of children: Further developmentof the Psychopathy Screening Device. Assessment,12, 382–393.Frick, P. J., Christian, R. C. & Wootton, J. M. (1999).Age trends in the association between parenting practicesand conduct problems. Behavior Modification, 23,106–128.Frick, P. J., & Dickens, C. (2006). Current perspectiveson conduct disorder. Current Psychiatry Reports, 8,59–72.Frick, P. J. & Hare, R. D. (2001). The antisocial processscreening device. Toronto: Multi-Health Systems.Frick, P. J., & Jackson, Y. K. (1993). Family functioningand childhood antisocial behavior. Yet another reinterpretation.Journal of Clinical Child Psychology, 22,410–419.Frick, P. J., Kamphaus, R. W., Lahey, B. B., Loeber, R.,Christ, M. A. G., Hart, E. L., et al. (1991). Academicunderachievement and the disruptive behavior disorders.Journal of Consulting and Clinical Psychology, 59,289–294.Frick, P. J., Kimonis, E. R., Dandreaux, D. M., & Farell,J. M. (2003). The 4 year stability of psychopathic traitsin non-referred youth. Behavioral Sciences & the Law,21(6), 713–736.Frick, P. J. & Kimonis, E. (2008). Externalizing disorders. InJ. E. Maddux & B. A. Winstead (Eds.), Psychopathology:Contemporary issues, theory, and research (pp. 349–374).Mahwah, NJ: Erlbaum.Frick, P. J., & Lahey, B. B. (1991). The nature and characteristicsof attention-deficit hyperactivity disorder.School Psychology Review, 20, 163–173.Frick, P. J., Lahey, B. B., Applegate, B., Kerdyck, L.,Ollendick, T., Ilynd, G. W., et al. (1994). DSM-IVfield trials for the disruptive behavior disorders. Journalof the American Academy of Child and AdolescentPsychiatry, 33, 529–539.Frick, P. J., Lahey, B. B., Christ, M. G., Loeber, R., & Green,S. (1991). History of childhood behavior problems in biologicalrelatives of boys with attention-deficit-hyperactivitydisorder and conduct disorder. Journal of Clinical ChildPsychology, 20, 445–451.Frick, P. J., Lahey, B. B., Hartdagen, S., & Hynd, C.W. (1989). Conduct problems in boys: Relations tomaternal personality, marital satisfaction, and socioeconomicstatus. Journal of Clinical Child Psychology,18, 114–120.Frick, P. J., Lahey, B. B., Loeber, R., Stouthamer-Loeber,M., Christ, M. G., & Hanson, K. (1992). Familial riskfactors to oppositional defiant disorder and conductdisorder: Parental psychopathology and maternal parenting.Journal of Consulting and Clinical Psychology,60, 49–55.Frick, P. J., Lahey, B. B., Loeber, R., Tannenbaum, L. E.,Van Horn, Y., Christ, M. A. C., et al. (1993). Oppositionaldefiant disorder and conduct disorder: A meta-

466 referencesanalytic review of factor analyses and cross-validation ina clinic sample. Clinical Psychology Review, 13, 319–340.Frick, P. J. & Loney, B. R. (1999). Outcomes of childrenand adolescents with conduct disorder and oppositionaldefiant disorder. In H. C. Quay & A. Hogan (Eds.),Handbook of disruptive behavior disorders (pp. 507–524).New York: Plenum Press.Frick, P. J. & Loney, B. R. (2000). The use of laboratoryand performance-based measures in the assessment ofchildren and adolescents with conduct disorders. Journalof Clinical Child Psychology, 29, 540–554.Frick, P. J. & Loney, B. R. (2002). Understanding theassociation between parent and child antisocial behavior.In R. J. McMahon & R. Dcv. Peters (Eds.), Theeffects of parental dysfunction on children (pp. 105–126).New York: Plenum Press.Frick, P. J., & Marsee, M. A. (2006). Psychopathic traitsand developmental pathways to antisocial behavior inyouth. In C. J. Patrick (Ed.), Handbook of psychopathictraits (pp. 355–374). New York: Guilford Press.Frick, P. J., & McMahon, R. J. (2008). Child and adolescentconduct problems. In J. Hunsley & E. J. Mash(Eds.), A guide to assessments that work (pp. 41–66). NewYork: Oxford University.Frick, P. J., & Morris, A. S. (2004). Temperament anddevelopmental pathways to conduct problems. Journalof Clinical Child and Adolescent Psychology, 33, 54–68.Frick, P. J., & O’Brien, B. S. (1994). Conduct disorders.In R. T. Ammerman & M. Hersen (Eds.), Handbookof child behaviour therapy in the psychiatric setting (pp.199–216). New York: Wiley.Frick, P. J., & Silverthorn, P. (2001). Psychopathologyin children. In H. E. Adams & P. B. Sutker (Eds.),Comprehensive handbook of psychopathology (3rd ed., pp.881–920). New York: Kluwer-Plenum.Frick, P. J., Silverthorn, P., & Evans, C. S. (1994).Assessment of childhood anxiety using structuredinterviews: Patterns of agreement among informantsand association with maternal anxiety. PsychologicalAssessment, 6, 372–379.Frick, P. J. & White, S. F. (2008). The importance ofcallous-unemotional traits for the development ofaggressive and antisocial behavior. Journal of ChildPsychology and Psychiatry, 49, 359–375.Friedlander, S., Weiss, D. S., & Traylor, J. (1986). Assessingthe influence of maternal depression on the validity of thechild behavior checklist. Journal of Abnormal Child Psychology,14, 123–133.Friedman, S. L., & Waches, T. D. (1999). Measuringenvironment across the life span: Emerging methods andconcepts. Washington, DC: American PsychologicalAssociation.Friman, P. C., Handwerk, M. L., Smith, G. L., Larzelere,R. E., Lucas, C. P., & Shaffer, D. M. (2000).External validity of conduct and oppositional defiantdisorders determined by the NIMH diagnostic interviewschedule for children. Journal of Abnormal ChildPsychology, 28, 277–286.Fuchs, D. (1987). Examiner familiarity effects on testperformance: Implications for training and practice.Topics in Early Childhood Special Education, 102,90–104.Fuchs, D., & Fuchs, L. S. (1986). Test procedure bias: Ameta-analysis of examiner familiarity effects. Reviewof Educational Research, 56, 243–262.Funderbunk, B. W., & Eyberg, S. M. (1989). Psychometriccharacteristics of the Sutter-Eyberg student behaviorinventory: A school behavior rating scale for use withpreschool children. Behavioral Assessment, 11, 297–313.Gadow, K. D., & Sprafkin, J. (1997). Early ChildhoodInventory-4: Screening manual. Stony Brook, NY:Checkmate Plus.Gadow, K. D., & Sprafkin, J. (1998). Adolescent SymptomsInventory-4 norms manual. Stony Brook, NY:Checkmate Plus.Gadow, K. D., & Sprafkin, J. (2002). Child Symptom Inventory-4screening and norms manual. Stony Brook, NY:Checkmate Plus.Gadow, K. D., Sprafkin, J., Salisbury, H., Schneider, J.,& Loney, J. (2004). Further validity evidence for theteacher version of the child symptom inventory-4.School Psychology Quarterly, 19, 50–71.Gallucci, N. T. (1989). Personality assessment withchildren of superior intelligence: Divergence versuspsychopathology. Journal of Personality Assessment, 53,749–760.Gallucci, N. T. (1997). On the identification of patternsof substance abuse with the MMPI-A scales. PsychologicalAssessment, 9(3), 224–232.Galton, F. (1884). Measurement of character. FortnightlyReview, 42, 179–185. Reprinted in L. D. Goodstein &R. I. Lanyon (Eds.), Readings in Personality Assessment,New York: Wiley.Ganmer, A. B., Graham, J. R., & Archer, R. A. (1992).Usefulness of the MAC scale in differentiating adolescentsin normal, psychiatric, and substance abusesettings. Psychological Assessment, 4, 133–137.Garb, H. N., Wood, J. M., & Nezworski, M. T. (2000).Projective techniques and the detection of child sexualabuse. Child Maltreatment, 5, 161–168.Garber, J. (1984). Classification of childhood psychopathology:A developmental perspective. Child Development,55, 30–48.Garcia, M., & Lega, L. I. (1979). Development of aCuban ethnic identity questionnaire. Hispanic Journalof Behavioral Sciences, 1, 247–261.Geisinger, K. F. (1992). Fairness and selected psychometricissues in the psychological testing of Hispanics.In K. F. Geisinger (Ed.), Psychological testing ofHispanics (pp. 17–42). Washington, DC: AmericanPsychological Association.Gelfand, D. M., & Hartmann, D. P. (1984). ChildBehavior analysis and therapy (2nd ed.). New York:Pergamon Press.Geller, B., Fox, L. W., & Fletcher, M. (1993). Effectof tricyclic antidepressants on switching to mania

eferenes467and on the onset of bipolarity in depressed 6- to12-year-olds. Journal of the American Academy ofChild and Adolescent Psychiatry, 32, 43–50.Gelman, C. F. (1998). Characteristics of a prenatallycocaine-exposed clinical population at school age.Dissertation Abstracts International Section B: The Sciencesand Engineering, 58(11-B), 6234.Ghiselli, E. E., Campbell, J. P., & Zedeck, S. (1981) Measurementtheory for the behavioral sciences. San Francisco:W. H. Freeman.Gibbs, J. T. & Huang, L. N. (1989). Children of color: Psychologicalinterventions with minority youth. San Francisco:Jossey-Bass, Inc.Gilliam, J. E. (1995). Gilliam autism rating scale. Austin,TX: Pro-Ed.Gilliam, J. E. (2001). Gilliam Asperger’s Disorder scale.Austin, Texas: Pro-Ed.Gilliam, J. E. (2008). Gilliam autism rating scale, 2nd edition.Austin, TX: Pro-Ed.Gittelman, R., Mannuzza, S., Shenker, R., & Bonegura,N. (1985). Hyperactive boys almost grown up:I. Psychiatric status. Archives of General Psychiatry, 42,937–947.Gittelman-Klein, R. (1986). Questioning the clinicalusefulness of projective psychological tests forchildren. Developmental and Behavioral Pediatrics, 7,378–382.Glueck, S., & Glueck, E. (1968). Delinquents and nondelinquentsin perspective. Cambridge, MA: HarvardUniversity Press.Glutting, J. J., & Kaplan, D. (1990). Stanford-Binet IntelligenceScale-Fourth Edition: Making the case forreasonable interpretations. In C. R. Reynolds & R. W.Kamphaus (Eds.), Handbook of psychological and educationalassessment of children: Personality, behavior, and context(pp. 277–295). New York: Guilford Press.Goldberg, L. R. (1992). The development of markers forthe big-five factor structure. Psychological Assessment,4, 26–42.Goldsmith, H. H., & Rieser-Danner, L. A. (1990).Assessing early temperament. In C. R. Reynolds andR. W. Kamphaus (Eds.), Handbook of psychological andeducational assessment of children: Personality, behavior,and context (pp. 245–278). New York: The GuilfordPress.Goodman, S. H. & Gotlib, I. H. (1999). Risk for psychopathologyin the children of depressed mothers: Adevelopmental model for understanding mechanismsof transmission. Psychological Review, 106, 458–490.Goodman, W. K., Price, L. H., Rasmussen, S. A.,Mazure, C., Fleischmann, R. L., Hill, C. L., et al.(1989). The Yale–Brown obsessive–compulsive scale:I. development, use, and reliability. Archives of GeneralPsychiatry, 46, 1006–1011.Gordon, M. (1983). The Gordon diagnostic system. DeWitt,NY: Gordon Systems.Gorsuch, R. L. (1988). Factor analysis (3rd ed.). Hillsdale,NJ: Erlbaum.Gottesman, I. I., McGuffin, P., & Farmer, A. E. (1987).Clinical genetics as clues to the “real” genetics ofschizophrenia. Schizophrenia Bulletin, 13, 23–47.Gottlieb, G. (1991). Experiential canalization of behavioraldevelopment: Theory. Developmental Psychology,27(1), 4–13.Green, K., Vosk, B., Forehand, R., & Beck, S. (1981).An examination of differences among sociometricallyidentified accepted, rejected, and neglected children.Child Study Journal, 11, 117–124.Green, M. (1992). Pediatric diagnosis: Interpretation ofsymptoms and signs in infants, children, and adolescents(5th ed.). Philadelphia: Saunders.Green, S. M., Loeber, R., & Lahey, B. B. (1991). Stabilityof mothers’ recall of the age of onset of their child’sattention and hyperactivity problems. Journal of theAmerican Academy of Child and Adolescent Psychiatry,30, 135–137.Greenberg, M. T., Lengua, L. J., Coie, J., Pinderhughes,E., & the Conduct Problems Prevention ResearchGroup. (1999). Predicting developmental outcomesat school entry using a multiple-risk model: FourAmerican communities. Developmental Psychology, 35,403–417.Greenspan, S. (2006). Functional concepts in mentalretardation: Finding the natural essence of an artificialcategory. Exceptionality, 14, 205–224.Greenspoon, P. J., & Saklofske, D. H. (1997). Validityand reliability of the multidimensional students’ lifesatisfaction scale with Canadian children. Journal ofPsychoeducational Assessment, 15, 138–155.Gresham, F. M., & Little, S. G. (1993). Peer-referencedassessment strategies. In T. H. Ollendick & M. Hersen(Eds.), Handbook of child and adolescent assessment (pp.165–179). Boston, MA: Allyn and Bacon.Gresham, R. M., & Elliott, S. N. (1990). Social skillsrating system. Circle Pines, MN: American GuidanceService.Grieten, H., De Haene, L., & Uyttebroek, K. (2007).Cross-cultural validation of the child abuse potentialinventory in Belgium (Flanders): Relations with demographiccharacteristics and parenting problems. Journalof Family Violence, 22, 223–229.Grills, A. E., & Ollendick, T. H. (2003). Multiple informantagreement and the Anxiety disorders interviewschedule for parents and children. Journal of theAmerican Academy of Child & Adolescent Psychiatry, 42,30–40.Gutterman, E. M., O’Brien, J. D., & Young, J. G. (1987).Structured diagnostic interviews for children and adolescents:Current status and future directions. Journalof the American Academy of Child and Adolescent Psychiatry,26, 621–630.Haak, R. A. (1990). Using the sentence completion toassess emotional disturbance. In C. R. Reynolds & R.W. Kamphaus (Eds.), Handbook of psychological andeducational assessment of children: Personality, behavior,and context (pp. 147–167). New York: Guilford Press.

468 referencesHaak, R. A. (2003). Using the sentence completion toassess emotional disturbance. In C. R. Reynolds & R.W. Kamphaus (Eds.), Handbook of psychological andeducational assessment of children: Personality, behavior,and context-2nd edition (pp. 159–181). New York:Guilford.Hammen, C., Rudolph, K., Weisz, J., Rao, U., & Burge,D. (1999). The context of depression in clinic-referredyouth: Neglected areas in treatment. Journal of theAmerican Academy of Child & Adolescent Psychiatry, 38,64–71.Hampson, S., Andrews, J., Barckley, M., & Peterson, M.(2007). Trait stability and continuity in childhood:Relating sociability and hostility to the five-factormodel of personality. Journal of Research in Personality,41, 507–523.Hand, C., Archer, R., Handel, R., & Forbey, J. (2007).The classification accuracy of the Minnesota multiphasicpersonality inventory-adolescent: Effectsof modifying the normative sample. Assessment, 14,80–85.Hankin, B. L., Grant, K. E., Cheeley, C., Wetter,E., Farahmand, F. K., & Westerholm, R. I. (2008).Depressive disorders. In M. Hersen (Series Ed.) & D.Reitman (Vol. Ed.), Handbook of psychological assessment,case conceptualization, and treatment (pp. 199–230). Hoboken, NJ: John Wiley & Sons.Hankin, B. L., Mermelstein, R., & Roesch, L. (2007).Sex differences in adolescent depression: Stress exposureand reactivity. Child Development, 78, 279–295.Harrison, P. L., & Oakland, T. (2003). Adaptive BehaviorAssessment System (2nd ed.). San Antonio, TX: ThePsychological Corporation.Hayes, S. C. (2005). Diagnosing intellectual disabilityin a forensic sample: Gender and age effects on therelationship between cognitive and adaptive functioning.Journal of Intellectual & Developmental Disability,30, 97–103.Hayes, S. C., Nelson, R. O., & Jarrett, R. B. (1987). Thetreatment utility of assessment: A functional approachto evaluating assessment quality. American Psychologist,42, 963–974.Hays, S., & McCallum, R. (2005). A comparison of thepencil-and-paper and computer-administered Minnesotamultiphasic personality inventory-adolescent. Psychologyin the Schools, 42, 605–613.Haynes, S.N., & O’Brien, W.H. (1988). The Gordianknot of DSM-III use: Integrating principles of behaviorclassification and complex causal models. BehavioralAssessment, 10, 95–105.Hawes, D. J. & Dadds, M. R. (2006). Assessing parentingpractices through parent-report and direct observationduring parenting-training. Journal of Child andFamily Studies, 15, 555–568.Holaday, M. (2000). Rorschach protocols from childrenand adolescents diagnosed with posttraumaticstress disorder. Journal of Personality Assessment, 75,143–157.Haley, G. M. T., Fine, S., Marriage, K., Moretti, M. M.,& Freeman, R. J. (1985). Cognitive bias and depressionin psychiatrically disturbed children and adolescents.Journal of Consulting and Clinical Psychology, 53,535–537.Hall, C. W., & Haws, H. D. (1989). Depressive symptomatologyin learning-disabled and nonlearning-disabledstudents. Psychology in the Schools, 26, 359–364.Hamilton, C., Fuchs, D., Fuchs, L. S., & Roberts, H.(2000). Rates of classroom participation and thevalidity of sociometry. School Psychology Review, 29,251–266.Hammen, C., Brennan, P. A., Keenan-Miller, D., & Herr,N. R. (2008). Early onset recurrent subtype of adolescentdepression: Clinical and psychosocial correlates.Journal of Child Psychology and Psychiatry, 49, 443–440.Hammen, C., Gordon, G., Burge, D., Adrian, C.,Jaenicke, C., & Hiroto, G. (1987). Maternal affectivedisorders, illness, and stress: Risk for children’spsychopathology. American Journal of Psychiatry, 144,736–741.Hammer, E. F. (1958). The clinical application of projectivedrawings. Springfield, IL: Thomas.Handfinger, A. S. (2000). The effects of behavior problemson the development of adaptive skills in children. A dissertationsubmitted to the graduate faculty of TheUniversity of Georgia.Handwerk, M. L., Friman, P. C., & Larzelere, R. (2000).Comparing the DISC and the youth self-report. Journalof the American Academy of Child and AdolescentPsychiatry, 39, 807–808.Handwerk, M. L., Larzelere, R. E., Soper, S. H., & Friman,P. C. (1999). Parent and child discrepancies inreporting severity of problem behaviors in three outof-homesettings. Psychological Assessment, 11, 14–23.Harrington, R. G., & Follett, G. M. (1984). The readabilityof child personality assessment instruments.Journal of Psychoeducational Assessment, 2, 37–48.Harrington, R. G., Fudge, H., Rutter, M., Pickles, A., &Hill, J. (1991). Adult outcomes of childhood and adolescentdepression: II. Links with antisocial disorders.Journal of the American Academy of Child and AdolescentPsychiatry, 30, 434–439.Harris, F. C., & Lahey, B. B. (1982a). Recording systembias in direct observational methodology: A reviewand critical analysis of factors causing inaccurate codingbehavior. Clinical Psychology Review, 2, 539–556.Harris, F. C., & Lahey, B. B. (1982b). Subject reactivityin direct observational assessment: A review and criticalanalysis. Clinical Psychology Review, 2, 523–538.Harrison, P. L. (1990). Assessment with the VinelandAdaptive Behavior Scales. In C. R. Reynolds & R. W.Kamphaus (Eds.), Handbook of psychological and educationalassessment of children: Personality, behavior, andcontext. New York: Guilford Press.Harrison, P. L., & Oakland, T. (2000). Adaptive BehaviorAssessment System manual. San Antonio, TX: The PsychologicalCorporation.

eferenes469Hart, D. H. (1972). The Hart sentence completion test forchildren. Salt Lake City: Educational Support Systems.Hart, D. H. (1986).The sentence completion techniques.In H. M. Knoff (Ed.), The assessment of childand adolescent personality (pp. 245–272). New York:Guilford Press.Hart, D. H., Kehle, T. J., & Davies, M. V. (1983). Effectivenessof sentence completion techniques: A reviewof the hart sentence completion test for children.School Psychology Review, 12, 428–434.Hart, E. L., & Lahey, B. B. (1999). General child behaviorrating scales. In D. Shaffer and C. P. Lucas (Eds.),Diagnostic assessment in child and adolescent psychopathology(pp. 65–87). New York: Guilford Press.Hartley, M. M. M. (1999). The relationship among disruptivebehaviors, attention, and academic achievementin a clinic referral sample. Dissertation AbstractsInternational Section A: Humanities and Social Sciences,60(2-A), 0333.Hartmann, D. P., Roper, B. L., & Bradford, D. C. (1979).Some relationships between behavioral and traditionalassessment. Journal of Behavioral Assessment, 1, 3–21.Harvey, V. S. (1989, March). Eschew obfuscation: Supportof clear writing. Communique, p. 12.Hasegawa, C. (1989). The unmentioned minority. In C.J. Maker & L. W. Schiever (Eds.), Defensive programsfor cultural and ethnic minorities (Vol. 2). Austin, TX:Pro-Ed.Haskett, M. E., Ahern, L. S., Ward, C. S., & Allaire, J.C. (2006). Factor structure and validity of the parentingstress index-short form. Journal of Clinical Childand Adolescent Psychology, 35, 302–312.Haskett, M. E., Scott, S. S. & Fann, K. D. (1995).Child abuse potential inventory and parentingbehavior: Relationships with high-risk correlates.Child Abuse & Neglect, 19, 1483–1495.Hathaway, S. R., & McKinley, J. C. (1940). A multiphasicpersonality schedule (Minnesota): I. Constructionof the schedule. Journal of Psychology, 10, 249–254.Hathaway, S. R., & McKinley, J. C. (1942). MinnesotaMultiphasic Personality Inventory. Minneapolis: Universityof Minnesota Press.Hathaway, S. R., & McKinley, J. C. (1989). MMPI-2:Minnesota Multiphasic Personality Inventory – SecondEdition. Minneapolis: University of Minnesota Press.Hayvren, M., & Hymnel, S. (1984). Ethical issues insociometric testing: Impact of sociometric measureson interaction behavior. Developmental Psychology, 20,849–884.Haz, A. M., & Ramirez, V. (1998). Preliminary validationof the child abuse potential inventory in Chile.Child Abuse & Neglect, 22, 869–879.Heber, R. (1959). A manual on terminology and classificationin mental retardation (2nd ed.). (Monographsupplement). American Journal of Mental Deficiency.Heber, R., & Helzer, J. E. (1988). Schizophrenia: Epidemiology.In R. Michels, J. O. Cavenar, A. M. Cooper,S. B. Guze, L. L. Judd, G. L. Merman, & A. J. Solnit(Eds.), Psychiatry. New York: Lippincott.Helms, J. E. (1986). Expanding racial identity theory tocover counseling process. Journal of Counseling Psychology,33, 62–64.Helzer, J.E. & Pryzbeck, T.T. (1988). The co-occurrenceof alcoholism with other psychiatric disorders in thegeneral population and its impact on treatment. Journalof Studies on Alcohol, 49, 2129–224.Helzer, J.E. & Robins, L.N. (1988). The DiagnosticInterview Schedule: Its development, evolution, anduse. Social Psychiatry and Psychiatric Epidemiology,23, 6–16.Hembree-Kigin, T. L., & McNeil, C. B. (1995). Parentchildinteraction therapy. New York: Plenum Press.Hendren, R. L., DeBacker, I., & Pandina, G. J. (2000).Review of neuroimaging studies of child and adolescentpsychiatric disorders from the past 10 years.Journal of the American Academy of Child and AdolescentPsychiatry, 39, 815–828.Henning-Stout, M. (1998). Assessing the behavior ofgirls: What we see and what we miss. Journal of SchoolPsychology, 36, 433–455.Hicks, M. M., Rogers, R., & Cashel, M. (2000). Predictionsof violent and total infractions amonginstitutionalized male juvenile offenders. Journal ofthe American Academy of Psychiatry and the Law, 28,183–190.Hill, A. B., Kemp-Wheeler, S. M. & Jones, S. A. (1986).What does the Beck depression inventory measurein students? Personality and Individual Differences, 7,39–47.Hiller, J. B., Rosenthal, R., Bornstein, R. F., Berry, D.T. R., & Brunell-Neulieb, S. (1999). A comparativemeta-analysis of Rorschach and MMPI validity. PsychologicalAssessment, 11, 278–296.Hinshaw, S. P. (1987). On the distinction between attentionaldeficits/hyperactivity and conduct problems/aggression in child psychopathology. PsychologicalBulletin, 101, 443–463.Hinshaw, S. P. (1991). Stimulant medications and theirtreatment of aggression in children with attention deficits.Journal of Clinical Child Psychology, 20, 301–312.Hinshaw, S. P. (1992). Externalizing behavior problemsand academic underachievement in childhood andadolescence: Causal relationships and underlyingmechanism. Psychological Bulletin, 111, 127–155.Hodges, K. (1994). Child and adolescent functional assessmentscale. Ypsilanti, MI: Eastern Michigan University.Hodges, K., Cool, J., & McKnew, D. (1989). Testretestreliability of a clinical research interview forchildren: The child assessment schedule (CAS). PsychologicalAssessment, 1, 317–322.Hodges, K., Doucette-Gates, A., & Liao, Q. (1999).The relationship between the child and adolescentfunctional assessment scale (CAFAS) and indicatorsof functioning. Journal of Child and Family Studies, 8,109–122.

470 referencesHodges, K., & Kim, C. (2000). Psychometric study ofthe child and adolescent functional assessment scale:Prediction of con-tact with the law and poor schoolattendance. Journal of Abnormal Child Psychology, 28,287–297.Hodges, K., & Saunders, W. (1989). Internal consistencyof a diagnostic interview for children: The childassessment schedule. Journal of Abnormal Child Psychology,17, 691–701.Hodges, K., & Wong, M. M. (1996). Psychometriccharacteristics of a multidimensional measure toassess impairment: The child and adolescent functionalassessment scale. Journal of Child and FamilyStudies, 5, 445–467.Hodges, K., & Zeman, J. (1993). Interviewing. In T. H.Ollendick & M. Hersen (Eds.), Handbook of child andadolescent assessment (pp. 65–81). Boston, MA: Allynand Bacon.Hojnoski, R. L., Morrison, R., Brown, M., & Matthews, W.J. (2006). Projective test use among school psychologists:A survey and critique. Journal of Psychoeducational Assessment,24, 145–159.Holaday, M., Smith, D. A., & Sherry, A. (2000). Sentencecompletion tests: A survey of the literature andresults of a survey of members of the Society for PersonalityAssessment. Journal of Personality Assessment,74, 371–383.Holden, E. W., & Banez, G. A. (1996). Child abusepotential and parenting stress within maltreatingfamilies. Journal of Family Violence, 11, 1–12.Holtzman, W. H., & Swartz, J. D. (2003). Use of theHoltzman Inkblot Technique (HIT) with children. InC. R. Reynolds & R. W. Kamphaus (Eds.), Handbookof psychological and educational assessment in children:Personality, behavior, and context (2nd ed., pp. 198–2218). New York: Guilford.Howlin, P. (2003). Outcome in high-functioningadults with autism with and without early languagedelays: Implications for the differentiation betweenautism and Asperger syndrome. Journal of Autismand Developmental Disorders, 33, 3–14.Hoza, B., Gerders, A. C., Mrug, S., Hinshaw, S. P.,Bukowski, W. M., Gold, J. A., et al. (2005). Peer–assessed outcomes in the multimodal treatmentstudy of children with attention deficit hyperactivitydisorder. Journal of Clinical Child and Adolescent Psychology,34, 74–86.Huberty, C. J., DiStefano, C. & Kamphaus, R. W.(1997). Behavioral clustering of school children. MultivariateBehavioral Research, 32, 105–143.Huber, C. J., Kamphaus, R. W., & DiStefano, C. (1997).Behavioral clustering of school children. MultivariateBehavioral Research, 32, 105–134.Hudziak, J. J., Heath, A. C., Madden, P. F., Reich, W.,Bucholz, K. K., Slutske, W., et al. (1998). Latent classand factor analysis of DSM-IV ADHD: A twin studyof female adolescents. Journal of the American Academyof Child and Adolescent Psychiatry, 37, 848–857.Hudziak, J. J., Wadsworth, B. A., Heath, A. C., &Achenbach, T. M. (1999). Latent class analysis ofchild behavior checklist attention problems. Journal ofthe American Academy of Child and Adolescent Psychiatry,38, 985–991.Huesmann, L. R., Eron, L. D., Lefkowitz, M. M., &Walder, L. O. (1984). Stability of aggression overtime and generations. Developmental Psychology, 20,1120–1134.Hughes, J. (1990). Assessment of social skills: Sociometricand behavioral approaches. In C. R. Reynolds &R. W. Kamphaus (Eds.), Handbook of psychological andeducational assessment of children: Personality, behavior,and context (pp. 423–444). New York: Guilford Press.Hughes, T. L., Gacono, C. B., & Owen, P. F. (2007).Current status of Rorschach assessment: Implicationsfor the school psychologist. Psychology in the Schools,44, 281–291.Hughes, J. N., Meehan, B. T., & Cavell, T. A. (2004).Development and validation of a gender-balancedmeasure of aggression-relevant social cognition.Journal of Clinical Child and Adolescent Psychology, 33,292–302.Hunsley, J. & Bailey, J. M. (2001). Whither the Rorschach?An analysis of the evidence. PsychologicalAssessment, 13, 472–485.Hunsley, J. & DiGiulio, G. (2001). Norms, norming, andclinical assessment. Clinical Psychology: Science and Practice,8, 378–382.Hutcheson, J. J. & Black, M. M. (1996). Psychometricproperties of the Parenting Stress Index in a sampleof low-income African-American mother of infantsand toddlers. Early Education and Development, 7,381–400.Huth-Bocks, A., Kerr, D., Ivey, A., Kramer, A., & King,C. (2007). Assessment of psychiatrically hospitalizedsuicidal adolescents: Self-report instruments as predictorsof suicidal thoughts and behavior. Journal ofthe American Academy of Child & Adolescent Psychiatry,46, 387–395.Hutt, M. L. (1980). The Michigan picture test – revised.New York: Grune & Stratton.Hynd, G. W., Hern, K. L., Novey, E. S., Eliopulos, D.,Marshall, R., Gonzales, J. J., et al. (1993). Attentiondeficithyperactivity disorder and asymmetry of thecaudate nucleus. Journal of Child Neurology, 8, 339–347.Hynd, G. W., & Willis, W. G. (1988). Pediatric neuropsychology.Boston, MA: Allyn and Bacon.Inclan, J. E., & Herron, D. G. (1985). Puerto Ricanadolescents. In G. T. Gibbs & L. N. Huang (Eds.),Children of color (pp. 251–277). San Francisco: Jossey-Bass.Ivanova, M. Y., Achenbach, T. M., Rescorla, L. A.,Dumenci, L., Almqvist, F., Bilenberg, N., et al.(2007). The generalizability of the youth self-reportsyndrome structure in 23 societies. Journal of Consultingand Clinical Psychology, 75, 729–738.

eferenes471Jackson, M. F., Barth, J. M., Powell, N., Lochman, J.E. (2006). Classroom contextual effects of race onchildren’s peer nominations. Child Development, 77,1325–1337.James, E. M., Reynolds, C. R., & Dunbar, J. (1994).Self-report instruments. In T. H. Ollendick & N. J.King, et al. (Eds.), International handbook of phobicand anxiety disorders in children and adolescents: Issuesin clinical child psychology (pp. 317–329). New York:Plenum Press.Janosz, M., Le Blanc, M., Boulerice, B., & Tremblay,R. E. (2000). Predicting different types of schooldropouts: A typological approach with two longitudinalsamples. Journal of Educational Psychology, 92,171–190.Janus, M. D., de Groot, C., & Toepfer, S. M. (1998). TheMMPI-A and 13-year-old inpatients: How young istoo young? Assessment, 5, 321–332.Jensen, P. S. (2003). Comorbidity and child psychopathology:Recommendations for the next decade. Journalof Abnormal Child Psychology, 31, 293–300.Jensen, P. S., Kettle, L., Roper, M. T., Sloan, M. T., Dulcan,M. K., Hoven, C., et al. (1999). Are stimulantsoverprescribed? Treatment of ADHD in four U.S.communities. Journal of the American Academy of Childand Adolescent Psychiatry, 38, 797–804.Jensen, P. S., Rubio-Stipec, M., Canino, G., Bird, H. R.,Dulcan, M. K., Schwab-Stone, M. E., et al. (1999).Parent and child contributions to diagnosis of mentaldisorder: Are both informants always necessary?Journal of the American Academy of Child & AdolescentPsychiatry, 38, 1569–1579.Jensen, P. S., Watanabe, H. E., & Richters, J. E. (1999).Who’s up first? Testing for order effects in structuredinterviews using a counterbalanced experimental design.Journal of Abnormal Child Psychology, 27, 439–446.Jersild, A. T., & Holmes, R. (1935). Children’s fears(Child Development Monograph No. 20). New York:Columbia University Press.Jesness, C. F. (1988). Jesness inventory of adolescent personality.North Tonawanda, NY: Multi-Health Systems.Jiang, X. L. & Cillessen, A. H. N. (2005). Stability ofcontinuous measures of sociometric status: A metaanalysis.Developmental Review, 25, 1–25.Johnson, S. B. (1988). Diabetes mellitus in childhood. InD. Routh (Ed.), Handbook of pediatric psychology (pp.9–31). New York: Guilford Press.Johnston, C. (1996). Parent charecteristics and parentchildinteractions in families of nonproblem childrenand ADHD children with higher and lower levels ofoppositional-defiant behaviour. Journal of AbnormalChild Psychology, 24, 85–104.Johnston, C. & Mash, E. J. (2001). Families of childrenwith attention-deficit/hyperactivity disorder: Reviewand recommendations for future research. Clinical Childand Family Review, 4, 183–207.Johnston, C., Pelham, W. E., & Murphy, H. A. (1985).Peer relationships in ADHD and normal children: Adevelopmental analysis of peer and teacher ratings.Journal of Abnormal Child Psychology, 13, 89–100.Joiner, T. E., Schmidt, K. L., & Barnett, J. (1996). Size,detail, and line heaviness in children’s drawings ascorrelates of emotional distress: (More) negative evidence.Journal of Personality Assessment, 67, 127–141.Joint Committee on Testing Practices (2004). Code ofFair Testing Practices in Education. Washington, DC:American Psychological Association.Jonsen, A. R., Siegler, M., & Winslade, W. J. (1992).Clinical ethics (3rd ed.). New York: McGraw-Hill.Jung, C. J. (1910). The association method. AmericanJournal of Psychology, 21, 219–235. Reprinted in L. D.Goodstein & R. I. Lanyon (Eds.), Readings in personalityassessment. New York: Wiley.Kachigan, S. K. (1982). Multivariate statistical analysis: Aconceptual introduction. New York: Radus Press.Kaczmarek, T., Hagan, M., & Kettler, R. (2006). Screeningfor suicide among juvenile delinquents: Reliabilityand validity evidence for the suicide screening inventory(SSI). International Journal of Offender Therapy and ComparativeCriminology, 50, 204–217.Kagan, J., & Snidman, N. (1991). Temperamental factorsin human development. American Psychologist, 46,856–862.Kamphaus, R. W. (1987). Conceptual and psychometricissues in the assessment of adaptive behavior. Journalof Special Education, 21, 27–36.Kamphaus, R. W. (1998). Intelligence test interpretation:Acting in the absence of evidence. In A. Prifitera & D.Saklofske (Eds.), WISC-III Clinical Use and interpretation:Scientist-practitioner perspectives (pp. 39–57). SanDiego: Academic Press.Kamphaus, R. W. (2001a). Clinical assessment of child andadolescent intelligence. Needham Heights, MA: Allyn& Bacon.Kamphaus, R. W. (2001b). Clinical assessment of children’sintelligence (2nd ed.). Boston, MA: Allyn and Bacon.Kamphaus, R. W., & Frick, P. J. (1996). Clinical assessmentof child and adolescent personality and behavior.Boston: Allyn and Bacon.Kamphaus, R.W., & Campbell, J.C. (2006). Psychodiagnosticassessment of children: dimensional and categoricalmethods. New York, NY: Wiley.Kamphaus, R. W., Huberty, C. J., DiStefano, C., &Petoskey, M. D. (1997). A typology of teacher ratedchild behavior for a national U. S. sample. Journal ofAbnormal Child Psychology, 25, 453–463.Kamphaus, R. W., Jimenez, M. E., Pineda, D. A.,Rowe, E. W., Fleckenstein, L., Restrepo, M. A., etal. (2000). Analisis transcultural de un instrumento dedimensiones multiples en el diagnostico del deficit deatencidn. Revista de Neuropsicologia, Neuropsygiatria yNeurociencias, 2, 51–63.Kamphaus, R. W., Morgan, A. W., Cox, M. R., & Powell,R. M. (1995). Personality and intelligence in thepsychodiagnostic process. In D. H. Saklofske & M.Zeidner (Eds.), International handbook of personality

472 referencesand intelligence (pp. 525–544). New York: PlenumPress.Kamphaus, R. W., Petoskey, M. D., Cody, A. H., Rowe,E. W., Huberty, C. J., & Reynolds, C. R. (1999a). Atypology of parent-rated child behavior for a nationalU. S. sample. Journal of Child Psychology and Psychiatryand Allied Disciplines, 40, 1–10.Kamphaus, R. W., Petoskey, M. D., Cody, A. H., Rowe,E. W., Huberty, C. J., & Reynolds, C. R. (1999b). Atypology of parent-rated child behavior for a nationalU.S. sample. Journal of Child Psychology and Psychiatryand Allied Disciplines, 40, 607–616.Kamphaus, R. W., Petoskey, M. D., & Rowe, E. W.(2000). Current trends in psychological testing ofchildren. Professional Psychology: Research and Practice,31, 155–164.Kamphaus, R. W., & Reynolds, C. R. (1998). BehaviorAssessment System for Children (BASC) ADHD Monitor.Circle Pines, MN: American Guidance Service.Kamphaus, R. W., Reynolds, C. R., & Imperato-McCammon, C. (1998). Roles of diagnosis and classificationin the schools. In T. B. Gutkin & C. R.Reynolds (Eds.), The Handbook of School Psychology (3rded., pp. 292–316). New York: Wiley.Kandel, D. B., & Davies, M. (1988). Adult sequelae ofadolescent depressive symptoms. Archives of GeneralPsychiatry, 43, 255–262.Karver, M. S. (2006). Determinants of multiple informantagreement on child and adolescent behavior.Journal of Abnormal Child Psychology, 34, 251–261.Karver, M. S., Handelsman, J. B., Fields, S., & Bickman,L. (2006). Meta-analysis of therapeutic relationshipvariables in youth and family therapy: The evidencefor different relationship variables in the child andadolescent treatment outcome literature. Clinical PsychologyReview, 26, 50–65.Kashani, J. H., Orvaschel, H., Burk, J. P., & Reid, J. C.(1985). Informant variance: The issue of parent-childdisagreement. Journal of the American Academy of Childand Adolescent Psychiatry, 24, 437–441.Kashner, T. M., Rush, A. J., Surís, A., Biggs, M. M.,Gajewski, V. L.Hooker, D. J., et al. (2003). Impact ofstructured clinical interviews on physicians’ practicesin community mental health settings. Psychiatric Services,54, 712–718.Kaufman, A. S. (1994). Intelligent testing with the WISC-III. New York: WileyKazdin, A. E. (1981). Behavioral observation. In M.Hersen & A. S. Bellack (Eds.), Behavioral assessment:A practical handbook (pp. 59–100). New York: PergamonPress.Kazdin, A. E. (1988). Childhood depression. In E. J.Mash & L. G. Terdal (Eds.), Behavioral assessment ofchildhood disorders (2nd ed., pp. 157–195). New York:Guilford Press.Kazdin, A. E. (1989). Childhood depression. In E. J.Mash & R. A. Barkley (Eds.), Treatment of childhooddisorders (pp. 135–166). New York: Guilford Press.Keenan, K. (2000). Emotional dysregulation as a riskfactor for child psychopathology. Clinical Psychology:Science and Practice, 7, 418–434.Keiley, M. K., Bates, J. E., Dodge, K. A., & Pettit, G.S. (2000). A cross-domain growth analysis: Externalizingand internalizing behaviors during 8 yearsof childhood. Journal of Abnormal Child Psychology,28(2), 161–179.Keith, T. Z. (1990). Confirmatory and hierarchical confirmatoryanalysis of the differential ability scales.Journal of Psychoeducational Assessment, 8, 391–405.Keller, H. R. (1986). Behavioral observation approachesto personality assessment. In H. M. Knoff (Ed.), Theassessment of child and adolescent personality (pp. 353–390). New York: Guilford Press.Kendall, P. C., & Ronan, K. R. (1990). Assessment ofchildren’s anxieties, fears, and phobias: Cognitivebehavioralmodels and methods. In C. R. Reynolds &R. W. Kamphaus (Eds.), Handbook of psychological andeducational assessment of children: Personality, behavior,and context (pp. 223–244). New York: Guilford Press.Kendall, P. C., & Clarkin, J. F. (1992). Introduction tospecial section: Comorbidity and treatment implications.Journal of Consulting and Clinical Psychology, 60,833–834.Keogh, B. K. (1994). A matrix of decision points in themeasurement of learning disabilities. In G. R. Lyon(Ed.), Frames of reference for the assessment of learning disabilities:New views on measurement issues (pp. 15–26).Baltimore, MD: Paul H. Brookes.Kerr, D., Lunkenheimer, E., & Olson, S. (2007). Assessmentof child problem behaviors by multiple informants:A longitudinal study from preschool to schoolentry. Journal of Child Psychology and Psychiatry, 48,967–975.Kim, J. A., Szatmari, P., Bryson, S. E., Streiner, D. L.,& Wilson, F. J. (2000). The prevalence of anxietyand mood problems among children with autism andAsperger syndrome. Autism, 4, 117–132.Kimonis, E. R., Frick, P. J., Skeem, J. L., Marsee, M.A., Cruise, K., Muñoz, L. C., & Aucoin, K. J. (2008).Assessing callous-unemotional traits in adolescentoffenders: Validation of the inventory of callousunemotionaltraits. Journal of the International Associationof Psychiatry and Law. Special Issue: ‘Psychopathyand risk taxation’ in adolescent offenders , 31, 241–251.Klein, D. N., Dougherty, L. R., & Olino, T. M. (2005).Toward guidelines for evidence-based assessmentof depression in children and adolescents. Journal ofClinical Child and Adolescent Psychology, 34, 412–432.Klein, D. N., Lewinsohn, P. M., & Seeley, J. R. (1997).Psychosocial characteristics of adolescents with a pasthistory of dysthymic disorder: Comparison with adolescentswith past histories of major depressive andnon-affective disorders, and never mentally ill controls.Journal of Affective Disorders, 42, 127–135.Kleinmmntz, B. (1967). Personality measurement: Anintroduction. Homewood, IL: Dorsey Press.

eferenes473Kline, R. B. (1989). Is the Fourth Edition Stanford-Bineta four-factor test? Confirmatory analysis of alternativemodels for ages 2 through 23. Journal of PsychoeducationalAssessment, 7, 913.Kline, R. B. (1994). Test review: New objective ratingscales for child assessment, I. Parent- and teacherinformantinventories of the behavior assessmentsystem for children, the child behavior checklist, andthe teacher report form. Journal of PsychoeducationalAssessment, 12, 289–306.Kline, R. B. (1995). New objective rating scales for childassessment, II. Self-report scales for children and adolescents:Self-report of personality of the behaviorassessment system for children, the youth self-report,and the personality inventory for youth. Journal of PsychoeducationalAssessment, 13, 169–193.Knight, L. W. (1997). The BASC parent rating scales:Patterns of behavior by gender, race, and socioeconomicstatus. Dissertation Abstracts International SectionA: Humanities and Social Sciences, 58(1-A), 0084.Knoff, H. M. (1983). Justifying projective/personalityassessment in school psychology: A response toBatsch and Peterson. School Psychology Review, 12,446–451.Kochanska, G. (1993). Toward a synthesis of parentalsocialization and child temperament in early developmentof con-science. Child Development, 64, 325–347.Kochanska, G., Kuczynski, L., & Radke-Yarrow, N.I. (1989). Correspondence between mothers’ selfreportedand observed child-rearing practices. ChildDevelopment, 60, 56–63.Konold, T. R., Walthall, J. C., & Pianta, R. C. (2004).The behavior of child ratings: Measurement structureof the child behavior checklist across time, informants,and child gender. Behavioral Disorders, 29, 372–383.Koocher, G. P. (1993). Ethical issues in the psychologicalassessment of children. In T. H. Ollendick & M.Hersen (Eds.), Handbook of child and adolescent assessment(pp. 51–61). Boston, MA: Allyn and Bacon.Koppitz, E. M. (1968). Psychological evaluation of children’shuman figure drawings. New York: Grune atStratton.Koppitz, E. M. (1983). Projective drawings with children,and adolescents. School Psychology Review, 12,421–427.Kovacs, M. (1989). Affective disorders in children andadolescents. American Psychologist, 44(2), 209–215.Kovacs, M. (1992). The Children’s Depression Inventorymanual. New York: Multi-Health Systems.Kovacs, M. (2001). Gender and the course of majordepressive disorder through adolescence in clinicallyreferredyoungsters. Journal of the American Academy ofChild and Adolescent Psychiatry, 40, 1079–1085.Kovacs, M., Paulauskas, S., Gatsonis, C., & Richards,C. (1988). Depressive disorders in childhood: III. Alongitudinal study of comorbidity with and risk forconduct disorders. Journal of Affective Disorders, 15,205–217.Krefetz, D., Steer, R., Gulab, N., & Beck, A. (2002).Convergent validity of the Beck depression inventory-IIwith the Reynolds adolescent depression scalein psychiatric inpatients. Journal of Personality Assessment,78, 451–460.Kroon, N., Goudena, P. P., & Rispens, J. (1998). Thematicapperception tests for child and adolescentassessment: A practitioner’s consumer guide. Journalof Psychoeducational Assessment, 16, 99–117.Krug, D. A., & Arick, J. R. (2003). Krug Asperger’s disorderindex. Austin, Texas: Pro-Ed.Kuehne, C., Kehle, T. J., & McMahon, W. (1987).Differences between children with attention deficitdisorder, children with specific learning disabilities,and normal children. Journal of School Psychology, 25,161–166.Kuramoto, H., Kanbayaski, Y., Nakata, Y., Fukui, T.,Mukai, T., & Negishi, Y. (1999). Standardization ofthe Japanese version of the youth self report (YSR).Japanese Journal of Child and Adolescent Psychiatry, 40,329–344.Lachar, D. (1982). Personality inventory for childrenrevised(PICR). Los Angeles: Western PsychologicalServices.Lachar, D. (1990). Objective assessment of child andadolescent personality: The Personality Inventoryfor Children. In C. R. Reynolds & R. W. Kantphaus(Eds.), Handbook of psychological and educationalassessment of children: Personality, behavior, andcontext (pp. 298–323). New York: Guilford Press.Lachar, D. (1993). Symptom checklists and personalityinventories. In T. R. Kratochwill & R. J. Morris(Eds.), Handbook of psychotherapy for children and adolescents(pp. 38–57). Boston, MA: Allyn and Bacon.Lachar, D. (1999). Personality inventory for Children,Second Edition (PTC-2), Personality Inventory forYouth (PM, and Student Behavior Survey (SBS). InM. Maiuish (Ed.), The use of psychological testing fortreatment planning and outcomes assessment (2nd ed.).Mahwah, NJ: Lawrence Erlbaum.Lachar, D., & Gruber, C. P. (1992). Personality Inventoryfor Youth (P1Y). Los Angeles: Western PsychologicalServices.Lachar, D., & Gruber, C. P. (1993). Development of thepersonality inventory for youth: A self-report companionto the personality inventory for children. Journalof Personality Assessment, 61, 81–98.Lachar, D., & Gruber, C. P. (1994). The PersonalityInventory for Youth. Los Angeles: Western PsychologicalServices.Lachar, D., & Gruber, C. P. (1997). Personality Inventoryfor Youth (PIP) manual: Administration and interpretationguide. Los Angeles: Western PsychologicalServices.Lachar, D. & Gruber, C. (2001). Personality Inventoryfor Children, second edition (PIC-2): Standard form andbehavioral summary manual. Los Angeles: WesternPsychological Services.

474 referencesLachar, D., Harper, R. A., Green, B. A., Morgan, S.T., & Wheeler, A. C. (1996, August). The PersonalityInventory fir Youth: Contribution to diagnosis. Posterpresented at the 104th Annual Convention of theAmerican Psychological Association, Toronto, Canada.Lachar, D., Morgan, S. T., Espades, A., & Schomer, O.(2000, August). Effect of defensiveness on two self-reportchild adjustment inventories. Poster presented at the108th Annual Convention of the American PsychologicalAssociation, Washington, DC.Lachar, D., Wingenfeld, S. A., Kline, R. B., & Gruber,C. P. (2000). Student behavior survey. Los Angeles,CA: Western Psychological Services.LaCombe, J. A., Kline, R. B., Lachar, D., Butkus, M., & Hillman, S. B.(1991). Case history correlates of a personality inventoryfor children (PIC) profile typology. Psychological Assessment,3, 678–687.Lahey, B. B., Applegate, B., Waldman, I. D., Loft, J.D., Hankin, B. L., & Rick, J. (2004). The structureof child and adolescent psychopathology: Generatingnew hypotheses. Journal of Abnormal Psychology, 113,358–385.Lahey, B. B., Applegate, B., Barkley, R. A., Garfinkel, B.,McBurnett, K., Kerdyck, L., et al. (1994). DSM-IVfield trials for oppositional defiant disorder and conductdisorder in children and adolescents. American Journal ofPsychiatry, 151, 1163–1171.Lahey, B. B., Applegate, B., McBurnett, K., Biederman,J., Greenhill, L., Hynd, G. W., et al. (1994). DSM-IVfield trials for attention-deficit hyperactivity disorderin children and adolescents. American Journal of Psychiatry,151, 1673–1685.Lahey, B. B., Carlson, C. L. & Frick, P. J. (1997). Attention-deficitdisorders: A review of research relevant todiagnostic classification. In T. A. Widiger, A. J. Frances,H. A. Pincus, R. Ross, M. B. First, & W. Davis (Eds.),DSM-IV sourcebook (Vol. 3, pp. 163–188). Washington,DC: American Psychiatric Association.Lahey, B.B., Goodman, S.H., Waldman, I.D., Bird, H.,Canino, G., Jenson, P., Regier, D., Leaf, P.J., Gordon,R., & Applegate, B. (1999). Relation of age of onsetto the type of severity of child and adolescent conductproblems. Journal of Abnormal Child Psychology, 27,247–260.Lahey, B. B., Hart, E. L., Pliszka, S., Applegate, B.,& McBurnett, K. (1993). Neurophysiological correlatesof conduct disorder: A rationale and a reviewof research. Journal of Clinical Child Psychology, 22,141–153.Lahey, B. B., & Loeber, R. (1994). Framework for adevelopmental model of oppositional defiant disorderand conduct disorder. In D. K. Routh (Ed.), Disruptivebehavior disorders in childhood (pp. 139–180). NewYork: Plenum.Lahey, B. B., Loeber, R., Quay, H. C., Applegate, B.,Shaffer, D., Waldman, I., et al. (1998). Validity ofDSM-IV subtypes of conduct disorder based on ageof onset. Journal of the American Academy of Child andAdolescent Psychiatry, 37, 435–442.Lahey, B. B., Loeber, R., Quay, H. C., Frick, P. J., &Grimm, J. (1992). Oppositional defiant disorder andconduct disorders: Issues to be resolved for DSM-IV.Journal of the American Academy of Child and AdolescentPsychiatry, 31, 539–546.Lahey, B. B., Loeber, R., Stouthamer-Loeber, M.,Christ, M. A. G., Green, S. M., Russo, M. F., et al.(1990). Comparison of DSM-III and DSM-III-Rdiagnoses for prepubertal children: Changes in prevalenceand validity. Journal of the American Academy ofChild and Adolescent Psychiatry, 29, 620–626.Lahey, B. B., Pelham, W. E., Loney, J., Lee, S. S., &Willcutt, E. (2005). Instability of the DSM-IV subtypesof ADHD from preschool through elementaryschool. Archives of General Psychiatry, 62, 896–902.Lahey, B. B., Pelham, W. E., Schaughency, E. A.,Atkins, M. S., Murphy, A., Hynd, G. W., et al. (1988).Dimensions and types of attention deficit disorder.Journal of the American Academy of Child and AdolescentPsychiatry, 27, 330–335.Lahey, B. B., & Piacentini, J. C. (1985). An evaluation ofthe Quay-Peterson revised behavior problem checklist.Journal of School Psychology, 23, 285–289.Lambert, M., Whipple, J., Hawkins, E., Vermeersch, D.,Nielsen, S., & Smart, D. (2003). Is it time for cliniciansto routinely track patient outcome? A meta-analysis.Clinical Psychology: Science and Practice, 10, 288–301.Lambert, N. M. (1988). Adolescent outcomes for hyperactivechildren: Perspectives on general and specificpatterns of child-hood risk for adolescent educational,social and mental health problems. American Psychologist,43, 786–799.Lambert, N., Nihira, K., & Leland, H. (1993). AAMRadaptive behavior scale-school (2nd ed.)(ABS-S:2).Austin, TX: Pro-Ed.Lanyou, R. I. (1995). Review of the Minnesota MultiphasicPersonality Inventory – Adolescent. ‘TwelfthMental Measurements Yearbook. Lincoln, NE: BurosInstitute.Last, C. G. (1993). Introduction. In C. G. Last (Ed.),Anxiety across the lifespan: A developmental perspective(pp. 1–6). New York: Springer.Last, C. G., Hersen, M., Kazdin, A. E., Francis, G., &Grubb, H. J. (1987). Psychiatric illness in the mothersof anxious children. American Journal of Psychiatry,144, 1580–1583.Le Couteur, A., Rutter, M., Lord, C., Rios, P., Robertson,S., Holdgrafer, M., et al. (1989). Autism diagnosticinterview: A standardized investigator-basedinstrument. Journal of Autism and Developmental Disorders,19, 363–387.Lefkowitz, M. M., & Tesiny, E. P. (1980). Assessmentof child-hood depression. Journal of Consulting andClinical Psychology, 48, 43–50.Lemery, K., Essex, M., & Smider, N. (2002). Revealingthe relation between temperament and behavior

eferenes475problem symptoms by eliminating measurement confounding:Expert ratings and factor analyses. ChildDevelopment, 73, 867–882.Lett, N. J. (1997). Differential validity of the BASC studentobservation system and the BASC teacher ratingscale. Canadian Journal of School Psychology, 13, 1–14.Lett, N. J., & Kamphaus, R. W. (1997). Differentialvalidity of the BASC student observation systemand the BASC teacher rating scales. Canadian Journalof School Psychology, 13, 1–14.Lewczyk, C. M., Garland, A. F., Hurlburt, M. S., Gearity,J., & Hough, R. L. (2003). Comparing DISC-IVand clinician diagnoses among youths receiving publicmental health services. Journal of the American Academyof Child & Adolescent Psychiatry, 42, 349–356.Lichtenberger, E. O., Mather, N., Kaufman, N. L., &Kaufman, A. S. (2004). Essentials of assessment reportwriting. Hoboken, NJ: Wiley.Lilienfeld, S. O. (2003). Comorbidity between andwithin childhood externalizing and internalizing disorders:Reflections and directions. Journal of AbnormalChild Psychology, 31, 285–291.Lochman, J. E., Wells, K. C., & Lenhart, L. A. (2008).Coping power: Child group program. New York:Oxford.Locke, H. J., & Wallace, K. M. (1959). Short maritaladjustmentand prediction tests: Their reliability andvalidity. Marriage and Family Living, 21, 251–255.Loeber, R. (1991). Antisocial behavior: More enduringthan changeable? Journal of the American Academy ofChild and Adolescent Psychiatry, 30, 393–397.Loeber, R., Burke, J. D., Lahey, B. B., Winters, A., &Zen, M. (2000). Oppositional defiant and conductdisorder: A review of the past 10 years, Part I. Journalof the American Academy of Child and Adolescent Psychiatry,39, 1468–1484.Loeber, R., Green, S. M., & Lahey, B. B. (1990). Mentalhealth professionals’ perception of the utility ofchildren, mothers, and teachers as informants onchildhood psychopathology. Journal of Clinical ChildPsychology, 19, 136–143.Loeber, R., Green, S. M., Lahey, B. B., & Stouthamer-Loeber, M. (1991). Differences and similaritiesbetween children, mothers, and teachers as informantson disruptive child behavior. Journal of Abnormal ChildPsychology, 19, 75–95.Loeber, R., Green, S. M., Lahey, B. B., Christ, M. A. G.,& Frick, P. J. (1992). Developmental sequences in theage of onset of disruptive child behaviors. Journal ofChild and Family Studies, 1, 21–41.Loeber, R., Green, S. M., Lahey, B. B., Christ, M. A. G.,& Frick, P. J. (1992). Developmental sequences in theage of onset of disruptive child behaviors. Journal ofChild and Family Studies, 1, 21–41.Loeber, R., & Keenan, K. (1994). Interaction betweenconduct disorder and its comorbid conditions: Effectsof age and gender. Clinical Psychology Review, 14,497–523.Loeber, R., & Stouthamer-Loeber, M. (1986). Familyfactors as correlates and predictors of juvenileconduct problems and delinquency. In M. Tonry& N. Morris (Eds.), Crime and justice (Vol. 7, pp.29–149). Chicago: University of Chicago Press.Loney, B. R., & Frick, P. J. (2003). Structured diagnosticinterviewing. In C. R. Reynolds & R. W. Kamphaus(Eds.), Handbook of educational assessment of children(2nd ed., pp. 235–247). New York: Guilford.Loney, B. R., Frick, P. J., Ellis, M. L., & McCoy, M.G. (1998). Intelligence, psychopathy, and antisocialbehavior. Journal of Psychopathology and BehavioralAssessment, 20, 231–247.Longman, Inc. (1984). Longman dictionary of psychologyand psychiatry. New York: Author.Lopez, S. R. (1999). Teaching culturally informed psychologicalassessment. In R. H. Dana (Ed.), Handbookof cross-cultural/multicultural personality assessment (pp.669–687). Hillsdale, NJ: Erlbaum.Lord, C., Rutter, M., DiLavore, P. C., & Risi, S. (1999).Autism diagnostic observation schedule-WPS (ADOS-WPS). Los Angeles: Western Psychological Services.Lord, C., Risi, S., Lambrecht, L., Cook, E. H., Jr., Leventhal,B. L., DiLavore, P. C., et al. (2000). The autismdiagnostic observation schedule-generic: A standardmeasure of social and communication deficits associatedwith the spectrum of autism. Journal of Autism andDevelopmental Disorders, 30(3), 205–223.Lord, C., Rutter, M., & Le Couteur, A. (1994). Autismdiagnostic interview-revised: A revised version of adiagnostic inter-view for caregivers of individuals withpossible pervasive developmental disorders. Journal ofAutism and Developmental Disorders, 24(5), 659–685.Lord, C., Rutter, M., DiLavore, P. C., & Risi, S. (2002).Autism Diagnostic Observation Schedule manual. LosAngeles: Western Psychological Services.Lord, C., Storoschuk, S., Rutter, M., & Pickles, A.(1993). Using the ADI-R to diagnose autism inpreschool children. Infant Mental Health Journal, 14,234–252.Lorr, M. (1965). Review of the Wittenborn PsychiatricRating Scales. In O. K. Buros (Ed.), Mental measurementsyearbook (5th ed.). Highland Park, NJ: GryphonPress.Lou, H. C., Henriksen, L., Bruhn, P., Borner, H., &Nielsen, J. B. (1989). Striatal dysfunction in attentiondeficit and hyperkinetic disorder. Archives of Neurology,46, 48–52.Love, A. S., & McIntosh, D. E. (March, 1994). Differentialdiagnosis of attention deficit hyperactivity disorderedand emotionally/behaviorally disordered children using theTemperament Assessment Battery for children. Paper presentedat the National Association of School Psychologyannual meeting, Seattle, WA.Lowman, M. G., Schwanz, K. A., & Kamphaus, R.W. (1996). WISC III third factor: Critical measurementissues. Canadian Journal of School Psychology, 12,15–22.

476 referencesLynam, D. R., Caspi, A., Moffitt, T. E., Raine, A., Loeber,R. & Stouthamer-Loeber, M. (2005). Adolescentpsychopathy and the big five: Results from two samples.Journal of Abnormal Psychology, 33, 431–443.Luby, J. L., Heffelfinger, A., Mrakotsky, C., Hessler, M.J., Brown, K. M., & Hildebrand, T. (2002). Preschoolmajor depressive disorder: Preliminary validation fordevelopmentally modified DSM-IV criteria. Journal ofthe American Academy of Child & Adolescent Psychiatry,41, 928–937.Lucia, V. C. & Breslau, N. (2006). Family cohesion andchildren’s behavior problems: A longitudinal investigation.Psychiatry Research, 141, 141–149.Luckasson, R., Borthwick-Duffy, S., Buntinx, W. H. E.,Coulter, D. L., Craig, E. M., Reeve, A., et al. (2002).Mental retardation: Definition, classification, and systemsof supports (10th ed.). Washington, DC: AmericanAssociation on Mental Retardation.Lyneham, H. J., Abbott, M. J., & Rapee, R. M. (2007).Interrater reliability of the anxiety disorders interviewschedule for DSM-IV: Child and parent version.Journal of the American Academy of Child & AdolescentPsychiatry, 46, 731–736.Lynskey, M. T., & Fergusson, D. M. (1995). Childhoodconduct problems, attention deficit behaviors, andadolescent alcohol, tobacco, and illicit drug use. Journalof Abnormal Child Psychology, 23, 281–302.Lytton, H. (1990). Child and parent effects in boys’ conductdisorder: A reinterpretation. Developmental Psychology,26, 683–697.Maccoby, E. E., & Martini, A. (1983). Socialization inthe context of the family: Parent-child interaction. InP. H. Mussen & E. M. Hetherington (Eds.), Handbookof child psychology: IV; Socialization, personality,and social development (pp. 1–101). New York: Wiley.Machover, K. (1949). Personality projection in the drawingof the human figure. Springfield, IL: Charles C.Thomas.Madsen, C. J. (1990, September). Attention deficit disordersand application of the EFL4 and Section 504. Paperpresented to the Seventh Annual Pacific NorthwestInstitute on Special Education and Law.Malgady, R. G. & Costantino, G. (1998). Symptomseverity in bilingual Hispanics as a function of clinicianethnicity and language of interview. PsychologicalAssessment, 10, 120–127.Mann, B. J., & MacKenzie, E. P. (1996). Pathwaysamong marital functioning, parental behaviors, andchild behavior problems in school-age boys. Journal ofClinical Child Psychology, 25, 183–191.Mannuzza, S., Gittelman-Klein, R., Konig, P. H., &Giampino, T. L. (1989). Hyperactive boys almostgrown up: IV. Criminality and its relationship topsychiatric status. Archives of General Psychiatry, 46,1073–1079.March, J. S., Parker, J. D. A., Sullivan, K., Stallings, P., &Conners, K. (1997). The Multidimensional AnxietyScale for Children (MASC): Factor, structure, reliability,and validity. Journal of the American Academy ofChild & Adolescent Psychiatry, 36, 554–565.Marchant, G. J., & Ridenour, T. A. (1998). [Review ofthe Personality Inventory for Youth]. Thirteenth MentalMeasurements Yearbook (pp. 757–758). Lincoln,NE: Burns Institute.Marcotte, D., Fortin, L., Potivin, P., & Papillon, M.(2002). Gender differences in depressive symptomsduring adolescence: Role of gender-typed characteristics,self-esteem, body image, stressful life events, andpubertal status. Journal of Emotional and BehavioralDisorders, 10, 29–42.Marcus, L. M., Lansing, M. D., & Schopler, E. (1993).Assessment of the autistic and pervasive developmentallydisordered child. In D. J. Willis & J. L. Culbertson(Eds.), Testing young children (pp. 319–344).Austin, TX: Pro-Ed.Maths, G. (1992). Issues in the measurement of acculturationamong Hispanics. In K. F. Geisinger (Ed.),Psychological testing of Hispanics (pp. 235–252). Washington,DC: American Psychological Association.Marin, G. (1992). Issues in the measurement of acculturationamong Hispanics. In K. F. Geisinger (Ed.),Psychological testing of Hispanics (pp. 235–252). Washington,DC: American Psychological Association.Marin, G., Sabogal, F., Marin, V. B., Otero-Sabogal, R.,& Perez-Stable, E. J. (1987). Development of a shortacculturation scale for Hispanics. Hispanic Journal ofBehavioral Sciences, 9, 183–205.Marin, G., Sabogal, F., VanOss Marin, B., Otero-Sabogal, R.,& Perez-Stable, E. J. (1987). Development of a short acculturationscale for Hispanics. Hispanic Journal of BehavioralScience, 9, 183–205.Marsee, M. A., & Frick, P. J. (2007). Exploring the cognitiveand emotional correlates to proactive and reactiveaggression in a sample of detained girls. Journal ofAbnormal Child Psychology, 35, 969–981.Martin, R. P. (1983). The ethical issues in the use andinterpretation of the Draw-a-Person Test and othersimilar projective procedures. The School Psychologist,38(6), 8.Martin, R. P. (1988). Assessment of personality and behaviorproblems: Infancy through adolescence. New York:Guilford Press.Martin, R. P. (1989). Activity level, distractibility, andpersistence: Critical characteristics in early schooling.In G. A. Kohnstamm, J. E. Bates, & M. K. Rothbart(Eds.), Temperament in childhood (pp. 451–462) London:Wiley.Martin, R. P., & Bridger, R. (1999). Temperament AssessmentBattery for Children – Revised (TABC-R). Athens,GA: Authors.Martin, R. P., & Holbrook, J. (1985). Relationshipof temperament characteristics to the academicachievement of first grade children. Journal of PsychoeducationalAssessment, 3, 131–140.Martin, R. P., Nagle, R., & Paget, K. (1984). Relationshipsbetween temperament and classroom behavior,

eferenes477teacher attitudes and academic achievement. Journalof Psychoeducational Assessment, 1, 377–386.Martin, R. P., Wisenbaker, J., Matthews-Morgan, J.,Holbrook, J., Hooper, S., & Spalding, J. (1986). Stabilityof teacher temperament ratings over 6 and 12months. Journal of Abnormal Child Psychology, 14,167–179.Martinez, R., Norman, R. D., & Delaney, H. D. (1984).A children’s Hispanic background scale. HispanicJournal of Behavioral Sciences, 6, 103–112.Mash, E. J., & Dozois, D. J. A. (2003). Child psychopathology:A developmental-systems perspective. In E.J. Mash & R. A. Barkley (Eds.), Child psychopathology(2nd ed., pp. 3–71). New York: Guilford.Mash, E., & Hunsley, J. (2005). Evidence-based assessmentof child and adolescent disorders: Issues andchallenges. Journal of Clinical Child and Adolescent Psychology,34, 362–379.Mash, E. J., & Terdal, L. G. (1988). Behavioral assessmentof child and family disturbance. In E. J. Mash& L. G. Terdal (Eds.), Behavioral assessment of childhooddisorders (2nd ed., pp. 3–68). New York: GuilfordPress.Massetti, G. M., Lahey, B. B., Pelham, W. E., Loney,J., Ehrhardt, A., Lee, S. S., et al. (2008). Academicachievement over 8years among children who metmodified criteria for attention-deficit/hyperactivitydisorder at 4–6 years of age. Journal of Abnormal ChildPsychology, 36, 399–410.Masuda, M., Matsumoto, G. H., & Meredith, G. M.(1970). Ethnic identity in three generations of JapaneseAmericans. Journal of Social Psychology, 81, 199–207.Matazow, G. S., & Kamphaus, R. N. T. (in press). Behaviorassessment system for children (BASC): Towardaccurate diagnosis and effective treatment. In J. J. W.Andrews, H. Janzen, & D. Saklofske (Eds.), Ability,achievement, and behavior assessment: A practical handbook.San Diego, CA: Academic Press.Mattison, R. E., Handford, H. A., Kales, H. C., Goodman,A. L., & McLaughlin, R. E. (1990). Four-yearpredictive value of the Children’s Depression Inventory.Psychological Assessment, 2, 169–174.Mayes, L., Volkmar, F. R., Hooks, M., & Cicchetti, D.V. (1993). Differentiating pervasive developmentaldisorder not otherwise specified from autism and languagedisorders. Journal of Autism & DevelopmentalDisorders, 23, 79–90.Mayfield, J. W. & Reynolds, C. R. (1998). Are ethnicdifferences in diagnosis of childhood psychopathologyan artifact of psychometric methods? An experimentalevaluation of Harrington’s hypothesis usingparent-reported symptomatology. Journal of SchoolPsychology, 36, 313–334.Mayeux, L., Bellmore, A. D., & Cillessen, A. H. N.(2007). Predicting changes in adjustment usingrepeated measures of sociometric status. The Journalof Genetic Psychology, 168, 401–424.Mayeux, L., Underwood, M. K., & Risser, S. D. (2007).Perspectives on the ethics of sociometric research withchildren. Merrill-Palmer Quarterly, 53, 53–78.Mazefsky, C., & Oswald, D. (2006). The discriminativeability and diagnostic utility of the ADOS-G,ADI-R, and GARS for children in a clinical setting.Autism, 10, 533–549.McArthur, D. S., & Roberts, G. E. (1982). Roberts ApperceptionTest for children. Los Angeles: Western PsychologicalServices.McCarney, S. B. (2004). The attention deficit disordersevaluation scale, home and school versions: Technicalmanual. Columbia, MO: Hawthorne.McCarthy, L., & Archer, R. P. (1998). Factor structureof the MMPI-A content scales: Item-level and scalelevelfindings. Journal of Personality Assessment, 71,84–97.McConaughy, S. H. (2005). Direct observational assessmentduring test sessions. School Psychology Review,34, 490–506.McConaughy, S. H., & Achenbach, T. M. (1989).Empirically based assessment of serious emotionaldisturbance. Journal of School Psychology, 27, 91–117.McConaughy, S. H., & Achenbach, T. M. (2004). Manualfor the Test Observation Form for ages 2–18. Burlington,VT: Research Center for Children, Youth, andFamilies.McConaughy, S. H., Achenbach, T. M., & Gent, C. L.(1988). Multiaxial empirically based assessment: Parent,teacher, observational, cognitive, and personalitycorrelates of child behavior profile types for 6- to11-year-old boys. Journal of Abnormal Child Psychology,16, 485–509.McConaughy, S. H., Stanger, C., & Achenbach, T. M.(1992). Three year course of behavioral emotionalproblems in a national sample of 4- to 16-year-olds:I. Agreement among informants. Journal of the AmericanAcademy of Child and Adolescent Psychiatry, 31,932–940.McConell, S. R., & Odom, S. L. (1986). Sociometrics:Peer-referenced measures and the assessment of socialcompetence. In P. S. Strain, M. J. Guralnick, & H. M.Walker (Eds.), Children’s social behavior: Development,assessment, and modification (pp. 215–286). New York:Academic Press.McCrae, R. R., Stone, S. V., Fagan, P. J., Costa, P. T.(1998). Identifying causes and disagreement betweenself-reports and spouse ratings of personality. Journalof Personality, 66, 285–312.McDonald, K. L., Putallaz, M., Grimes, C. L.,Kupersmidt, J. B., & Coie, J. D. (2007). Girl talk:Gossip, friendship, and sociometric status. Merrill-Palmer Quarterly, 53, 381–411.McGoldrick, M., Gerson, R., & Shellenberger, S.(2008). Genograms: Assessment and intervention (2nded.). New York: W. W. Norton.McGrew, K. S., Bruininks, R. H., & Thurlow, M. L.(1992). Relationship between measures of adaptive

478 referencesfunctioning and community adjustment for adultswith mental retardation. Exceptional Children, 58,517–529.McMahon, R. J., & Frick, P. J. (2005). Evidence-basedassessment of conduct problems in children and adolescents.Journal of Clinical Child and Adolescent Psychology,34, 477–505.McMahon, R. J. & Frick, P. J. (2007). Conduct andoppositional disorders. In E. J. Mash & R. A. Barkley(Eds)., Assessment of childhood disorders (4th ed.,pp. 132–183). New York: Guilford.McMahon, R. J., & Peters, R. D. V. (2002). The effectsof parental dysfunction on children. New York: PlenumPress.McNally, S., Eisenberg, N., & Harris, J. D. (1991).Consistency and change in maternal child-rearingpractices and values: A longitudinal study. ChildDevelopment, 62, 190–198.McNamara, J. R., Hollmann, C., & Riegel, T. (1994).A preliminary study of the usefulness of theBehavior Assessment System for Children in theevaluation of mental health needs in a Head Startpopulation. Psychological Reports, 75, 1195–1201.McNeil, C. B., Eyberg, S., Eisenstadt, T. H., Newcomb,K., & Funderbunk, B. (1991). Parent-child InteractionTherapy with behavior problem children: Generalizationof treatment effects to the school setting.Journal of Clinical Child Psychology, 20, 140–151.McReynolds, P., Ballachey, E.L., & Ferguson, J.T.Development and evaluation of a behavioral scale forappraising the adjustment of hospitalized patients.American Psychologists, 1952, 7, 340Measelle, J. R., Ablow, J. C., Cowan, P. A., & Cowan,C. P. (1998). Assessing young children’s views of theiracademic, social, and emotional lives: An evaluationof the self-perception scales of the Berkeley PuppetInterview. Child Development, 69, 1556–1576.Meehl, P. E. (1973). Psychodiagnosis. Selected papers. Minneapolis:University of Minnesota Press.Meehl, P. E. & Hathaway, S. R. (1946). The K factoras a suppressor variable in the Minnesota MultiphasicPersonality Inventory. Journal of Applied Psychology,30, 525–564.Mendoza, R. H. (1989). An empirical scale to measuretype and degree of acculturation in Mexican-Americanadolescents and adults. Journal of Cross CulturalPsychology, 20, 372–385.Merenda, P. F. (1996). BASC: Behavior AssessmentSystem for Children. Measurement and Evaluation inCounseling and Development, 28, 229–232.Merrell, K. W., & Popinga, M. R. (1994). The alliance ofadaptive behavior and social competence: An examinationof relationships between the Scales of IndependentBehavior and the Social Skills Rating System.Research in Developmental Disabilities, 15, 39–47.Merydith, S. P. (2000). Aggression Intervention SkillTraining: Moral reasoning and moral emotions.NASP Communique, 28, 6–8.Merydith, S. P., & Joyce, F. K. (1998, August). Temporalstability and convergent validity of the BASC Parent andTeacher Rating Scales. Paper presented at the AnnualMeeting of the American Psychological Association,San Francisco.Meyer, G. J., Erdberg, P., & Shaffer, T. W. (2007). Towardinternational normative reference data for the ComprehensiveSystem. Journal of Personality Assessment,89(Suppl. 1), S201–S216.Mikami, A. Y. & Hinshaw, S. P. (2003). Buffers of peerrejection among girls with and without ADHD: Therole of popularity with adults and goal-directed solitaryplay. Journal of Abnormal Child Psychology, 31, 381–397.Mildenberger, K., Sitter, S., Noterdaeme, M., & Amorosa,H. (2001). The use of the ADI–R as a diagnostictool in the differential diagnosis of children withinfantile autism and children with receptive languagedisorder. European Child and Adolescent Psychiatry, 10,248–255.Milich, R. (1984). Cross-sectional and longitudinalobservations of activity level and sustained attentionin a normative sample. Journal of Abnormal Child Psychology,12, 261–276.Milich, R., Loney, J., & Landau, S. (1982). The independentdimensions of hyperactivity and aggression:A validation with play-room observations. Journal ofAbnormal Psychology, 91, 183–198.Miller, D. C. (1994, Winter). Behavior assessment systemfor children (BASC): Test critique. Texas Associationof School Psychologists Newsletter, 23–29.Miller, L. C. (1967). Louisville Behavior Checklist formales 6–12 years of age. Psychological Reports, 21,885–896.Milliones, J. (1980). Construction of a Black consciousnessmeasure: Psychotherapeutic implications. Psychotherapy:Theory, Research and Practice, 17(2), 175–182.Milner, J. S. (1986). The Child Abuse Potential inventory:Manual – Second Edition. De Kalb, IL: Psytec.Milner, J. S., & Crouch, J. L. (1997). Impact and detection ofresponse distortions on parenting measures used to assessrisk for child physical abuse. Journal of Personality Assessment,69, 633–650.Milner, J. S., & Wimberley, R. C. (1980). Prediction andexplanation of child abuse. Journal of Psychology, 36,875–884.Mintz, N. L., & Schwartz, D. T. (1964). Urban ecologyand psychosis: Community factors in the incidence ofschizophrenia and manic-depression among Italiansin Greater Boston. Inter-national Journal of Social Psychiatry,10, 101–117.Mischel, W. (1968). Personality and assessment. New York:Wiley.Moffitt, T. E. (1993). Adolescence-limited and lifecoursepersistent antisocial behavior: A developmentaltaxonomy. Psychological Reports, 100, 674–701.Moffitt, T. E. (2003). Life-course persistent and adolescence-limitedantisocial behavior: A 10-year researchreview and research agenda. In B. B. Lahey, T. E. Mof-

eferenes479fitt, & A. Caspi (Eds.), Causes of conduct disorder andjuvenile delinquency (pp. 49–75). New York: Guilford.Monica, R. F., Poole, C., Son, L., & Murry, C. C. (1997).Effects of war trauma on Cambodian refugee adolescents’functional health and mental health stains.Journal of the American Academy of Child & AdolescentPsychiatry, 36, 1098–1106.Monuteaux, M. C., Faraone, S. V., Gross, L. M., & Biederman,J. (2007). Predictors, clinical characteristics,and outcome of conduct disorder in girls with attentiondeficit hyperactivity disorder: A longitudinalstudy. Psychological Medicine, 37, 1731–1741.Moore, S. G., & Updegraff, R. (1964). Sociometric statusof pre-school children as related to age, sex, nurturance-giving,and dependency. Child Development,35, 519–524.Moos, R. H., & Moos, B. S. (1986). Family EnvironmentScale Manual – Second Edition. Palo Alto, CA:Consulting Psychologists Press.Morgan, A. B. W. (1998). Childhood social withdrawal:The role of social skills in the relation of withdrawaland anxiety. Dissertation Abstracts International SectionB: The Sciences and Engineering, 59(6-B), 3067.Morgan, C. J., & Cauce, A. M. (1999). PredictingDSM-III-R disorders from the Youth Self-Report:Analysis of data from a field study. Journal of theAmerican Academy of Child and Adolescent Psychiatry,38, 1237–1245.Morgenthaler, W. (1954). Hermann Rorschach,deceased, at his 70th birthday. Schweizerische Zeitschriftfuer Psychologie and ihre Anwendungen, 13, 315.Morris, R. J., & Kratochwill, T. R. (1983). Treatingchildren’s fears and phobias: A behavioral approach.New York: Pergamon Press. Murray, H. A. (1943).Thematic apperception test. Cambridge, MA: HarvardUniversity Press.Morsbach, S. K., & Prinz, R. J. (2006). Understandingand improving the validity of self-report of parenting.Clinical Child and Family Review, 9, 1–21.Munoz, L. C. & Frick, P. J. (2007). The reliability, stability,and predictive utility of the self-report versionof the Antisocial Process Screening Device. ScandinavianJournal of Psychology, 48, 299–312.Mussman, M. C. (1964). Teacher’s evaluations of psychologicalreports. Journal of School Psychology, 3, 35–37.Myers, K., & Winters, N. C. (2002). Ten-year review ofrating scales: II. Scales for internalizing disorders. Journalof the American Academy of Child & Adolescent Psychiatry,41, 634–659.Myles, B. S., Bock, S. J., & Simpson, R. L. (2001).Asperger syndrome diagnostic scale. Los Angeles, CA:Western Psychological Services.Nadeau, K. G. (1995). Attention-deficit disorder in adults:Research, diagnosis, and treatment. New York: Brunner& Mazel.Naglieri, J. A., LeBuffe, P. A., & Pfeiffer, S. I. (1994).Devereux scales of mental disorders. New York: The PsychologicalCorporation.Nail, J. M., & Evans, J. G. (1997). The emotional adjustmentof gifted adolescents: A view of global functioning.Roeper Review, 20(1), 18–21.Neeper, R., & Lahey, B. B. (1986). The Children’sBehavior Rating Scale: A factor analytic developmentalstudy. School Psychology Review, 15, 277–288.Neeper, R., Lahey, B. B., & Frick, P. J. (1990). Comprehensivebehavior rating scale for children. New York:The Psychological Corporation.Negy, C., Lachar, D., Gruber, C., & Garza, N. (2001).The Personality Inventory for Youth: Validity andcomparability of English and Spanish versions forregular education and juvenile justice samples. Journalof Personality Assessment, 76, 250–263.Nelson, B., Martin, R. P., Hodge, S., Havill, V., & Kamphaus,R. (1999). Modeling the prediction of elementaryschool adjustment from preschool temperament.Personality and Individual Differences, 26, 687–700.Neuman, R. J., Todd, R. D., Beath, A. C., Reich, W.,Hudziak, J. J., Bucholz, K. K., et al. (1999). Evaluationof ADHD typology in three contrasting samples: Alatent class approach. Journal of the American Academyof Child and Adolescent Psychiatry, 38(1), 25–33.Newman, J., & Matsopoulos, A. (1994, March). Temperamentprecursors of early reading ability. Paper presentedat the annual meeting of the National Association ofSchool Psychologists, Seattle, WA.Nezu, A. M., & Nezu, C. M. (1993). Identifying andselecting target problems for clinical interventions:A problem-solving model. Psychological Assessment, 5,254–263.Nigg, J. T. (2006). What causes ADHD? Understandingwhat goes wrong and why. New York: Guilford.Nihira, K., Foster, R., Shellhaus, M., Leland, H., Lambert,N., & Windmiller, M. (1981). AAMR adaptivebehavior scale, school edition. Monterey, CA: Publisher’sTest Service.Nihira, K., Leland, H., & Lambert, N. (1993). AAMRadaptive behavior scales-residential and community(2nd. ed.) (AAMR-RC: 2). Austin, TX: Pro-Ed.Nitko, A. (1983). Educational tests and measurement: Anintroduction. New York: Harcourt Brace Jovanovich.Nixon, R. D. V., Sweeney, L., Erickson, D. B., & Touyz,S. W. (2003). Parent-child interaction therapy: A comparisonof standard and abbreviated treatments for oppositionaldefiant preschoolers. Journal of Consulting andClinical Psychology, 71, 251–260Nurnberger, J. I., Jr., Blehar, A. S. C., Kaufmann, C.A., York-Cooler, C., Simpson, S. G., Harkavy-Friedman, J., et al. (1994). Diagnostic interviewfor genetic studies: Rationale, unique features, andtraining. Archives of General Psychiatry, 51, 849–859.O’Brien, B. S., & Frick, P. J. (1996). Reward dominance:Associations with anxiety, conduct problems, and psychopathyin children. Journal of Abnormal Child Psychology,24, 223–240.Oehler-Stinnett, J., & Stinnett, T. A. (1995). Teacherrating scales for attention deficit hyperactivity: A

480 referencescomparative review. Journal of Psychoeducational Assessment,Monograph Series, Special Issue on Assessment ofAttention-Deficit/Hyperactivity Disorders. Cordova,TN: Psychoeducational Corporation.Oesterheld, J. R., & Haber, J. (1997). Acceptability ofthe Conners Parent Rating Scale and Child BehaviorChecklist to Dakotan/Lakotan parents. Journal of theAmerican Academy of Child and Adolescent Psychiatry,36, 55–64.Offord, D. R., Boyle, M. H., Fleming, J. E., Blum, H.M., & Grant, N. I. R. (1989). Ontario child healthstudy: Summary of selected results. Canadian Journalof Psychiatry, 34, 483–491.Okiishi, J., Lambert, M., Eggett, D., Nielsen, L., Dayton,D., & Vermeersch, D. (2006). An analysis of therapisttreatment effects: Toward providing feedback to individualtherapists on their clients’ psychotherapy outcome.Journal of Clinical Psychology, 62, 1157–1172.Ollendick, T. H. (1978). The Fear Survey Schedule forChildren-Revised. Unpublished rating scale: IndianaState University.Ollendick, T. H., & Hersen, M. (1993). Handbook ofchild and adolescent assessment. Boston, MA: Allyn andBacon.Ollendick, T. H., & King, N. J. (1994). Diagnosis,assessment, and treatment of internalizing problemsin children: The role of longitudinal data.. Journal ofConsulting and Clinical Psychology, 62, 918–927.Ollendick, T. H., Seligman, L. D., Goza, A. B., Byrd,D. A., & Singh, K. (2003). Anxiety and depression inchildren and adolescents: A factor-analytic examinationof the tripartite model. Journal of Child and FamilyStudies, 12, 157–170.Olmedo, E. L., Martinez, J. L., Jr., & Martinez, S. R.(1978). Measure of acculturation for Chicano adolescents.Psychological Reports, 42, 159–170.Ondersma, S. J. (2007). Introduction to the special sectionon substance abuse and child maltreatment. ChildMaltreatment, 12, 111–113.Ondersma, S. J., Chaffin, M. J., Mullins, S. M., & LeBreton,J. M. (2005). A brief form of the Child AbusePotential Inventory: Development and validation.Journal of Clinical Child and Adolescent Psychology, 34,301–311.Orvaschel, H., Ambrosini, P., & Rabinovich, H. (1993).Diagnostic issues in child assessment. In T. H. Ollendick& M. Hersen (Eds.), Handbook of child and adolescentassessment (pp. 26–40). Boston, MA: Allyn andBacon.Orvaschel, H., Welsh-Allis, G., & Ye, W. (1988). Psychopathologyin children of parents with recurrentdepression. Journal of Abnormal Child Psychology,16(1), 17–28.Osofsky, J. D., Wewers, S., Hann, D. M., & Fick, A. C.(1993). Chronic community violence: What is happeningto our children? Psychiatry, 56, 36–45.Ostrander, R., Weinfurt, K. P., Yarnold, P. R., & August,G. J. (1998). Diagnosing attention-deficit disorderswith the Behavioral Assessment System for Childrenand the Child Behavior Checklist: Test and constructvalidity analyses using optimal discriminant classificationtrees. Journal of Consulting and Clinical Psychology,66, 660–672.Ostrov, J. M. & Keating, C. F. (2004). Gender differencesin preschool aggression during free play andstructured interactions: An observational study. SocialDevelopment, 13, 255–277.Ownby, R. L., & Wallbrown, F. (1986). Improvingreport writing in school psychology. In T. R. Kratochwill(Ed.), Advances in school psychology (Vol. 5, pp.7–49). Hillsdale, NJ: Erlbaum.Ozonoff, S., Goodlin-Jones, B., & Solomon, M. (2005).Evidence-based assessment of autism spectrum disordersin children and adolescents. Journal of ClinicalChild and Adolescent Psychology, 34, 523–540.Paikoff, R. L., & Brooks-Gunn, J. (1991). Do parent-childrelationships change during puberty? Psychological Bulletin,110, 47–66.Papadopoulos, L., Bor, R., & Stanion, P. (1997). Genogramsin counselling practice: A review (part 1).Counseling Psychology Quarterly, 10, 17–28.Pardini, D. A. (2008). Novel insights into longstandingtheories of bidirectional parent-child influences:Introduction to the special section. Journal of AbnormalChild Psychology, 36, 627–632.Parker, J. G., & Asher, S. R. (1987). Peer relations andlater personal adjustment: Are low-accepted childrenat risk? Psychological Bulletin, 102, 357–389.Patterson, G. R. (1982). Coercive family process. Eugene, OR:Castalia.Patterson, G. R., & Dishion, T. J. (1985). Contributionsof families and peers to delinquency. Criminology, 23,63–79.Patterson, G. R., & Yoerger, K. (1993). Developmentalmodels for delinquent behavior. In S. Hodgins (Ed.),Mental Disorder and Crime. (pp. 140–172). NewburyPark: Sage.Payne, A. F. (1928). Sentence completions. New York: NewYork Guidance Clinic.Pearson, D. A., & Aman, M. G. (1994). Ratings ofhyperactivity and developmental indices: Shouldclinicians correct for developmental level? Journal ofAutism and Developmental Disorders, 24, 395–411.Pearson, D. A., & Lachar, D. (1994). Using behavioralquestionnaires to identify adaptive deficits in elementaryschool children. Journal of School Psychology, 32,33–52.Pearson, K. (1901). On lines and planes of closest fitto systems of points in space. Philosophical Magazine(Series 6), 2, 559–572.Peeples, F., & Loeber, R. (1994). Do individual factorsand neighborhood context explain ethnic differences injuvenile delinquency? Journal of Quantitative Criminology,10, 141–157.Pelham, W. E., Fabiano, G. A., & Massetti, G. M.(2005). Evidence-based assessment of Attention-

eferenes481Deficit Hyperactivity Disorder in children andadolescents. Journal of Clinical Child and AdolescentPsychology, 34, 477–505.Pelham, W. E., Gnabry, E. M., Greenslade, K. E., &Milich, R. (1992). Teacher ratings of DSM-III-Rsymptoms for the disruptive behavior disorders. Journalof the American Academy of ’ Child and AdolescentPsychiatry, 31, 210–218.Perry, M. A. (1990). The interview in developmentalassessment. In J. H. Johnson & J. Goldman (Eds.),Developmental assessment in clinical child psychology: Ahandbook (pp. 58–77). New York: Pergamon Press.Petersen, A. C., Compas, B. E., Brooks-Gunn, J., Stemmler,M., Ey, S., & Grant, K. E. (1993). Depression inadolescence. American Psychologist, 48, 155–168.Petersen, N. S., Kolen, M. J., & Hoover, H. D. (1989)Scaling, norming, and equating. In R. L. Linn (Ed.),Educational measurement (3rd ed., pp. 221–262). NewYork: Macmillan.Peterson, D. R. (1961). Behavior problems of middle childhood.Journal of Consulting Psychology, 25, 205–209.Pfiffner, L. J. & McBurnett, K. (2006). Family correlatesof comorbid anxiety disorders in children with attentiondeficit hyperactivity disorder. Journal of AbnormalChild Psychology, 34, 725–736.Phares, E. J. (1984). Clinical psychology: Concepts, methods,and profession (rev.). Homewood, IL: Dorsey Press.Piacentini, J. (1993). Checklists and rating scales. In T.H. Ollendick & M. Hersen (Eds.), Handbook of childand adolescent assessment (pp. 82–97). Boston, MA:Allyn and Bacon.Piacentini, J. C., Cohen, P., & Cohen, J. (1992). Combiningdiscrepant diagnostic information from multiplesources: Are complex algorithms better thansimple ones? Journal of Abnormal Child Psychology, 20,51–63.Piacentini, J., Robper, M., Jensen, P., Lucas, C., Fisher,P., Bird, H., et al. (1999). Informant-based determinantsof symptom attenuation in structured childpsychiatric interviews. Journal of Abnormal Child Psychology,27, 417–428.Piacentini, J., Shaffer, D., Fisher, P., Schwab-Stone, M.,Davies, A. T., & Gioia, P. (1993). The DiagnosticInterview Schedule for Children – Revised Version(DISC-R): III. Concurrent criterion validity. Journalof the American Academy of Child and Adolescent Psychiatry,32, 648–656.Pinderhughes, E. E., Dodge, K. A., Bates, J. E., Pettit,G. S., & Zelli, A. (2000). Discipline responses:Influences of parents’ socioeconomic status, ethnicity,beliefs about parenting, stress, and cognitiveemotionalprocesses. Journal of Family Psychology, 14,280–400.Piotrowski, C. (1999). Use of tests and measures inmarital and family research. Psychological Reports, 84,1251–1252.Pisecco, S., Lachar, D., Gruber, C. P., Gallen, R. T.,Kline, R. B., & Huzinec, C. (1999). Developmentand validation of disruptive behavior scales for theStudent Behavior Survey (SBS). Journal of PsychoeducationalAssessment, 17, 314–331.Pope, K. S., Butcher, J. N., & Seelen, J. (2000). TheMMPI, MMPI-2 and MMPI-A in court: A practicalguide for expert witnesses and attorneys (2nd ed.). Washington,DC: American Psychological Association.Porter, B., & O’Leary, K. D. (1980). Marital discord andchild-hood behavior problems. Journal of AbnormalChild Psychology, 8, 287–295.Power, T., Doherty, B., Panichelli-Mindel, S., Karustis,J., Eiraldi, R., Anastopoulos, A., et al. (1998). Thepredictive validity of parent and teacher reports ofADHD symptoms. Journal of Psychopathology andBehavioral Assessment, 20, 57–81.Poznanski, E. O., Cook, S. C., & Carroll, J. B. (1979).A depression rating scale for children. Pediatrics, 64,442–450.Prifitera, A., Weiss, L. G., & Saklofske, D. H. (1998).The WISC-III in context. In A. Prifitera & D. Saklofske(Eds.), WISC-III: Clinical use and interpretation(pp. 1–38). San Diego: Academic Press.Pressman, L. J., Loo, S. K., Carpenter, E. M., Asarnow,J. R., Lynn, D., McCracken, J. T., et al. (2006). Relationshipof family environment and parental psychiatricdiagnosis to impairment in ADHD. Journal of theAmerican Academy of Child and Adolescent Psychiatry,45, 346–354.Prinstein, M. J. (2007). Assessment of adolescents’ preference-and reputation-based peer status using sociometricexperts. Merrill-Palmer Quarterly, 53, 243–261.Prinstein, M. J., & Cillessen, A. H. N. (2003). Formsand functions of adolescent peer aggression associatedwith high levels of peer status. Merrill-Palmer Quarterly,49, 310–342.Prinzie, P., Onghena, P., & Hellinckx, W. (2007).Reexamining the Parenting Scale: Reliability, factorstructure, concurrent validity of a scale for assessingthe discipline practices of mothers and fathers ofelementary-school-aged children. European Journalof Psychological Assessment, 23, 24–31.Puig-Antich, J. (1982). Major depression and conductdisorder in prepuberty. Journal of the American Academyof Child and Adolescent Psychiatry, 21, 118–128.Puig-Antich, J., Blau, S., Marx, N., Greenhill, L. L., &Chambers, W. (1978). Prepubertal major depressivedisorders. Journal of the American Academy of Child andAdolescent Psychiatry, 17, 695–707.Puig-Antich, J., & Chambers, W. (1978). The schedulefor affective disorders and schizophrenia for school-agedchildren (Kiddie-SADS). New York: New York StatePsychiatric Institute.Quay, H. C. (1986). Classification. In H. C. Quay & J. S.Werry (Eds.), Psychopathological disorders of childhood(3rd ed., pp. 1–34). New York: Wiley.Quay, H. C., & Love, C. T. (1977). The effect of a juvenilediversion program on rearrests. Criminal Justiceand Behavior, 4, 377–396.

482 referencesQuay, H. C., & Peterson, D. R. (1983). Interim manualfor the Revised Behavior Problem Checklist. Coral Gables,FL: Author.Rabin, A. I. (1986). Concerning projective techniques.In A. I. Rabin (Ed.), Projective techniques for adolescentsand children (pp. 3–13). New York: Springer.Raine, A. (2002). Biosocial studies of antisocial and violentbehavior in children and adults: A review. Journalof Abnormal Child Psychology, 30, 311–326.Ramirez, M. (1984). Assessing and understandingbiculturalism–multiculturalism in Mexican–Americanadults. In J. L. Martinez, Jr. & R. H. Mendoza(Eds.), Chicano Psychology (pp. 77–94). Orlando, FL:Academic Press.Rapoport, J. L., & Castellanos, F. X. (1996). Attentiondeficithyperactivity disorder. In J. M. Wiener (Ed.),Diagnosis and psychopharmacology of childhood and adolescentdisorders (2nd ed., pp. 265–292). New York: Wiley.Rapport, M. D., Chung, K. M., Shore, G., Denney, C.B., & Isaacs, P. (2000). Upgrading the science andtechnology of assessment and diagnosis: Laboratoryand clinic-based assessment of children with ADHD.Journal of Clinical Child Psychology, 29, 555–568.Realmuto, G. M., August, G. J., Sieler, J. D., & Pessoa-Brandao,L. (1997). Peer assessment of socialreputation in community samples of disruptive andnondisruptive children: Utility of the Revised ClassPlay method. Journal of Clinical Child Psychology, 26,67–76.Reed, M. L., & Edelbrock, C. (1983). Reliability andvalidity of the direct observational form of the ChildBehavior Checklist. Journal of Abnormal Child Psychology,11, 521–530.Reich, W. (2000). Diagnostic interview for children andadolescents (DICA). Journal of the American Academyof Child and Adolescent Psychiatry, 39, 59–66.Reich, W., & Earls, F. (1987). Rules for making psychiatricdiagnoses in children on the basis of multiplesources of information: Preliminary strategies. Journalof Abnormal Child Psychology, 1S, 601–616.Reich, W., Herjanic, B., Weiner, Z., & Gandhy, P. R.(1982). Development of a structured psychiatric interviewfor children: Agreement in diagnosis comparingchild and parent interviews. Journal of Abnormal ChildPsychology, 10, 325–336.Reitman, D., Currier, R. O., Hupp, S. D. A., Rhode, P.C., Murphy, M. A., & O’Callahan, P. M. (2001). Psychometriccharacteristics of the parenting Scale in aHead Start population. Journal of Clinical Child Psychology,30, 514–524.Rende, R., & Weissman, M. M. (1999). Assessment offamily history of psychiatric disorder. In D. Shaffer, C.P. Lucas, & J. E. Richters (Eds.), Diagnostic assessmentin child and adolescent psychopathology (pp. 230–255).New York: Guilford Press.Renk, K., & Phares, V. (2004). Cross-informant ratingsof social competence in children and adolescents.Clinical Psychology Review, 24, 239–254.Reynolds, C. R., & Kaiser, S. (1990). Test Bias in psychologicalassessment. In C. R. Reynolds and R. W.Kamphaus (Eds.), Handbook of psychological and educationalassessment of children. New York: Guilford.Reynolds, C. R., & Kamphaus, R. W. (1992). Behaviorassessment system. for children (BASC). Circle Pines,MN: American Guidance Services.Reynolds, C. R., & Kamphaus, R. W. (2004). Behaviorassessment system for children, 2nd edition (BASC-2).Circle Pines, MN: American Guidance Services.Reynolds, C. R., Kamphaus, R. W., & Hatcher, N. M.(1999). Treatment planning and evaluation with theBASC: The Behavior Assessment System for Children.In M. Maruish (Ed.) (2nd ed.), The use of psychologicaltesting for treatment planning and outcome assessment (pp.563–597). Mahwah, NJ: Erlbaum.Reynolds, C. R., & Richmond, B. O. (1985). Revisedchildren’s manifest anxiety scale (RCMAS). Los Angeles:Western Psychological Services.Reynolds, W. M. (1986a). A model for screening andidentification of depressed children and adolescentsin school settings. Professional School Psychology, 1,117–129.Reynolds, W. M. (1986b). Reynolds Adolescent DepressionScale. Odessa, FL: Psychological AssessmentResources.Reynolds, W. M. (1989). Reynolds Child Depression Scale.Odessa, FL: Psychological Assessment Resources.Reynolds, W. M. (1990). Introduction to the nature andstudy of internalizing disorders in children and adolescents.School Psychology Review, 19(2), 137–141.Reynolds, W. M. (1998). Adolescent psychopathology scale.Odessa, FL: Psychological Assessment Resources.Reynolds, W. M. (2002). Reynolds Adolescent DepressionScale, 2nd edition (RADS-2). Lutz, FL: PsychologicalAssessment Resources.Rhoades, K. A. & O’Leary, S. G. (2007). Factor structureand validity of the Parenting Scale. Journal of ClinicalChild and Adolescent Psychology, 36, 137–146.Rice, F., Harold, G. T., & Thapar, A. (2002). The geneticaetiology of childhood depression: A review. Journal ofChild Psychology and Psychiatry, 43, 65–79.Riccio, C., & Reynolds, C. (2003). The assessment ofattention via continuous performance tests. Handbookof Psychological and Educational Assessment ofChildren: Personality, Behavior, and Context (2nd ed.,pp. 291–319). New York: Guilford Press.Richardson, R. D., & Burns, M. K. (2005). AdaptiveBehavior Assessment System (2nd edition) by Harrison,P. L, & Oakland, T. (2002). San Antonio, TX:Psychological Corporation. Assessment for EffectiveIntervention, 30, 51–54.Richmond, B. O., & Kicklighter, R. H. (1979). Children’sadaptive behavior scale. Atlanta, GA: Humantics Limited.Richters, J. E. (1992). Depressed mothers as informantsabout their children A critical review of the evidenceof distortion. Psychological Bulletin, 112, 485–499.

eferenes483Ricks, J. H. (1959). On telling parents about test results.Test Service Bulletin, 54, 1–4.Roberts, B. W. & DelVecchio, W. F. (2000). The rankorderconsistency of personality traits from childhoodto old age: A quantitative review of longitudinal studies.Psychological Bulletin, 126, 3–25.Roberts, C., McCoy, M., Reidy, D., & Crucitti, F.(1993). A comparison of methods of assessing adaptivebehaviour in pre-school children with developmentaldisabilities. Australia and New Zealand Journalof Developmental Disabilities, 18, 261–272.Roberts, G. C., Block, J. H., & Block, J. (1984). Continuityand change in parents’ child-rearing practices.Child Development, 55, 586–597.Roberts, M. A., Milich, R., & Loney, J. (1984). Structuredobservations of academic and play settings (SOAPS).Iowa City, IA: University of Iowa.Roberts, M. A., Ray, R. S., & Roberts, R. J. (1984). Aplayroom observational procedure for assessing hyperactiveboys. Journal of Pediatric Psychology, 9, 177–191.Roberts, R. E., Roberts, C. R., & Xing, Y. (2007). Ratesof DSM-IV psychiatric disorders among adolescentsin a large metropolitan area. Journal of PsychiatricResearch, 41, 959–967.Robertson, D., & Hyde, J. (1982). The factorial validityof the Family Environment Scale. Educational andPsychological Measurement, 42, 1233–1241.Robins, L. N. (1966). Deviant children grown up. Baltimore,MD: Williams and Wilkins.Robins, L. N., Helzer, J. E., Croughan, J., & Ratcliff, K. S.(1981). The National Institute of Mental Health DiagnosticInterview Schedule: Its history, characteristic andvalidity. Archives of General Psychiatry, 38, 381–389.Robins, R. W., John, O. P., Caspi, A., Moffitt, T. E., &Stouthamer-Loeber, M. (1996). Resilient, overcontrolled,and undercontrolled boys: Three replicablepersonality types. Journal of Personality and Social Psychology,70, 157–171.Robinson, E. A., & Eyberg, S. M. (1981). The DyadicParent Child Interaction Coding System: Standardizationand validation. Journal of Consulting and ClinicalPsychology, 49, 245–250.Rockett, J. L., Murrie, D. C., & Boccacini, M. T. (2007).Diagnostic labeling in juvenile justice settings: Dopsychopathy and conduct disorder findings influenceclinicians? Psychological Services, 4, 107–122.Roff, J. D., & Wirt, R. D. (1984). Childhood aggressionand social adjustments as antecedents of delinquency.Journal of Abnormal Child Psychology, 12, 111–126.Roid, G. (2003). Stanford-Binet Intelligence Scale:Fifth Edition. Itasca, IL: Riverside.Rorschach, H. (1921). Psychodiagnostik. Bern: Bircher.Rorschach, H. (1951). Psychodiagnostics: A diagnostictest based on perception (P. Lemkau & B. Kronenberg,Trans.). New York: Grune & Stratton.Rosenthal, G. (1996). The Behavior Assessment System forChildren: Locus of control scale. Child Assessment News,5(5), 1–5, 12.Ross, A. O. (1981). Child behavior therapy: Principles,procedures, and empirical basis. New York: Wiley.Rotter, J. B., & Rafferty, J. E. (1950). Manual for theRotter In-complete Sentences Blank, College Form. NewYork: The Psychological Corporation.Rourke, B. P., & Strang, J. D. (1983). Subtypes of readingand arithmetical disabilities: A neuropsychologicalanalysis. In M. Rutter (Ed.), Developmental neuropsychiatry(pp. 473–488). New York: Guilford Press.Rounsaville, B. J., Alarcon, R. D., Andrews, G., Jackson,J. S., Kendell, R. E., & Kendler, K. (2002). Basicnomenclature issues for DSM-V. In D. J. Kupfer, M.B. First, & D. E. Regier (Eds.), A research agenda forDSM-V (pp. 1–30). Washington, DC: American PsychiatricAssociation.Routh, D. K., & Ernst, A. R. (1983). Somatization disorderin relatives of children and adolescents withfunctional abdominal pain. Journal of Pediatric Psychology,9, 427–437.Rowe, E. W., Circle, E. L., Reynolds, C. R., & Kamphaus,R. W. (1998). Behavior Assessment System forChildren (BASC) Monitor for ADHD: An introductionto applications in practice. The ADHD Report,6(6), 113–129.Rucker, C. N. (1967). Technical language in the school psychologists’report. Psychology in the Schools, 4, 146–150.Rudolph, K. D., Hammen, C., & Burge, D. (1994).Interpersonal functioning and depressive symptomsin childhood: Addressing the issues of specificity andcomorbidity. Journal of Abnormal Child Psychology, 22,355–373.Ruffalo, S. L., & Elliott, S. N. (1997). Teachers’ and parents’ratings of children’s social skills: A closer look atcross-informant agreements through an item analysisprotocol. School Psychology Review, 26, 489–501.Rutter, M. (1983). Behavioral studies: Questions and findingson the concept of a distinctive syndrome. In M.Rutter (Ed.), Developmental neuropsychiatry (pp. 259–280). New York: Guilford.Rutter, M. (1999). Forward. In D. Shaffer, C. P. Lucas,& J. E. Lucas (Eds.), Diagnostic assessment in childand adolescent psychopathology (pp. xii–xiv). New York:Guilford.Rutter, M., & Garmezy, N. (1983). Developmental psychopathology.In E. M. Hetherington (Ed.), Manualof child psychology, IV: Social and personality development(pp. 775–912). New York: Wiley.Rutter, M., & Giller, H. (1983). Juvenile delinquency:Trends and perspectives. London: Penguin.Rutter, M., Chadwick, O., & Shaffer, D. (1983). Headinjury. In M. Rutter (Ed.), Developmental neuropsychiatry(pp. 83–111). New York: Guilford.Saavedra, L. M., & Silverman, W. K. (2002). Classificationof anxiety disorders in children: What a differencetwo decades make. International Review ofPsychiatry, 14, 87–100.Saemundsen, E., Magnusson, P., Smari, J., & Sigurdardottir,S. (2003). Autism Diagnostic Interview

484 references– Revised and the Childhood Autism Rating Scale:Convergence and discrepancy in diagnosing autism.Journal of Autism and Developmental Disorders, 33,319–328.Sandberg, S. T., Wieselberg, M., & Shaffer, D. (1980).Hyperkinetic and conduct problem children in a primaryschool population: Some epidemiological considerations.Journal of Child Psychology, Psychiatry andApplied Disciplines, 21, 293–311.Sanders, M. R., & Dadds, M. R. (1989). Depressionobservation schedule: An observers manual. Brisbane,Australia: University of Queensland.Sanders, M. R., Dadds, M. R., Johnston, B. M., & Cash,R. (1992). Childhood depression and conduct disorder:I. Behavioral, affective, and cognitive aspects of familyproblem-solving interactions. Journal of Abnormal Psychology,101, 495–504.Sanders, M. R., Markie-Dadds, C., Tully, L. A., & Bor,W. (2000). The Triple P-Positive Parenting Program:A comparison of enhanced, standard, and self-directedbehavioral family intervention for parents of childrenwith early onset conduct problems. Journal of Consultingand Clinical Psychology, 68, 624–640.Sandoval, J., & Echandia, A. (1994). Behavior assessmentsystem for children. Journal of School Psychology,32, 419–425.Sandoval, J., & Mille, M. (1979, September). Accuracyjudgments of WISC-R item difficulty for minority groups.Paper presented at the annual meeting of the AmericanPsychological Association, New York.Sandstrom, M. J., & Cillessen, A. H. N. (2006). Likeableversus popular: Distinct implications for adolescentadjustment. International Journal of Behavioral Development,30, 305–314.Sanford, K., Bingham, C. R., & Zucker, R. A. (1999).Validity issues with the Family Environment Scale:Psychometric resolution and research applicationwith alcoholic families. Psychological Assessment, 11,315–325.Sanger, M., MacLean, W., & Van Slyke, D. (1992).Relation between maternal characteristics and childbehavior ratings: Implications for interpreting behaviorchecklists. Clinical Pediatrics, 31, 461–466.Sarason, I.G., Johnson, J.H, & Siegel, J.M. (1978).Assessing the impact of life changes: Developmentof the Life Experiences Survey. Journal of Consultingand Clinical Psychology, 46, 932–946.Sargent, J., Liebman, R., & Silber, M. (1985). Familytherapy for anorexia nervosa. In D. M. Garner& P. E. Garfinkel (Eds.), Handbook of psychotherapyfor anorexia nervosa and bulimia (pp. 257–279). NewYork: Guilford Press.Sattler, J. M. (1988). Assessment of children (3rd ed.). SanDiego, CA: Author.Scahill, L., Schwab-Stone, M., Merikangas, K. R.,Leckman, J. F., Zhang, H., & Kasl, S. (1999). Psychosocialand clinical correlates of ADHD in a communitysample of school-age children. Journal of theAmerican Academy of Child and Adolescent Psychiatry,38, 976–984.Schachar, R., Rutter, M., & Smith, A. (1981). The characteristicsof situationally and pervasively hyperactivechildren: Implications for syndrome definition. Journalof Child Psychology and Psychiatry, 22, 375–392.Schaughency, E. A., & Lahey, B. B. (1985). Mothers’and fathers’ perceptions of child deviance: Roles ofchild behavior, parental depression, and marital satisfaction.Journal of Consulting and Clinical Psychology,53, 718–723.Schaughency, E. A., & Rothlind, J. (1991). Assessmentand classification of attention-deficit hyperactivitydisorder. School Psychology Review, 20, 187–202.Schinka, J. A., Elkins, D. E., & Archer, R. P. (1998).Effects of psychopathology and demographic characteristicson MMPI-A scale scores. Journal of PersonalityAssessment, 71, 295–305.Schniering, C. A., & Rapee, R. M. (2002). Developmentand validation of a measure of children’s automaticthoughts: The Children’s Automatic Thoughts Scale.Behaviour Research and Therapy, 40, 1091–1109.Scholte, E., Van Berckelaer-Onnes, I., & Van der Ploeg,J. (2008). Rating scale to screen symptoms of psychiatricdisorders in children. European Journal of SpecialNeeds Education, 23, 47–62.Schorre, B. E. H., & Vandvik, I. H. (2004). Globalassessment of psychosocial functioning in child andadolescent psychiatry: A review of three unidimensionalscales (CGAS, GAF, GAPD). European Childand Adolescent Psychiatry, 13, 273–286.Schwab-Stone, M., Fisher, P., Piacentini, J., Shaffer, D.,Davies, M., & Briggs, M. (1993). The Diagnostic InterviewSchedule for Children – Revised Version (DISC-R): H. Test-retest reliability. Journal of the AmericanAcademy of Child and Adolescent Psychiatry, 32, 651–657.Schwanz, K. A., & Kamphaus, R. W. (1997). Assessmentand diagnosis of ADHD. In W. N. Bender(Ed.), Understanding ADHD: A practical guide forparents and teachers. Upper Saddle River, NJ: Merrill/Prentice-Hall.Schwean, V. L., Burt, K., & Saklofske, D. H. (1999). Correlatesof mother- and teacher-ratings of hyperactivityimpulsivityand inattention in children with AD/HD.Canadian Journal of School Psychology, 15, 43–62.Scotti, J. R., Morris, T. L., McNeil, C. B., & Hawkins,R. P. (1996). DSM–IV and disorders of childhood andadolescence: Can structural criteria be functional?Journal of Consulting and Clinical Psychology, 64,1177–1191.Semrud-Clikeman, M. (1990). Assessment of childhooddepression. In C. R. Reynolds & Kamphaus, R. W.(Ed.), Handbook of psychological and educational assessmentof children: Personality, behavior, and context (pp.279–297). New York: Guilford Press.Semrud-Clikeman, M. (1991). ADHD and majordepression in children and adolescents. Child AssessmentNews, 1, 6–7.

eferenes485Sevin, J. A., Matson, J. L., Coe, D. A., & Fee, V. F. (1991).A comparison and evaluation of three commonly usedautism scales. Journal of Autism and Developmental Disorders,21, 417–432.Shaevel, B., & Archer, R. P. (1996). Effects of PI-2 andMMPI-A norms on T-Score elevations for 18-yearolds.Journal of Personality Assessment, 67(1), 72–78.Shaffer, D., Fisher, P. W., & Lucas, C. P. (1999).Respondent-based interviews. In D. Shaffer, C. P.Lucas, & J. E. Richters (Eds.), Diagnostic assessmentin child and adolescent psychopathology (pp. 3–33). NewYork: Guilford.Shaffer, D., Fisher, P., Lucas, C. P., Dulcan, M. K., &Schwab-Stone, M. E. (2000). NLMH diagnosticinterview schedule for children version, IV IMHDISC-IV): Description, differences from previousversions, and reliability of some common diagnoses.Journal of the American Academy of Child and AdolescentPsychiatry, 39, 28–38.Shaffer, D., Fisher, P., Piacentini, J., Schwab-Stone, M.,& Wicks, J. (1991). NIMH diagnostic interview schedulefor children-version 2.3. New York: New York StatePsychiatric Institute.Shaffer, D., Garland, A., Gould, M., Fisher, P., & Trautman,P. (1988). Preventing teenage suicide: A criticalreview. Journal of the American Academy of Child andAdolescent Psychiatry, 27, 675–687.Shaffer, D., Gould, M. S., Brasic, J., Ambrosini, P.,Fisher, P., Bird, H., et al. (1983). A children’s globalassessment scale (CGAS). Archives of General Psychiatry,40, 1228–1231.Shaffer, D., Schwab-Stone, M., Fisher, P., Cohen,P., Piacentini, J., Davies, M., Conners, C. K., et al.(1993). The Diagnostic Interview Schedule for Children-RevisedVersion (DISC-R): I. Preparation, fieldtesting, interrater reliability, and acceptability. Journalof the American Academy of Child and Adolescent Psychiatry,32, 643–650.Shapiro, E. S. (1987). Behavioral assessment in school psychology.Hillsdale, NJ: Lawrence Erlbaum.Shapiro, E. S., & Skinner, C. H. (1990). Principles ofbehavioral assessment. In C. R. Reynolds & R. W.Kamphaus (Eds.), Handbook of psychological and educationalassessment of children: Personality, behavior, andcontext (pp. 343–363). New York: Guilford Press.Sharp, C. & Kine, S. (2008). The assessment of juvenilepsychopathy: Strengths and weaknesses of currentlyused questionnaire measures. Child and AdolescentMental Health, 13, 85–95.Shaw, J. G., Hammer, D., & Leland, H. (1991). Adaptivebehavior of preschool children with developmentaldelays: Parent versus teacher ratings. MentalRetardation, 29, 49–53.Shelby, M. D., Nagle, R. J., Barnett-Queen, L. L., Quattlebaum,P. D., & Wuori, D. F. (1998). Parental reportsof psychosocial adjustment and social competence inchild survivors of acute lymphocytic leukemia. Children’sHealth Care, 27(2), 113–129.Shellenberger, S., Dent, M. M., Davis-Smith, M., Seale,J. P., Weintraut, R., & Wright, T. (2007). Culturalgenogram: A tool for teaching and practice. Families,Systems, & Health, 25, 367–381.Shelton, K. K., Frick, P. J., & Wootton, J. (1996). Theassessment of parenting practices in families of elementaryschool-aged children. Journal of Clinical ChildPsychology, 25, 317–327.Sherrill, J. T., & Kovacs, M. (2000). Interview schedulefor children and adolescents (ISCA). Journal of theAmerican Academy of Child and Adolescent Psychiatry,39, 67–75.Shiner, R., & Caspi, A. (2003). Personality differences inchildhood and adolescence: Measurement, development,and consequences. Journal of Child Psychology andPsychiatry, 44, 2–32.Shopler, E., Reichler, R. J., & Renner, R. R. (1988).Child autism rating scale. Los Angeles: Western PsychologicalServices.Sikora, D. M., Hall, T. A., Hartley, S. L., Gerrard-Morris,A. E., & Cagle, S. (2008). Does parent report ofbehavior differ across ADOS-G classifications: Analysisof scores from the CBCL and GARS. Journal ofAutism and Developmental Disorders, 38, 440–448.Silverman, W. K. (1993). DSM and classification ofanxiety disorders in children and adults. In C. G. Last(Ed.), Anxiety across the lifespan: A developmental perspective(pp. 7–36). New York: Springer.Silverman, W. K., & Albano, A. M. (1996). AnxietyDisorders Interview Schedule for Children for DSM-IV:(Child and Parent Versions). San Antonio, TX: PsychologicalCorporation/Graywind.Silverman, W. K., & Nelles, W. B. (1989). The stability ofmothers’ rating of child fearfulness. Journal of AnxietyDisorders, 3, 1–5.Silverman, W. K., & Ollendick, T. H. (2005). Evidencebasedassessment of anxiety and its disorders in childrenand adolescents. Journal of Clinical Child andAdolescent Psychology, 34, 380–411.Silverman, W. K., & Rabian, B. (1999). Rating scalesfor anxiety and mood disorders. In D. Shaffer, C. P.Lucas, & J. E. Richters (Eds.), Diagnostic assessmentin child and adolescent psychopathology (pp. 127–166).New York: Guilford.Silverman, W. K., Saavedra, L. M., & Pina, A. A. (2001).Test–retest reliability of anxiety symptoms and diagnosesusing the Anxiety Disorders Interview Schedulefor DSM-IV: Child and Parent Versions (ADIS forDSM-IV: C/P). Journal of the American Academy ofChild & Adolescent Psychiatry, 40, 937–944.Silverthorn, P., & Frick, P. J. (1999). Developmentalpathways to antisocial behavior: The delayed-onsetpathway in girls. Development and Psychopathology, 11,101–126.Singleton, L. C., & Asher, S. R. (1979). Peer preferencesand social interaction among third-grade children inan integrated school district. Journal of EducationalPsychology, 69, 330–336.

486 referencesSiskind, G. (1967). Fifteen years later: A replication ofa semantic study of concepts of clinical psychologistsand psychiatrists. Journal of Psychology, 65, 3–7.Skinner, B. F. (1963). Operant behavior. American Psychologist,18, 503–515.Smith, B. H., Barkley, R. A., & Shapiro, C. J. (2007). Attention-deficit/hyperactivitydisorder. In E. J. Mash & R. A.Barkley (Eds.), Assessment of childhood disorders (4th ed.).New York: Guilford.Smith, S. R. (2007). Making sense of multiple informantsin child and adolescent psychopathology: Aguide for clinicians. Journal of Psychoeducational Assessment,25, 139–149.Smith, S. R., Baity, M. R., Knowles, E. S., & Hilsenroth,M. J. (2001). Assessment of disordered thinking inchildren and adolescents: The Rorschach Perceptual-Thinking Index. Journal of Personality Assessment, 77,447–463.Smolla, N., Valla, J., Bergeron, L., Berthiaume, C., &St-Georges, M. (2004). Development and reliabilityof a pictorial mental disorders screen for youngadolescents. The Canadian Journal of Psychiatry, 49,828–837.Snyder, J. (1991). Discipline as a mediator of the impactof maternal stress and mood on child conduct problems.Development and Psychopathology, 3, 263–276.Snyder, M., Tanke, E. D., & Berscheid, F. (1977). Socialperception and interpersonal behavior on the self-fulfillingnature of social stereotypes. Journal of Personalityand Social Psychology, 35, 656–666.Solis, M. L., & Abidin, R. R. (1991). The Spanish versionof the Parenting Stress Index: A psychometric study.Journal of Clinical Child Psychology, 20, 372–378.Solomon, I. L., & Starr, B. D. (1968). School apperceptionmethod: SAM. New York: Springer.Sorneson, E., & Johnson, E. (1996). Subtypes of incarcerateddelinquents constructed via cluster analysis.Journal of Child Psychology, 37, 293–303.Sourander, A., Helstelae, L., & Helenius, H. (1999).Parent-adolescent agreement on emotional andbehavioral problems. Social Psychiatry and PsychiatricEpidemiology, 34, 657–663.South, M., Williams, B. J., McMahon, W. M., Owley,T., Filipek, P. A., Shernoff, E., et al. (2002). Utilityof the Gilliam Autism Rating Scale in research andclinical populations. Journal of Autism and DevelopmentalDisorders, 32, 593–599.Southam-Gerow, M. A., & Chorpita, B. F. (2007). Anxietyin children and adolescents. In E. J. Mash & R.A. Barkley (Eds.), Assessment of childhood disorders (4thed., pp. 347–397). New York: Guilford.Spanier, G. B. (1976). Measuring dyadic adjustment:New scales for assessing the quality of marriage andsimilar dyads. Journal of Marriage and the Family, 38,15–28.Sparrow, S. S., Balla, D. A., & Cicchetti, D. V. (1984).Vineland adaptive behavior scales. Circle Pines, MN:American Guidance Services.Sparrow, S. S., Cicchetti, D. V., & Balla, D. A. (2005).Vineland adaptive behavior scales (2nd ed.). CirclePines, MN: American Guidance Services.Speece, D. L., & Cooper, D. H. (1990). Ontogeny ofschool failure: Classification of first grade children.American Educational Research Journal, 27, 119–140.Spielberger, C. D. (1973). Manual for the State-TraitAnxiety Inventory for children. Palo Alto, CA: ConsultingPsychologists Press.Spielberger, C. D., Gorsuch, R. L., & Lushene, R. E.(1970). STAI manual for the State-Trait Inventory. PaloAlto, CA: Consulting Psychologist Press.Spitzer, R., & Cantwell, D. (1980). The DSM-III classificationof psychiatric disorders of infancy, childhood,and adolescence. Journal of the American Academy ofChild and Adolescent Psychiatry, 19, 356–370.Spitzer, R. L., Davies, M., & Barkley, R. A. (1990). TheDSM-III-R field trials of disruptive behavior disorders.Journal of the American Academy of Child and AdolescentPsychiatry, 29, 690–697.Spivack, G., & Spotts, J. (1966). Devereux child behaviorrating scale. Devon, PA: The Devereux Foundation.Spivack, G., Spotts, J., & Haines, P. E. (1967). Devereuxadolescent behavior rating scale. Devon, PA: TheDevereux Foundation.Stanger, C., Dumenci, L., Kamon, J., & Burstein, M.(2004). Parenting and children’s externalizing problemsin substance-abusing families. Journal of ClinicalChild and Adolescent Psychology, 33, 590–600.Strakaki, C., & Ellis, J. (1989). The relationship ofanxiety to depression in children and adolescents.In C. Stavrakaki (Ed.), The psychiatric clinics of NorthAmerica: Affective disorders and anxiety in the child andadolescent (pp. 777–789). Philadelphia, PA: W. B.Saunders.Steele, R. G., Nesbitt-Daly, J. S., Daniel, R. C., & Forehand,R. (2005). Factor structure of the ParentingScale in a low-income African-American sample.Journal of Child and Family Studies, 14, 535–549.Stein, L. A. R., & Graham, J. R. (1999). Detecting fakegoodMMPI-A profiles in a correctional facility. PsychologicalAssessment, 11(3), 386–395.Stein, L., & Graham, J. (2005). Ability of substanceabusers to escape detection on the Minnesota MultiphasicPersonality Inventory-Adolescent (MMPI-A) in a juvenile correctional facility. Assessment, 12,28–39.Steinberg, L., Lamborn, S., Dornbusch, S., & Darling,N. (1992). Impact of parenting practices on adolescentachievement: Authoritative parenting, schoolinvolvement, and encouragement to succeed. ChildDevelopment, 63, 1266–1281.Steinhausen, H. C., & Metzke, C. W. (1998). Youthself-report of behavioral and emotional problems in aSwiss epidemiological study. Journal of Youth and Adolescence,27, 429–441.Stone, W. L., Coonrod, E. E., Pozdol, S. L., & Turner, L.M. (2003). The Parent Interview for Autism – clinical

eferenes487version (PIA-CV): A measure of behavioral changefor young children with autism. Autism, 7, 9–30.Stone, W. L., & Hogan, K. L. (1993). A structuredparent interview for identifying young children withautism. Journal of Autism and Developmental Disorders,23(4), 639–652.Stoolmiller, M., Eddy, J. M., & Reid, J. B. (2000). Detectingand describing preventive intervention effects ina universal school-based randomized trial targetingdelinquent and violent behavior. Journal of Consultingand Clinical Psychology, 68, 296–306Stowers-Wright, L. E. (2000). A behavioral cluster analysisof male juvenile defenders. Unpublished doctoral dissertation:The University of Georgia.Strauss, A. A., & Lehtinen, L. F. (1947). Psychopathologyand education of the brain-injured child. New York:Grune & Stratton.Strauss, C. C. (1990). Overanxious disorder in childhood.In M. Hersen & C. G. Last (Eds.), Handbookof child and adult psychopathology (pp. 237–246). NewYork: Pergamon Press.Strauss, C. C. (1993). Developmental differences inexpression of anxiety disorders in children and adolescents.In C. G. Last (Ed.), Anxiety across the lifespan:A developmental perspective (pp. 63–77). New York:Springer.Strauss, C. C., Lahey, B. B., Frick, P. J., Frame, C. L.& Hynd, G. W. (1988). Peer social status of childrenwith anxiety disorders. Journal of Consulting and ClinicalPsychology, 56, 137–141.Stredny, R. V. & Ball, J. D. (2005). The utility of the RorschachCoping Deficit Index as a measure of depressionand social skills deficits in children and adolescents.Assessment, 12, 295–302.Sundberg, N. D. (1961). The practice of psychologicaltesting in clinical settings in the United States. AmericanPsychologist, 16, 79–83. Reprinted in L. D. Goodstein& R. I. Lanyon (Eds.), Readings in PersonalityAssessment. New York: Wiley.Sundre, D. L. (1998). [Review of the Personnel ReactionBlank]. Thirteenth Mental Measurements Yearbook.Lincoln, NE: Buros Institute.Sutter, J., & Eyberg, S. M. (1984). Sutter-Eyberg studentbehavior inventory. Gainesville, FL: University ofFlorida.Suveg, C., Zeman, J., Flannery-Schroeder, E., & Cassano,M. (2005). Emotion socialization in families of childrenwith an anxiety disorder. Journal of Abnormal Child Psychology,33, 145–155.Swensen, C. H. (1968). Empirical evaluations of humanfigure drawings: 1957–1966. Psychological Bulletin, 70,20–44. Reprinted in L. D. Goodstein & R. I. Lanyon(Eds.), Readings in Personality Assessment. New York:Wiley.Szapocznik, J., & Kurtines, W. (1980). Acculturation,biculturalism and adjustment among Cuban Americans.In A. M. Padilla (Ed.), Acculturation: Theory,models and some new findings (pp. 139–159). AmericanAssociation for the Advancement of ScienceSelected Symposium 39. Boulder, CO: Westview.Szapocznik, J., Scopetta, M. A., & Aranalde, M. A.(1978). Theory and measurement of acculturation.Interamerican Journal of Psychology, 12, 113–130.Szatmari, P., Saigal, S., Rosenbaum, P., & Campbell,D. (1993). Psychopathology and adaptive functioningamong extremely low birth weight children ateight years of age. Development and Psychopathology,5, 345–357.Tallent, N. (1988). Psychological report writing (3rd ed.).Englewood Cliffs, NJ: Prentice-Hall.Tallent, N. (1993). Psychological report writing (4th ed.).Englewood Cliffs, NJ: Prentice-Hall.Tanaka, J. S., & Westerman, M. A. (1988). Commondimensions in the assessment of competence inschool-aged girls. Journal of Educational Psychology,80, 579–584.Tannock, R. (2000). Attention deficit disorders withanxiety disorders. In T. Brown (Ed.), Attention deficitdisorders and comorbidities in children, adolescents andadults (pp. 125–170). Washington, DC: AmericanPsychiatric PressTapscott, M., Frick, P. J., Wootton, J. & Kruh, I. (1996).The intergenerational link to antisocial behavior:Effects of paternal contact. Journal of Child and FamilyStudies, 5, 229–240.Teglasi, H. (1983). Report of a psychological assessmentin a school setting. Psychology in the Schools, 20,466–479.Tellegen, A., & Ben-Porath, Y. S. (1992). The new uniformT-scores for the A[MPI-2: Rationale, derivation,and appraisal. Psychological Assessment, 4, 145–155.Teodori, J. B. (1993). Neurological assessment. In T. H.Ollendick & M. Hersen (Eds.), Handbook of child andadolescent assessment (pp. 41–50). Boston, MA: Allynand Bacon.Teufel-Shone, N., Staten, L. K., Irwin, S., Rawiel, U.,Bravo, A. B., & Waykayuta, S. (2005). Family cohesionand conflict in an American Indian community.American Journal of Health Behavior, 29, 413–422.Thomas, A., & Chess, S. (1977). Temperament and development.New York: Bruner & Mazel.Thompson, C. E. (1949). Manual for thematic apperceptiontest: Thompson modification. Cambridge, MA:Harvard University Press.Thompson, L. L., Riggs, P. D., Mikulich, S. K., & Crowley,T. J. (1996). Contribution of ADHD symptomsto substance problems and delinquency in conductdisorderedadolescents. Journal of Abnormal Child Psychology,24, 325–348.Thurber, S., & Hollingsworth, D. K. (1992). Validityof the Achenbach and Edelbrock Youth Self-Reportwith hospitalized adolescents. Journal of Clinical ChildPsychology, 21, 249–254.Tibbetts, S. G., & Piquero, A. R. (1999). The influenceof gender, low birth weight, and disadvantaged environmentin predicting early onset of offending: A test

488 referencesof Moffitt’s interactional hypothesis. Criminology, 37,843–877.Tillman, R., Geller, B., Craney, J. L., Bolhofner, K.,Williams, M., Zimerman, B., Frazier, J., et al. (2003).Temperament and character factors in a prepubertaland early adolescent bipolar disorder phenotypecompared to attention deficit hyperactive and normalcontrols. Journal of Child and Adolescent Psychopharmacology,13, 531–543.Tripp, G., Schaughency, E., & Clarke, B. (2006).Parent and teacher rating scales in the evaluationof attention-deficit hyperactivity disorder: Contributionto diagnosis and differential diagnosisin clinically referred children. Journal of Developmental& Behavioral Pediatrics, 27, 209–218.Trites, R. L., Bouin, A. G. A., & Laprade, K. (1982).Factor analysis of the Conners teacher rating scalebased on a large normative sample. Journal of Consultingand Clinical Psychology, 50, 615–623.Tsoubris, K. T. (1998). Relationship between prenatalcocaine exposure and behavioral patterns among preschoolchildren with disabilities. Dissertation AbstractsInternational Section B: The Sciences and Engineering,59(5-B), 2441.Tuma, J. M., & Elbert, J. C. (1990). Critical issues andcurrent practice in personality assessment of children.In C. R. Reynolds & R. W. Kamphaus (Eds.), Handbookof psychological and educational assessment of children:Personality, behavior, and context (pp. 3–29). NewYork: Guilford.Tupes, E. C., & Christal, R. E. (1961). Recurrent personalityfactors based on trait ratings. Technical ReportASD-TR-61–97. Lackland Air Force Base, TX: U.S.Air Force.Vaillancourt, T., & Hymel, S. (2006). Aggression andsocial status: The moderating roles of sex and peervaluedcharacteristics. Aggressive Behavior, 32, 396–408.Valdez, G. M., & McNamara, J. R. (1994). Matching toprevent adoption disruption. Child & Adolescent SocialWork Journal, 11, 391–403.Valla, J. P., Bergeron, L., Bidaut-Russell, M., St., M., &Gaudet, N. (1997). Reliability of the Dominic-R: Ayoung child mental health questionnaire combiningvisual and auditory stimuli. Journal of Child Psychologyand Psychiatry, 38, 717–724.Valla, J., Bergeron, L., & Smolla, N. (2000). The Dominic-R:A pictorial interview for 6- to 11-year-oldchildren. Journal of the American Academy of Child andAdolescent Psychiatry, 39, 85–93.Vance, F. I. (1965). I was an imaginary playmate.American Psychologist, 20, 990.Van Hasselt, V. B., Hersen, M., Bellack, A. S., Rosenblum,N. D., & Lamparski, D. (1979). Tripartiteassessment of the effects of systematic desensitizationin a multi-phobic child: An experimental analysis.Journal of Behavior Therapy and Experimental Psychiatry,10, 51–55.Van Horn, M. L., Atkins-Burnett, S., Karlin, E., Ramey,S. L., & Snyder, S. (2007). Parent ratings of children’ssocial skills: Longitudinal psychometric analyses ofthe Social Skills Rating System. School PsychologyQuarterly, 22, 162–199.Vasay, M. W., & Lonigan, C. J. (2000). Consideringthe clinical utility of performance-based measures ofchildhood anxiety. Journal of Clinical Child Psychology,29, 493–508.Vaughn, M. L., Riccio, C. A., Hynd, G. W., & Hall, J.(1997). Diagnosing ADHD subtypes: Discriminantvalidity of the Behavior Assessment System for Children(BASC) and the Achenbach paren and teacherrating scales. Journal of Clinical Child Psychology, 26,349–357.Verhulst, F. C., Koot, H. M., & Van der Ende, J. (1994).Differential predictive value of parents’ and teachers’reports of children’s problem behaviors: A longitudinalstudy. Journal of Abnormal Child Psychology, 22,531–546.Victor, J. B. (1994). The five-factor model applied toindividual differences in school behavior. In C. F.Halverson, G. A. Kohnstamm, & R. P. Martin (Eds.),The developing structure of temperament and personalityfrom infancy to adulthood (pp. 355–366). Hillsdale, NJ:Erlbaum.Vincent, G. M. (2006). Psychopathy and violence riskassessment in youth. Child and Adolescent PsychiatricClinics of North America, 15, 407–428.Voelker, S. L., Shore, D. L., Brown-More, C., & Hill, L.C. (1990). Validity of self-report of adaptive behaviorskills by adults with mental retardation. Mental Retardation,28, 305–309.Waldman, I. D. & Gizer, I. R. (2006). The genetics ofattention deficit hyperactivity disorder. Clinical PsychologyReview, 26, 396–432.Walker, C. E., Milling, L. S., & Bonner, B. L. (1988).Incontinence disorders: Enuresis and encopresis. InD. K. Routh (Ed.), Handbook of pediatric psychology(pp. 363–398). New York: Guilford Press.Walker, H. M., & McConnell, S. R. (1988). The Walker-McConnell scale of f social competence and school adjustment.Austin, TX: Pro-Ed.Walker, J. L., Lahey, B. B., IIynd, G. W., & Frame, C. L.(1987). Comparison of specific patterns of antisocialbehavior in children with conduct disorder with or withoutcoexisting hyper-activity. Journal of Consulting andClinical Psychology, 55, 666–678.Walker, J. L., Lahey, B. B., Russo, M. F., Frick, P. J., Christ,M. A. G., McBurnett, K., et al. (1991). Anxiety, inhibition,and conduct disorder in children: I. Relations tosocial impairment. Journal of the American Academy ofChild and Adolescent Psychiatry, 30, 187–191.Walsh, C., MacMillan, H. L., & Jamieson, E. (2003).The relationship between parental substance abuseand child maltreatment: Findings from the OntarioHealth Supplement. Child Abuse and Neglect, 27,1409–1425.

eferenes489Wang, J. J. (1992, April). An analytical approach to generatingnorms for skewed normative distributions. Paperpresented at the Annual Meeting of the AmericanEducational Research Association, San Francisco.Ward, J. H. (1963). Hierarchical grouping to optimizean objective function. Journal of the American StatisticalAssociation, 58, 236–244.Waschbusch, D. A. (2002). A meta-analytic examinationof comorbid hyperactive-impulsive-attentionproblems and conduct problems. Psychological Bulletin,128, 118–150.Wasielewski, S. M., Gridley, B. J., & Hall, J. (1998,April). Construct validity of the Behavior AssessmentSystem for Correlations with the Personality Inventoryfor Children-Revised with a clinical sample. Paper presentedat the Annual Meeting of the National Associationof School Psychologists, Orlando.Watkins, C. E., Campbell, V. L., Nieberding, R., &Hallmark, R. (1995). Contemporary practice of psychologicalassessment by clinical psychologists. ProfessionalPsychology: Research and Practice, 26, 54–60.Webster-Stratton, C., & Lindsay, D. W. (1999). Socialcompetence and conduct problems in young children:Issues in assessment. Journal of Clinical Child and AdolescentPsychology, 28, 25–43.Wechsler, D. (1974). Wechsler intelligence scale for children- revised. New York: Psychological Corporation.Wechsler, R. D. (1991). Wechsler intelligence scale for children(3rd ed.). New York: Psychological Corporation.Weems, C. F., Hammond-Laurence, K., Silverman, W.K., & Ginsburg, G. S. (1998). Testing the utility of theanxiety sensitivity construct in children and adolescentsreferred for anxiety disorders. Journal of ClinicalChild Psychology, 27, 69–77.Weems, C. F., Silverman, W. K., Rapee, R. R., & Pina, A.A. (2003). The role of control in childhood anxiety disorders.Cognitive Therapy and Research, 27, 557–568.Weems, C. F., & Stickle, T. R. (2005). Anxiety disordersin childhood: Casting a nomological net. ClinicalChild and Family Psychology Review, 8, 107–134.Weiner, I. B. (1986). Assessing children and adolescentswith the Rorschach. In H. M. Knoff (Ed.), The adjustmentof child and adolescent personality (pp. 141–172).New York: Guilford Press.Weiner, I. B. (2001). Advancing the science of psychologicalassessment: The Rorschach Inkblot Method asexemplar. Psychological Assessment. 13, 423–432.Weiss, B., & Garber, G. (2003). Developmental differencesin the phenomenology of depression. Developmentand Psychopathology, 15, 403–430.Weiss, B., & Weisz, J. R. (1988). Factor structure of selfreporteddepression: Clinic-referred children versus adolescents.Journal of Abnormal Psychology, 97, 492–495.Weiss, G., Hechtman, L., & Perlman, T. (1978). Hyperactivesas young adults: School, employer, and selfrating scales obtained during ten-year follow-upevaluation. American Journal of Orthopsychiatry, 48,438–445.Weissman, M. M., Warner, V., & Fendrich, M. (1990).Applying impairment criteria to children’s psychiatricdiagnosis. Journal of the American Academy of ChildAdolescent Psychiatry, 29, 789–795.Weissman, M. M., Warner, V., Wickramaratne, P., &Prusoff, B. A. (1988). Early-onset major depressionin parents and their children. Journal of Affective Disorders,15, 269–277.Weissman, M. M., Wickramaratne, P., Adams, P., Wolk,S., Verdeli, H., & Olfson, M. (2000). Brief screeningfor family psychiatric history: The family historyscreen. Archives of General Psychiatry, 57, 675–682.Weller, E. B., Weller, R. A., Fristad, M. A., Rooney,M. T., & Schecter, J. (2000). Children’s interviewfor psychiatric syndromes (ChIPS). Journal of theAmerican Academy of Child and Adolescent Psychiatry,39, 76–84.Wender, P. H. (1995). Attention deficit disorder in adults.London: Oxford University.West, M. O., & Prinz, R. J. (1987). Parental alcoholismand childhood psychopathology. PsychologicalBulletin, 102, 204–218.Whalen, C. K., & Henker, B. (1998). Attention-deficithyperactivity disorder. In T. H. Ollendick & M.Hersen (Eds.), Handbook of child psychopathology (3rded., pp. 181–212). New York: Plenum Press.Whalen, C. K., Henker, B., & Dotemoto, S. (1980).Methylphenidate and hyperactivity: Effects on teacherbehaviors. Science, 208(4449), 1280–1282.Whalen, D. K., Henker, B., Jamner, L. D., Ishikawa, S.S., Floro, J. N., Swindle, R., et al. (2006). Toward mappingdaily challenges of living with ADHD: Maternaland child perspectives using electronic diaries. Journal ofAbnormal Child Psychology, 34, 115–130.Whipple, E.E. & Webster-Stratton, C. (1991). The roleof parental stress in physically abusive families. ChildAbuse & Neglect, 15, 2779–291.Whiteside, S. P., McCarthy, D. M. & Miller, J. D.(2007). An examination of the factor structure of theSocial Skills Rating System parent elementary form.Assessment, 14, 246–254.Whiteside-Mansell, L., Ayoub, C., McKelvey, L., Faldowski,R. A., Hart, A., & Shears, J. (2007). Parentingstress of low-income parents of toddlers andpreschoolers: Psychometric properties of a short formof the Parenting Stress Index. Parenting: Science andPractice, 7, 27–56.Widaman, K. F., Stacy, A. W., & Borthwick-Duffy, S.A. (1993). Construct validity of dimensions of adaptivebehavior: A multitrait–multimethod evaluation.American Journal on Mental Retardation, 98, 219–234.Widiger, T. A., Frances, A. J., Davis, W., & First, M.(Eds.) (1994). DSM-IV sourcebook (Vol. 1). Washington,DC: American Psychiatric Association.Widiger, T. A., Frances, A. J., Pincus, H. A., Davis, W.W., & First, M. B. (1991). Toward an empirical classificationfor the DSM-IV. Journal of Abnormal Psychology,100, 280–288.

490 referencesWidiger, T. A., Frances, A. J., Pincus, H. A., Ross, R.,First, M. B., Davis, W., et al. (1998). DSM-IV sourcebook:Vol. 4. Washington, DC: American PsychiatricAssociation.Wiener, J. (1985). Teachers’ comprehension of psychologicalreports. Psychology in the Schools, 22, 60–64.Wiender, I. B. (2001). Advancing the science of psychologicalassessment: The Rorschach Inkblot Method asexemplar. Psychological Assessment, 13, 423–432.Wiener, J., & Kohler, S. (1986). Parents’ comprehensionof psychological reports. Psychology in the Schools, 23,265–270.Williams, C. L., Butcher, J. N., Ben-Porath, Y. S., &Graham, J. R. (1992). MMPI-A Content Scales: Assessingpsychopathology in adolescents. Minneapolis: Universityof Minnesota Press.Wingenfeld, S. A., Lachar, D., Gruber, C. P., & Kline,R. B. (1998). Development of the teacher-informantStudent Behavior Survey. Journal of PsychoeducationalAssessment, 16, 226–249.Winokur, G. (1990). The concept of secondary depressionand its relationship to comorbidity. In G. Winokur& R. B. Wesner (Eds.), The psychiatric clinicsof North America: Anxiety and depression as secondaryphenomena (pp. 567–583). Philadelphia, PA: W. B.Saunders.Wirt, R. D., Lachar, D., Klinedinst, J. K., & Seat, P. S.(1984). Multidimensional description of child personality:A manual for the personality inventory for children.Los Angeles: Western Psychological Services.Wirt, R. D., Lachar, D., Klinedinst, J. K., & Seat, P. S.(1990). Personality inventory for children-1990 edition.Los Angeles: Western Psychological Services.Witt, J. C., Heffer, R. W., & Pfeiffer, J. (1990). Structuredrating scales: A review of self-report and informantrating processes, procedures, and issues. In C.R. Reynolds & R. W. Kamphaus (Eds.), Handbookof psychological and educational assessment of children:Personality, behavior and context (pp. 364–394). NewYork: Guilford Press.Witt, J. C. & Jones, K. M. (1998). Review of the BehaviorAssessment System for Children. In J. C. Impara& B. S. Plake (Eds.), The Thirteenth, Mental MeasurementsYearbook (pp. 131–133). Lincoln, NE: BurosInstitute.Wolf, T. H. (1966). Intuition and experiment: AlfredBinet’s first efforts in child psychology. Journal of Historyof Behavioral Sciences, 2, 233–239.Wolfe, D. A., Edwards, B., Manion, I., & Koverola, C.(1988). Early intervention of parents at risk for childabuse and neglect: A preliminary investigation. Journalof Consulting and Clinical Psychology, 56, 40–47.Woodcock, R. W., & Johnson, M. B. (1989). Woodcock-Johnson tests of achievement – revised. New York: Riverside.Woodin, M. F. (1999). Screening for the interfacebetween attention, executive functioning, and workingmemory: A cluster and profile analytic study. DissertationAbstracts International Section B: The Sciencesand Engineering, 59(10-B), 5590.Woodworth, R. S. (1918). Personal data sheet. Chicago:Stoelting.Wootten, S. A. (1999). Attention-deficit/hyperactivitydisorder. Dissertation Abstracts International Section B:The Sciences & Engineering, 60(1-B), 0380.Wootton, J. M., Frick, P. J., Shelton, K. K., & Silverthorn,P. (1997). Ineffective parenting and childhoodconduct problems: The moderating role of callousunemotionaltraits. Journal of Consulting and ClinicalPsychology, 65, 301–308.Worchel, F. F., & Dupree, J. L. (1990). Projective storytellingtechniques. In C. R. Reynolds & R. W. Kamphaus(Eds.), Handbook of psychological and educationalassessment of children: Personality, behavior, and context(pp. 70–88). New York: The Guilford Press.World Health Organization (1993). The ICD-10 classificationof mental and behavioural disorders: Diagnostic criteriafor research. Geneva: World Health Organization.Wrobel, T. A., Lachar, D., Wrobel, N. H., Morgan, S.T., Gruber, C. P., & Neher, J. A. (1999). Performanceof the Personality Inventory for Youth validity scales.Assessment, 6, 367–376.Wrobel, N., Lachar, D., & Wrobel, T. (2005). Selfreportproblem scales and subscales and behavioralratings provided by peers: Unique evidence of testvalidity. Assessment, 12, 255–269.Youngstrom, E., Izard, C., & Ackerman, B. (1999).Dysphoria-related bias in maternal ratings of children.Journal of Consulting and Clinical Psychology, 67,905–916.Zettergren, P. (2007). Cluster analysis in sociometricresearch: A pattern-oriented approach to identifyingtemporally stable peer status groups of girls. Journal ofEarly Adolescence, 27, 90–114.Zettergren, P., Bergman, L. R., & Wangby, M. (2006).Girls’ stable peer status and their adult adjustment: Alongitudinal study from age 10 to age 43. InternationalJournal of Behavioral Development, 30, 315–325.Zimmerman, I. L., & Woo-Sam, J. M. (1967). Reportingresults. In A. J. Glasser & I. L. Zimmerman (Ed.),Clinical interpretation of the Wechsler Intelligence Scalefor Children (WISC). New York: Grune & Stratton.Zimmerman, I. L., & Woo-Sam, J. M. (1985). Clinicalapplications. In B. B. Wolman (Ed.), Handbookof intelligence: Theories, measurements, and applications(pp. 873–898). New York: Wiley.Zuckerman, M. (1990). Some dubious premises inresearch and theory on racial differences: Scientific,social, and ethical issues. American Psychologist, 45,1297–1303.

A u t h o r I n d e xAbad, J., 115Abbott, M.J., 428Abidin, 289, 291Abidin, R.R., 289, 291Abikoff, H., 400Ablow, J.C., 264Abramowitz, A.J., 397Achenbach, T.M., 15, 16, 18–20, 52, 61–63, 73, 89,103, 109, 110, 113–116, 126, 132, 146, 148,155, 156, 158–164, 170, 174, 178, 187, 200,201, 212, 235, 333, 334, 335, 336, 340, 346,350, 377, 386, 389, 413, 425,Acker, M.M., 280Adrien, J.L., 439Ageton, S.S., 345, 408Ahern, L.S., 288Albano, A.M., 256, 431Albert-Corush, J., 294Alderman, E., 72Allaire, J.C., 288Allen, J.C., 235Allen, M.J., 21Altepeter, T.S., 184Amador, J.A., 115Amato, P.R., 277, 291Ambrosini, P.J., 257, 415Ames, L.B., 11, 231Amorosa, H., 440Anastasi, 186, 226–228, 245Anastasi, A., 13, 21, 22, 34, 37, 39, 40, 44, 67, 76Anastopoulos, A.D., 382, 393, 394Anderson, J.C., 260Andrews, J., 22Angold, A., 256, 378, 380, 417Applegate, B., 380, 381Aranalde, M.A., 79Arbelaez, C., 416Archer, R., 122, 128Archer, R.P., 19, 121, 122, 125–128, 130, 235, 277,Arick, J.R., 446Arita, A.A., 126Arney, F., 286, 393Arnold, D.S., 280, 286Aronen, E.T., 115, 116Arseneault, L., 263Asher, S.R., 212–214, 217, 219, 223Aten, J.D., 305Austin, 291Baer, R.A., 121, 126Baghurst, P., 286Bailey, A., 438Bailey, J.M., 226, 227Baity, M.R., 236Baker, B.L., 285Baldwin, 78Baldwin, J.S, 429Ball, J.D., 229, 235, 341Balla, 316, 320–322Balla, D.A., 33, 316, 320–322Ballenger, J., 121Banez, 291Bao-Yu, W., 281Barckley, M., 22Barker, P., 89, 94, 95Barkley, R.A., 55, 184, 207, 279, 294, 378, 381–384,395, 397, 399Barlow, D.H., 427, 428Barnett, D.C., 42Barnett, J., 249Barrett, P.M., 295Barrios, B.A., 192, 198, 199, 204Barry, C.T., 116, 167Barth, J.M., 219Barthelemy, C., 439Bates, J.E., 291Bauer, D.H., 58Baumrind, D, 280Beck, A., 420Beck, S., 218Beck, S.T., 295, 296Behar, L.B., 413Belardinelli, C., 281Belevich, J.K.S., 125Bell, 78Bell, N.L., 240, 243Bellack, A.S., 196, 203Bellak, L., 11, 237491

492 author indexBellak, S.S., 11, 237Bellmore, A.D., 212Belter, R.W., 19, 229, 248Ben-Porath, Y.S., 29, 30, 120Berg, I., 429Bergman, L.R., 19, 212Berscheid, F., 51Bertrand, J., 437Beutler, L.E., 23Bickman, L., 23, 82Biederman, J., 62, 142, 294Bigras, M., 291Bingham, C.R., 282Bird, H.R., 62, 185, 344, 345Birmaher, B., 416Black, K., 130Black, M.M., 291Blader, J.C., 285Blair, R.J.R., 401Blashfield, R.K., 49Blum, G.S., 237Boccacini, M.T., 51Bock, S.J., 446Bodin, 167Boggs, S.R., 280Bohnert, A., 262Bolte, S., 317Bonner, B.L., 58Bor, R., 305Bor, W., 286Borgen, F.H., 42Bornstein, M.R., 196Borthwick-Duffy, S.A., 318, 336Bower, E.M., 18Bowlby, J., 58Boyd, C.P., 282Boyle, M.H., 384Boyle-Whitesel, M., 164Braaten, E., 358Bradford, D.C., 190Brakkc, N., 218Bravo, M., 74Breen, M.J., 184Brennan, 416Brent, D., 416Breslau, N., 282Bretzing, B., 30Briggs-Gowan, M.J., 276Brooks-Gunn, J., 346Brown, M., 226Brown, R.A., 115Brown, T.E., 393Brubaker, R.G., 285Bruininks, R.H., 320, 328, 329Bryson, S.E., 439Budd, K.S., 289Burge, D., 204Burke, J.D., 402Burns, G.L., 186Burns, M.K., 331Burns, R.C., 249Burstein, M., 285Burt, T., 282Buss, A.H., 57Butcher, J.N., 40, 41, 103, 117–123, 125, 126, 128Cabiya, J., 122Cagle, S., 443Campbell, 18Campbell, J., 446Campbell, J.P., 51Campbell, M., 443Campbell, S.B., 58, 381Campbell, V.L., 226Canino, G., 74Cantwell, 19Cantwell, D., 54Cantwell, D.P., 342, 345Capaldi, D., 402Capaldi, D.M., 358, 400Carlson, C.L., 144, 215, 379Caron, 19Carrion, V.G., 429Carroll, A., 190, 192, 193Carter, A.S., 229, 235, 276Carter, C.L., 229Cas-soano, M., 281Casat, C., 164Cashel, M., 126Caspi, A., 22Cassen, E.H., 417Castellanos, F.X., 384Cauce, A.M., 115Cavell, T.A., 407, 408Chaffin, M.J., 288Chambers, W., 258Chandler, L.A., 6, 7, 11Chorpita, B.F., 203, 204Christ, M.A.G., 381, 384Christ, M.G., 415, 417Christensen, A., 345Christian, R.C., 283Chung, K.M., 394Cicchetti, 306, 316, 320–322Cicchetti, D., 306, 385, 440Cicchetti, D.V., 46, 306, 316, 320–322, 439Cillessen, A.H.N., 212, 219Claiborn, C.D., 130Clarke, B., 142Clarkin, 62, 64Clerkin, S.M., 283, 284Coe, D.A., 441Cohen, J., 170, 343, 344Cohen, N.J., 16, 414

author index493Cohen, P., 76, 341, 342Coie, J.D., 214, 216–219, 221, 408Conners, C.K., 6, 104, 109, 110, 131–133, 138,139, 147–149, 170–174, 180, 181, 187,390–393, 413Conoley, J.C., 96Constantino, G., 237Cookus, B.A., 257, 263Coonrod, E.E., 440Cooper, D.K., 415, 419Coppotelli, H., 218Costa, N.M., 295Costa, P.T., 4, 62, 295Costantino, G., 76Costello, A.J., 265, 269Costello, E.J., 260, 269, 378, 417Costello, F.J., 380Courchesne, E, 438Cowan, C.P., 264Cowan, P.A., 264Coyne, J.C., 293Craighead, W.E., 49Crick, N.R., 401Crnic, K., 262Crooks, C.V., 278, 279Crouch, J.L., 288Crowley, T.J., 400Cuellar, L., 78Cumella, E.J., 126Cummings, E.M., 285Cummings, J.A., 11, 248, 249Cunningham, C.E., 384Curry, J.F., 49Cytryn, L., 49Dacey, C.M., 229, 235Dadds, M.R., 283, 285, 286, 295, 429Dahlstrom, W.G., 22Dana, R.H., 77, 78Dandreaux, D.M., 283Danforth, J.S., 184Daniel, R.C., 286Darling, N., 280David-Ferdon, C., 191Davies, M., 55, 267, 399Davies, P.T., 285Davis, W.W., 16de Bildt, A., 443, 451De Groot, A., 115de Groot, C., 122De Haene, L., 289De Los Reyes, A., 142, 143, 275, 336, 345DeBacker, I., 384DeBaryshe, B.D., 356Delaney, H.D., 79DelVeccio, W.F., 60Denney, C.B., 394Desombre, H., 440DeStefano, L., 137, 138, 316–318, 329Dickens, C., 403, 410DiGiulio, G., 232DiGiuseppe, R., 237DiLavore, P.C., 193, 441Dishion, T.J., 402DiSibio, M., 317DiStefano, C., 42Dodge, K.A., 193, 196, 197, 205, 214, 218,219, 221, 291, 306, 401, 408Doleys, D.M., 58Doll, E.A., 315, 316, 321–323, 334Dopke, C.A., 288Dornbusch, S., 280Dougherty, L.R., 142, 301Downey, G., 293Dozois, D.J.A., 23DuBois, P.H., 6–8, 10, 11Duke, M., 280Dulcan, M.K., 407Dumas, J.E., 291Dumenci, L., 113, 158, 285Dunsterville, E., 288DuPaul, G.J., 184, 382, 386, 393, 394Durbin, C., 22Earls, F., 343, 345Eaves, R.C., 441Edelbrock, C., 15, 16, 20, 184, 201, 262,269, 345, 346Edelbrock, C.S., 144, 255, 269Edelsohn, G., 416Edwards, B., 289Egger, H., 378Egger, H.I., 380Eiraldi, R., 54Eisenstadt, T.H., 286, 288Elbert, J.C., 19Elgar, F.J., 283, 285Elkins, D.E., 121, 122, 125Elliot, D.S., 345, 408Elliott, S., 30Elliott, S.N., 332–334, 345Ellis, J., 417, 427Ellis, M.L., 401Emery, R.E., 292Endicott, J., 254Erdberg, P., 232Erickson, D.B., 291Erickson, M.T, 271Erikson, E.H., 60Erkanli, A., 280, 417Ernst, M., 257, 263Eron, L.D., 221Essau, C.A., 285, 410, 411Evans, C.S., 276, 345

494 author indexExner, J.E., 10, 12, 49, 126, 227, 229–236,250, 251, 346Eyberg, S., 194, 286Eyberg, S.M., 186, 280, 286–288Eyde, L.D., 67, 68Eysenck, H.J., 15Fabiano, G.A., 43, 383Fagan, P.J., 62Fann, K.D., 289Faraone, S.V., 294, 340Farmer, A.E., 295Farrell, J.M., 283Fee, V.F., 441Feighner, J.P., 254Feinfield, K.A., 285Feldman, E., 196, 197Fendrich, M., 185Ferdinand, R., 24, 160Fergusson, D.M., 19, 400–402Fiala, S.E., 407Fick, A.C., 401Fields, S., 82Figueroa, R.A., 75Filipek, P.A., 437Finch, A.J., 229, 248Firestone, P., 294First, M.B., 16, 293, 296Fischer, M., 294, 383, 384Fisher, L., 401Fisher, P., 258, 265, 267, 400, 407Flanagan, R., 237Flannery-Schroeder, E., 281Flens, 291Fletcher, J.M., 316Fletcher, K., 294, 382, 383Fletcher, M., 273Foley, D.L., 347Follett, G.M., 43Forbey, J.D., 122Forehand, R., 218, 286, 292Forness, S.R., 18Forns, M., 115Forth, A.E., 411Fortin, L., 417Fowler, P., 282Fox, L.W., 273Frame, C.L., 216Frances, A.J., 16Francis, D.J., 316Francis, G., 6, 8, 295Franco, J.N., 78Fraser, J.A., 283Frazier, T.W., 440French, J.L., 7, 76Freud, S., 8, 11, 13Frick, P.J.,Friman, P.C., 115, 268Fuchs, D., 89, 217Fudge, H., 400Funderbunk, B., 286Gacono, C.B., 225Gadow, K.D., 146, 148, 166, 168, 169Gallucci, N.T., 126, 236Galton, F., 6, 18, 19Gao, H., 142Garb, H.N., 249Garber, G., 415Garber, J., 48Garbin, M.G., 296Garcia, M., 79Garland, A., 400Garmezy, N., 48, 58Garrido, M., 122Geisinger, K.F., 77, 78Gelfand, D.M., 192Geller, B., 273, 293Gent, C.L., 201Gerrard-Morris, A.E., 443Gerson, R., 305Ghiselli, E.E., 51, 61Giampino, T.L., 383Gibbon, M., 296Gilliam, J.E., 443, 446Ginsburg, G.S., 429Gittelman-Klein, R., 226, 236, 383Gizer, I.R., 384Glueck, E., 294Glueck, S., 294Glutting, J.J., 41Goldberg, L.R., 4Goldsmith, H.H., 5Goldstein, 54Goodlin-Jones, B., 316, 438Goodman, J.T., 277Goodman, S.H., 293Goodman, W.K., 432Gordon, M., 393Gorsuch, R.L., 39, 296Gotlib, I.H., 293Gotlieb, H., 16Gottesman, I.I., 295Goudena, P.P., 237Gould, M., 400Gould, M.S., 62, 344, 345Graham, J., 121Graham, J.R., 121Green, K., 218Green, M., 302, 313Green, S., 384Green, S.M., 96, 142, 222, 342, 344, 381, 410Greenspan, S., 318Gresham, F.M., 214, 217, 219, 223Gresham, R.M., 332–334Grills, A.E., 428

author index495Grimes, C.L., 219Grimm, J., 399Gross, L.M., 294Grubb, H.J., 295Gruber, C., 102, 136, 147, 174, 176, 177, 276, 391,Gruber, C.P., 75, 104, 133–138, 149, 178, 180, 276,352, 391, 405Guevremont, I.C., 382, 394Gulab, N., 420Gullone, E., 282Gutterman, E.M., 258Haak, R.A., 11, 227, 244, 245Hagan, M., 420Hale, R.L., 7, 76Haley, G.M.T., 415Hall, J., 154Hall, T.A., 443Hallmark, R., 226Halperin, J.M., 283Hamilton, C., 217, 220, 420Hammen, C., 204, 293, 417Hammer, D., 321Hammer, E.F., 248Hammond-Laurence, K., 429Hampson, S., 22Hand, C., 122Handel, R., 122Handel, R.W., 120, 130Handelsman, J.B., 82Handwerk, M.L., 115Hankin, B.L., 415, 417, 418Hann, D.M., 401Hare, R.D., 406, 410, 411Harold, G.T., 416Harrington, R.G., 43Harris, F.C., 190Harris, I.C., 78Harrison, P.L., 317, 321, 330, 331Hart, D.H., 243, 244, 246Hart, E.L., 25, 142. 161, 401Hartdagen, S., 401Hartley, S.L., 443Hartmann, D.P., 190, 192, 204Harvey, V.S., 365Hasegawa, C., 77Haskett, M.E., 288, 289, 291Hathaway, S.R., 12, 117, 118, 121, 296Hawes, D.J., 285Hawkins, R.P., 23Hayden, E., 22Hayes, S.C., 44, 317Haynes, 19Hays, S., 120Hayvren, M., 214Haz, A.M., 289Hedrick, M., 292Heffer, 142Heilman, N.E., 289Helenius, H., 115Hellinckx, W., 286Helms, 78Helstelae, L., 115Helzer, J.E., 295, 296Hembree-Kigin, T.L., 288Hendren, R.L., 384Henker, B., 384Herr, 416Herron, D.G., 77Hersen, M., 89, 196, 199, 203, 295Hicks, M.M., 126Hill, B.K., 320, 328Hill, J., 400Hiller, J.B., 229, 233Hilsenroth, M.J., 236Hinshaw, S.P., 63, 383, 384, 400Hodges, K., 255, 258, 261Hogan, K.L., 440Hojnoski, R.L., 226Holaday, M., 236, 243, 244Holden, 291Hollifield, J., 235Hollingsworth, D.K., 115Holmes, R., 203Holtzman, W.H., 231Hooks, M., 439Hoover, H.D., 30, 37, 341Horwood, L.J., 19, 401Houghton, S., 190Howell, C.T., 61, 63, 155, 212, 340Howlin, P., 439Hoza, B., 112, 212Huang, L.N., 77Huberty, C.J., 42Hudziak, J.J., 18Huesmann, L.R., 221Hughes, J., 217, 218Hughes, J.N., 408Hughes, T.L., 225Huizinga, D., 343, 408Hunsley, J., 23, 43, 44, 185, 186, 226, 227, 254,260, 359Hutcheson, J.J., 291Huth-Bocks, A., 420Hutt, M.L., 237Hyde, J., 282Hymel, S., 214, 217, 219, 223Hynd, C.W., 401Hynd, G.W., 154, 216Inclan, J.E., 77Isaacs, P., 394Ivanova, M.Y., 16, 113Jackson, M.F., 219Jackson, Y.K., 272

496 author indexJamieson, E., 294Janus, M.D., 122Jarrett, R.B., 44Jasso, R., 78Jensen, P.S., 86, 261, 262, 380Jersild, A.T., 203Jiang, X.L., 219Johnson, 288Johnston, 288Johnston, C., 384Joiner, T.E., 249Jung, C.J., 8Kaczmarek, T., 420Kagan, J., 51, 54Kaiser, S., 73–75Kamon, J., 285Kamphaus, R.W., 4, 6, 16, 18, 22, 23, 29, 30, 32, 36,39, 40, 42, 69, 73–76, 102, 103, 105–110, 145,146, 148, 150–155, 201–203, 277, 308, 315,317, 335, 346, 374, 390, 396, 413, 419Kane, D., 289Kaplan, D., 41, 44Karustis, J., 54Karver, M.S., 82, 341, 347Kasen, S., 76Kashani, J.H., 345Kashner, T.M., 260Kaslow, N.J., 191Kaufman, A.S.,358, 396Kaufman, N.L., 358Kaufman, S.H., 35, 249Kazdin, A.E., 190, 193, 259, 275, 295, 340, 347Keating, C.F., 195, 196Keenan, K., 61, 62, 294, 340, 400, 416Keenan-Miller, 416Keith, B., 277, 291Keith, T.Z., 41Keller, H.R., 191, 197, 199Keller, P.S., 285Kendall, P.C., 62, 64, 432Kennedy, C., 72Kerr, D., 24, 164Kerr-Almcida, N., 126Kershner, J., 16Kettler, R., 420Kicklighter, 321Kim, J.A., 439Kimonis, E., 87, 89, 378, 399, 410Kimonis, E.R., 283King, N.J., 61Klein, D., 22Klein, D.N., 39, 142, 226, 236, 260, 301, 303,414–417, 423, 424, 435Klein, R.G., 400Kleinmuntz, B., 3, 7, 9, 12, 14, 15, 22Klin, A., 440Kline, R.B., 130, 149, 159, 178, 180, 391Kine, S., 410Klinedinst, J.K., 174Knoff, H.M., 228, 245, 246, 250Knowles, E.S., 236Koenig, K., 440Kohler, S., 364Kolen, M.J., 30, 37, 341Konig, P.H., 383Konold, T.R., 340Koot, H.M., 115Koppitz, E.M., 11, 230, 245–249Kosson, D.S., 410Kovacs, M., 23, 38, 48, 255, 415, 418, 419Kratochwill, T.R., 203Krefetz, D., 420Krishnamurthy, R., 126, 128Kritzer, J., 18Kroll, L.S., 121Kroon, N., 237, 238Krug, D.A., 446Kruh, I, 294Kruse, S.J., 305Kubu, C.S., 440Kupersmidt, J.B., 216, 219Kuramoto, H., 115Kurkela, S.A., 116Kurtines, W., 79Lachar, D., 14, 75, 102, 105, 133–138, 141, 147,149, 174, 176–181, 276, 352, 391, 405LaCombe, J.A., 42LaFrenier, P.J., 291Lahey, B.B., 25, 53, 54, 56, 96, 142, 144, 161, 185,190, 215, 216, 222, 342, 344, 350, 378–381,383, 384, 386, 399, 401, 402, 410Lambert, M., 186, 187Lamborn, S., 280Lamparski, D., 203Landau, S., 207Lansing, M.D., 444Larzelere, R., 115Last, C.G., 429Le Couteur, A., 440LeBreton, J.M., 288Lee, S.S., 380Lefkowitz, M.M., 221, 222, 422Lega, L.I., 79Leland, H., 321Lenhart, L.A., 401Lett, N.J., 203Lewczyk, C.M., 268Lewinsohn, P.M., 342, 417Li, F., 402Lichtenberger, E.O., 358, 365, 369–371, 373Liebman, R., 273Lilienfeld, S.O., 273Lin-Yan, S., 281Lindsay, D.W., 408

author index497List-Kerz, M.L., 190Little, S.G., 214, 217, 219, 223Lochman, J.E., 219, 401, 408Locke, H.J., 292Loeber, R., 96, 142, 144, 222, 261, 342, 344, 345,381, 399–402, 410Loney, B.R., 61, 206, 294, 400–402, 406, 408Loney, J., 207, 380Long, N., 292Lonigan, C.J., 203, 204Lord, C., 193, 440–442Lorr, M., 15Lovejoy, M.C., 288Luby, J.L., 416Lucas, C.P., 407Lucia, V.C., 282Lucio, E., 122Luckasson, R., 316, 336Lundahl, B.W., 288Lunkenheimer, E., 24Lushene, R.E., 296Lynam, D.R., 5Lynch, K.D., 130Lyneham, H.J., 428Lynskey, M.T., 400Lyon, G.R., 316Maccoby, E.E., 289MacKenzie, E.P., 292MacMillan, H.L., 294Madsen, C.J., 398Madson, M.B., 305Magnusson, D., 19Magnusson, P., 441Malgady, R.G., 76, 237Malik, M., 23Manion, I., 289Mann, B.J., 292Mannuzza, S., 383March, J.S., 431Marchant, G.J., 137Marcotte, D., 417Marcus, L.M., 444Margolin, G., 320Marin, 78Marin, G., 79Markie-Dadds, C., 286Marks, D.J., 283Marsee, M.A., 402, 409Martin, L.S., 230, 231Martin, R.P., 3, 5, 13, 19, 22, 30, 32, 60, 61, 76, 79,83, 84, 230, 231, 250, 289Martinez, S.R., 79Martorell, B., 115Mash, E., 23, 43, 44, 185, 186, 254, 260, 359Mash, E.J., 23, 190, 191, 198, 384, 385Mason, B., 234, 346Massetti, G.M., 43, 383, 384Masuda, 78Mather, N., 358Matson, J.L., 441Matsumoto, 78Matthews, W., 226Mattison, R.E., 416Maujean, A., 283Mayes, L., 439Mayeux, L., 212, 214Mazefsky, C., 443McArthur, D.S., 227, 230, 236–240, 242McBurnett, K., 285, 380, 401McCallum, R., 120McCarney, S.B., 393McCarthy, 334McCarthy, L., 125McClaskey, C.L., 196, 197McConaughy, S.H., 61, 63, 89, 155, 161, 200, 201,212, 235, 340McConnell, S.R., 214, 216, 223McCoy, M.G., 401McCrae, R.R., 4, 62, 295McDonald, K.L., 219McGoldrick, M., 305McGrew, K.S., 329McGuffin, P., 295McKinley, J.C., 12, 117, 296McKnew, D.H., 49McMahon, R.J., 110, 273, 279, 292, 296, 399, 404,406–409McNeil, C.B., 23, 286, 288McReynolds, 15Measelle, J.R., 264Meehan, B.T., 408Meehl, 121Meehl, P, 12, 13Mendoza, R.H., 79Meredith, 78Mermelstein, R., 417Merrell, K.W., 332, 422Metevia, L., 394Metraux, R., 11Metzke, C.W., 115Meyer, G.J., 232Mikami, A.Y., 384Mikulich, S.K., 400Mildenberger, K., 440Milich, R., 196, 207Mille, 74Miller, 330Miller, L.C., 413Milling, L.S., 58Milliones, J., 78Milner, B., 441Milner, J.S., 288, 289Mischel, W., 60Moffitt, T.E., 60, 401–403Monuteaux, M.C., 294

498 author indexMoore, S.G., 216Moos, B.S., 280–282Moos, R.H., 280–282Moravec, B.C., 257, 263Morgan, C.J., 115Morgenthaler, W., 9, 10Morris, A.S., 403Morris, R.D., 316Morris, R.J., 203Morris, T.L., 23Morrison, R., 226Morsbach, S.K., 279Mullins, S.M., 288Munoz, L.C., 410Murray, H.A., 9, 228, 230, 237Murrie, D.C., 51Myers, K., 419, 421Myles, B.S., 446Nagle, R.J., 240, 243Needleman, G.L., 282Neeper, R., 215, 350, 386Negy, C., 136Neiberding, R., 226Nelles, W.B., 61Nelson, M.M., 280Nelson, R.O., 44Nesbitt-Daly, J.S., 286Newcomb, K., 286Newsom, C.R., 19Nezu, A.M., 347–349Nezu, C.M., 347–349Nezworski, M.T., 249Nigg, J.T., 384Nitko, A., 35Nixon, R.D.V., 291Norman, R.D., 79Norton, H., 164Noterdaeme, M., 440Nurnberger, J.I., Jr., 304O’Brien, 19, 63O’Brien, J.D., 258, 401O’Leary, K.D., 292O’Leary, S.G., 280, 286, 397Oakland, T., 321, 330, 331Odom, S.L., 214, 216, 223Offord, D.R., 59Okiishi, J., 187Olino, T.M., 22, 142, 301Ollendick, T.H., 61, 89, 142, 199, 260, 388, 414,427–433, 435Olmedo, E.L., 79Olson, S., 24Ondersma, S.J., 288, 294Onghena, P., 286Orvaschel, H., 293Osofsky, J.D., 401Ostrov, J.M., 195, 196Oswald, D., 443Owen, P.F., 225Ownby, R.L., 358Ozonoff, S., 316, 438, 439, 442–445, 451Paikoff, R.L., 346Pandina, G.J., 384Papadopoulos, L., 305Papillon, M., 417Pardini, D.A., 272Parker, J.G., 212, 213Patterson, D.R., 186Patterson, G.R., 206, 207, 272, 288, 358, 402, 409Payne, A.F., 10Pearson, K., 39Peeples, F., 401Pelham, W.E., 43, 260, 380, 383, 385, 388,393, 394, 397Perriello, L.M., 364Perrot, A., 439Perry, M.A., 94Peters, R.D.V., 273Petersen, A.C., 415Petersen, N.S., 30, 33, 37, 415Peterson, D.R., 16, 341, 413Peterson, M., 22Petit, G.S., 291Petoskey, M.D., 42Pettit, G.S., 306, 401Pfeiffer, 142Pfiffner, L.J., 285Phares, E.J., 89Phares, V., 62Piacentini, J., 144, 164, 261, 262, 267Piacentini, J.C., 170, 343, 344Pianta, R.C., 340Pickles, A., 400, 440Pina, A.A., 256, 430, 431Pincus, H.A., 16Pinderhughes, E.E., 291, 408Piotrowski, C., 19, 281, 282Piquero, A.R., 409Pisecco, S., 180Pliszka, S., 401Policaro, K., 283Pope, K.S., 128Popinga, M.R., 332Porter, B., 292Potvin, P., 417Powell, N., 219Power, T., 54Power, T.J., 393Pozdol, S.L., 440Pressman, L.J., 281, 282Prinstein, M.J., 216, 219Prinz, R.J., 278, 293Prinzie, P., 286

author index499Prior, M., 286Prusoff, B.A., 293Puig-Antich, J., 258, 354Putallaz, M., 219Quay, H.C., 49, 51, 52, 54, 377, 399Rabian, B., 419, 430, 435Rabin, A.I., 225, 227, 228Rafferty, J.E., 11, 244, 246Raine, A., 401Ramirez, M., 79Ramirez, V., 289Rapee, R.M., 295, 428, 431Rapee, R.R., 431Rapport, M.D., 386, 394Ray, R., 429Ray, R.S., 207Reed, M.L., 201Reich, W., 256, 343, 345Reichler, R.J., 441Reid, R., 393Reiss, A.L., 429Reitman, D., 286Rende, R., 304Renk, K., 62Renner, R.R., 441Rescorla, L.A., 103, 113, 114, 116, 146, 148,156, 158–160, 162, 163, 413Reynolds, C., 43Reynolds, C.R., 6, 16, 23, 43, 73, 102, 103,105–108, 110, 145, 146, 148, 150–155,201, 202, 277, 308, 335, 346, 388, 390,413, 421, 429, 430, 435Reynolds, W.M., 37, 74, 75, 414, 415, 419, 420Rezai, A., 440Rhode, P., 342Rhoades, K.A., 286Riccio, C., 43, 154Rice, F., 416Richardson, R.D., 327Richmond, 321Richmond, B.O., 388, 429, 430Richters, J.E., 261, 275, 276, 341Ricks, J.H., 372Ridenour, T.A., 137Rieser-Danner, L.A., 5Riggs, P.D., 400Risi, 193Risi, S., 441Rispens, J., 237Risser, S.D., 214Roberts, B.W., 60, 217Roberts, C., 329Roberts, C.R., 261Roberts, G.C., 207Roberts, G.E., 227, 230, 236Roberts, M.A., 207Roberts, R.E., 261Robertson, D., 282Robins, L.N., 254, 265, 296, 409Robinson, E.A., 194, 286–288Rockett, J.L., 51Rodell, J.L., 11Roesch, L., 417Roff, J.D., 401Rogers, A, 142Rogers, H., 286Rogers, R., 126Rogler, L.H., 237Roid, 201Ronan, 432Roper, B.L., 190Rorschach, H., 9–11, 20, 230Rosenblum, N.D., 203Ross, A.O., 272Ross, A.W., 186Rothlind, J., 222Rotter, J.B., 11, 244, 246Rounsaville, B.J., 19Roux, S., 439Rubian, B., 419Rudolph, K.D., 204Ruffalo, S.L., 333Rutter, M., 19, 48, 58, 193, 400–401Ryan, S., 295Saavedra, L.M, 430Saavedra, L.M., 430Saccuzzo, 44Saemundsen, E., 441Sanders, M.R., 286Sandoval, J., 74Sandstrom, M.J., 219Sanford, K., 282Sarason, 288Sargent, J., 273Sasagawa, S., 285, 410Sattler, J.M., 89Sawyer, M., 286Scahill, L., 18Schaughency, E., 142Schaughency, E.A., 222Schinka, J.A., 122Schmidt, K.L., 249Schniering, C.A., 431Scholte, E., 160Schopler, E., 441, 444Schorre, B.E.H., 185Schwab-Stone, M., 267, 276Schwab-Stone, M.E., 407Scopetta, M.A., 79Scott, S.S., 289Scotti, J.R., 23Seat, P.S., 174Seelen, J., 128

500 author indexSeeley, J.R., 342, 417Semrud-Clikeman, M., 418, 422Sevin, J.A., 441Shaevel, B., 122Shaffer, D., 184, 256, 267, 400Shaffer, D.M., 74, 254, 259, 265, 266, 268, 386, 407Shaffer, T.W., 232Shapiro, C.J., 378Shapiro, E.S., 189, 196–199Sharp, C., 410Shaw, J.G., 321Shellenberger, S., 305Shelton, 280Shelton, K.K., 283, 285Shelton, T.L., 394Sherrill, J.T., 255Sherry, 243Shiner, R., 22Shore, G., 394Siegel, 288Sigmund Freud, 8Sigurdardottir, S., 441Sikora, D.M., 160, 443Silver, M., 273Silverman, W.K., 61, 142, 256, 260, 419,427–433, 435Silverthorn, P., 103, 168, 259, 345Simpson, R.L., 446Sinclair, L., 440Singleton, L.C., 219, 223Siskind, G., 360Sitter, S., 440Sivaldason, N., 283Skinner, B.F., 5Skinner, C.H., 189Slesinger, D., 126Smallish, L., 294, 383Smari, J., 441Smith, B.H., 343Smith, G.L., 243Smith, M., 343, 378, 382, 384, 394Smith, S.R., 236, 378, 388Smolla, N., 257, 263Snidman, N., 51, 54Snyder, 288Snyder, M., 51Solis, M.L., 291Solomon, I.L., 237,Solomon, M., 316, 438Sourander, A., 115, 116South, M., 443Southam-Gerow, M.A., 203Spanier, G.B., 292Sparrow, 316Sparrow, S.S., 33, 316, 320–324Spencer, T., 142Spielberger, C.D., 296, 431Spitzer, R., 54Spitzer, R.I., 296Spitzer, R.L., 55, 254, 399Spracklen, K.M., 402Sprafkin, J., 146, 148, 166, 168, 169Stacy, A., 318Stacy, A.W., 336Stagheeza, B., 345Staghezza, B.M., 62Stanger, C., 235, 285Stanion, P., 305Starr, B.D., 237Stavrakaki, C., 417, 427Steele, R.G., 286Steer, R., 420Steer, R.A., 296Stein, L., 121Stein, L.A.R., 121Steinberg, L., 280Steinhausen, H.C., 115Stickle, T.R., 428Stone, S.V., 62Stone, W.L., 440Storoschuk, S., 440Stouthamer-Loeber, M., 222, 401Strauss, C.C., 63, 215–219, 429Stredny, R., 128Stredney, R.V., 229Streiner, D.L., 439Stringfield, S.A., 413Sullaway, M., 345Sundberg, N.D., 12Suveg, C., 281, 282Swain, N.R., 401Swartz, J.D., 231Sweeney, L., 291Swensen, C.H., 11Szakowski, A., 285Szapocznik, J., 79Szatmari, P., 317, 439Tallent, N., 358, 360–362, 367, 368Tanaka, J.S., 316Tanke, E., 51Tannock, R., 384Tapscott, M., 294Taylor, M., 190Teerikangas, O.M., 115Teglasi, H., 365, 369–372, 374Tellegen, A., 29, 30Teodori, J.B., 307Terdal, L.G., 190, 191, 198Tesiny, E.P., 222, 422Teufel-Shone, N., 281Thapar, A., 416Thomas, E.A., 234, 346Thompson, D.S., 316–318, 329Thompson, L.L., 400Thurber, S., 115Thurlow, M.L., 329Tibbetts, S.G., 409

author index501Tinsley, B.R., 223Toepfer, S.M., 122Toth, S.L., 306Touyz, S.W., 291Trautman, P., 400Tripp, G., 142Tsuang, M.T., 294Tully, L.A., 286Turner, L.M., 440Underwood, M.K., 214Updegraff, R., 216Urbina, S., 14, 21, 22, 25, 34, 37, 39, 40, 43, 45Uyteeborek, K., 289Vaillancourt, T., 219Valla, J., 255, 257, 263Valla, J.P., 263Van Berckelaer-Onnes, I., 160Van der Ploeg, J., 160Van Hasselt, V.B., 203Van Horn, M.L., 333, 334Vandvik, I.H., 185Vasay, M.W., 203, 204Vaughn, 154Verhulst, F.C., 39, 73, 115Vincent, G.M., 410Voelker, S.L., 321Volkmar, F., 439Volkmar, F.R., 440Vosk, B., 218Walder, L.O., 221Waldman, I.D., 384Walker, E.F., 385Walker, H.M., 58Walker, J.L., 221Walker, R., 11Wall, A.D., 126Wallace, K.M., 292Wallbrown, F., 358Walsh, C., 294Walthall, J.C., 340Wangby, M., 212Ward, C.S., 288Ward, J.H., 42Warner, V., 185, 293Waschbusch, D.A., 283, 400Watanabe, H.E., 261Watkins, C.E., 226Weatherman, R.E., 320, 328Webster-Stratton, 288Webster-Stratton, C., 408Wechsler, R.D., 396Weems, C.F., 428–431Weems, C.W., 295Wehrspann, W., 16Weiner, 227Weiner, I.B., 225, 230–232, 234–236Weiner, Z., 10, 12, 49Weiss, B., 31, 415Weissman, M.M., 185, 293, 304Weisz, J.R., 31Wells, K.C., 401Welsh-Allis, G., 293West, J., 190West, M.O., 293Westerman, M.A., 316Wewers, S., 401Whalen, C.K., 199Whalen, D.K., 384Whipple, 288White, S.F., 403, 410Whiteside, 334Whiteside-Mansell, L., 288, 291Wickramaratne, P., 293Widaman, K.F., 318, 336Widiger, T.A., 16, 17, 56, 166Wiener, J., 232, 362, 364Wilcutt, E., 380Williams, C.L., 119, 122, 126Williams, J.B.W., 296Wilson, F.J., 439Wimberley, R.C., 289Wingenfeld, S.A., 178, 180, 391Winokur, G., 418Winters, A., 419Winters, N.C., 402, 419, 422Wirt, R.D., 101, 174, 401Witt, 142Wolf, T.H., 3Wolfe, D.A., 278, 279, 289Wolff, L.S., 280Wood, J.M., 249Woodcock, R.W., 320, 325, 328, 330Woodworth, R.S., 7Wootton, J., 280, 294Wootton, J.M., 283Wrobel, T., 136Wrobel, T.A., 134Xing, Y., 261Ye, W., 293Yen, W.M., 21Yoerger, K., 409Young, J.G., 258Youngstrom, E.A., 440Zedeck, S., 51Zelli, A., 291Zeman, J., 255, 258, 261, 281Zera, M., 402Zettergren, P., 212, 219Zucker, R.A., 282Zuckerman, M., 77

S u b j e c t I n d e xAAID. See American association on intellectual anddevelopmental disabilities (AAID)Adaptive behavior assessment system-II (ABAS-II)administration and scoring, 329–330, 332content, 329interpretation, 329norming, 329reliability, 329strengths and weaknesses, 330–331validity, 329Adaptive behavior scalescharacteristicsadministration format, 321domains assess, 318–320informant choice, 320–321norm vs. criterion referencing, 320general recommendationsadaptive behavior assessment system-II(ABAS-II), 319–324scales of independent behavior, 324–326historyAAIDD criteria, 316definition, 316intelligence relationship, 317–318mental retardation, 316scales uses, 318social quotient, 316measuring social skillsdevelopment and revision, 332social skills rating system, 332strengths and weaknesses, 334omnibusrating scales component, 335referral reason, 325Vineland adaptive behavior scales, 2nd edition,322–325Administer diagnostic interviews, 264African American measures, 78Aggression, peer-referenced assessment, 221Agoraphobia, 428Alabama parenting questionnaire (APQ)content, 283–285norms, 285reliability and validity, 285American association on intellectual anddevelopmental disabilities (AAID), 316American psychological association (APA), 68–69Anxiety disordersassessment strategyalternative causes and treatment, 432–433classification and co-morbidity, 432–433questions and implications, 433screening process, 432–433case study, 434characteristics features, 428–429cognitive-behavioral assessment, 432definition, 428interviews and clinician ratings, 431–432specialized measuresfear survey schedule for children-revised(FSSCR), domains, 431revised children’s manifest anxiety scale(RCMAS), 429–431APQ. See Alabama parenting questionnaireAsian American measures, 78Asperger’s disorder, 446Assessor-administered exercise with unlimitedchoiceformat, 215Attention-deficit hyperactivity disorder (ADHD)alternative causes assessmentintensity and duration, symptom, 396medication side effects, 397primary and secondary dysfunction, 397assessment implicationscombined type, 385predominantly inattentive type, 385behavior rating scalesAchenbach system for empirically basedassessment, 389behavior assessment system for children–2ndedition, 390Conners rating scales–3rd edition, 390–392callous-unemotional traits assessment, 408–409classification and subtypeswith (ADD/H) and without hyperactivity(ADD/WO), 379conceptualization, 378–379core features, 378–379DSM-IV-TR, 380–383proposed behaviors, 384comorbiditiesconduct problems/aggression, 383learning disability, 383comorbid problems assessment503

504 subject IndexAttention-deficit hyperactivity disorder (ADHD)(Continued)family functioning, 396learning difficulties, 396conceptual modelbehavioral inhibition, 384environmental factors, 385neurological abnormalities, 384conduct problemsassessment implications, 404–406callous-unemotional traits, 403case study, 405–406causal factors assessment, 407–408classification and diagnosis, 398–399cognitive and learning style, 401comorbidities, 400, 407core features assessment, 406–407developmental pathways assessment,408–410ecological factors, 401onset subtype model, 402–403core behavior assessmentage and gender based comparison, 393close attention and scale selection, 388–389continuous performance test (CPT), 393diagnosis limitations, 394ecological validity, 395freedom from distractibility (FD) factor, 396multiple modalities, 393ratings scales, 389–393subtype differentiation, 389task performance, 394DISC-IV evaluation, 262special education placementindividuals with disabilities act (IDEA), 398requirement, 397–398Autism diagnostic interview-revised (ADI-R)assessment behaviors, 440strength, 440Autism diagnostic observation schedule (ADOS)assessment of, 441–442differential validity, 441modules, 441Autism spectrum disorders (ASD)and Asperger’s disorder,assessment strategyalternative cause rule out, 444co-morbidity, 444multidisciplinary assessment, 444questions and implications, 445screening process and classification, 444case studyacademic achievement, 448adaptive behavior, 448–449background information, 445–447behavioral observations, 447–448intellectual functioning analysis, 448rating scale assess, 449recommendations, 450referral information, 445visual motor integration, 448characteristics, 438–439co-morbidity, 439DSM-IV-TR criteriadefinition of, 437–438symptoms, 438functional analysis, 443specialized measuresautism diagnostic interview-revised (ADI-R), 440autism diagnostic observation schedule (ADOS),441–443childhood autism rating scale (CARS), 441Gilliam autism rating scale–2nd Edition(GARS-2), 443infant behavioral summarized evaluation(IBSE), 439–440parent interview for autism (PIA), 440BASC-2. See Behavior assessment system forchildren, 2nd edition (BASC-2)Behavioral avoidance tests (BATs), 203Behavior assessment system for children, 2nd edition(BASC-2)parent rating scaleadministration and scoring, 151content, 145interpretation, 152–153norming, 151reliability, 152strength and weakness, 153validity, 152self-report of personality (SRP)administration and scoring, 105age, 102interpretation, 110norming and reliability, 108–109scale content, 102–105strengths and weakness, 110, 112validity scales, 109teacher rating scaleadministration and scoring, 154composites and scales, 154content, 153–154interpretation, 155–156norming, 154reliability, 154–155strength and weakness, 156validity, 155Behavior problems, interpretation, 345Bias and cultural competenceconstruct validity, 74–75content validity, 73–74predictive validity, 75test, 73CAPI-second edition. See Child abuse potentialinventory (CAPI)-second editionCAT. See Children’s apperception test (CAT)

subject Index505CBCL-TRF. See Child behavior checklist- teacherreport form (CBCL-TRF)Child abuse and neglect assessment strategychild-rearing demands, parental responses,278–279general strengths and problem areas identify, 278identify needs of the child, 279parent–child relationship and abuse dynamics, 279suspected child abuse, 277Child abuse potential inventory (CAPI)-secondeditioncomposite abuse scale, 288–289norms, 289reliability and validity, 289Child behavior checklist (CBCL)parent reportadministration and scoring, 158–159content, 158interpretation, 160–161norming, 159reliability, 159–160sample case, 157strength and weakness, 161–162validity, 160teacher report form (TRF)administration and scoring, 162–163content, 162interpretation, 164norming, 163reliability, 163sample case, 165strength and weakness, 166validity, 163–164Child behavior checklist-teacher report form(CBCL-TRF), 333Child global assessment scale (CGAS), 184–185Childhood autism rating scale (CARS), 441Childhood problems, family contextaggression, social learning perspective, 272–273anxiety, an operant perspective, 272eating disorders, family systems perspective, 273Children’s apperception test (CAT), 11, 237Children’s depression inventory (CDI)administration and scoring, 418norming and reliability, 418–419scale content, 418strengths and weaknesses, 419validity, 419Children’s MMPI, 14Child symptom inventory-4 (CSI-4)parent and teacher report checklistsadministration and scoring, 167–168content, 167DSM-IV disorders, 167interpretation, 169–170norming, 168reliability, 168strength and weakness, 170validity, 168–169Classificationcommunication, 49–50dangers and limitations, 50–51documentation, 50evaluating systems, 51–52integration, medical and multivariate, 53–54levels, 48medical models, 52multivariate approaches, 52–53Commonalities across interview schedulescharacteristics, 256–257DISC-IV question format, 254Comorbidity, 62Confidence bands, 36–37Conflict negotiation task, 204–205Conners-3. See Conners, 3rd edition(Conners-3)Conners, 3rd edition (Conners-3)parent rating scaleadministration and scoring, 170interpretation, 172norming, 171reliability, 171scale content, 170–171strength and weakness, 172validity, 160self-report (Conners-3 SR), 131–133administration and scoring, 131–132interpretation, 133reliability, 132scale content, 131strengths and weakness, 133validity, 132–133teacher rating scaleadministration and scoring, 173content, 173interpretation, 174norming, 173reliability, 173strength and weakness, 174validity, 173–174Constructs vs. categories, 18–19Construct validity bias, 74–75Content validity, 37–38Contextual factors, 306Controversy surrounding projective techniquesprojection/behavioral sampledemand characteristics, 230hypothesis, 228, 230interpretations types, 230perceptual test, 230stimulus type selection, 230psychometric/clinical technique testapproaches and limitations inherent, 228clinical impressions, 228person’s unique qualities, 227reliability and validity, 227standardization efforts problem, 227traditional measurement theory, 227

506 subject IndexConvergent/discriminant validity, 39Criterion-related validityconcurrent and predictive, 38–39correlation, 39Data reanalyses, 17Depressionassessment strategyalternative causes, 425classification, 423comorbidity, 423questions and implications, 423screening process, 422treatment, 423case study, 424characteristics featureschronicity, 416duration, 416early and late onset, 416gender differences, 415heritability, 416prevalence rate, 415comorbidityacademic functioning, 417anxiety disorder, 417developmental level difference, 417internalizing disorders, 417–418medical illness, 417diagnostic criteriaimportance and inter-rater reliability, 415symptoms, 414internalizing disorders, 413–414peer-referenced assessment,specialized measureschildren’s depression inventory (CDI),418–419peer nomination inventory for depression(PNID), 422Reynolds adolescent depression scale, 2ndedition (RADS-2), 419–421Reynolds child depression scale (RCDS),421–422Developmental psychopathologycomorbidities, 62–64definition, 48Erikson’s psychosocial theory, 60meta-analysis of cross-informant correlation, 63norms, 58practical implications, 64process-oriented approach., 59situational stability, 61–62stability and continuity, 60–61Diagnostic and statistical manual of mental disorders(DSM)changes, 54–55children and adolescents, 56comparison, diagnostic criteria, 55Direct observation form (DOF), 200DISC-IV. See NIMH diagnostic interview schedulefor childrenDPICS. See Dyadic parent-child interaction codingsystem (DPICS)Drawing techniques. See Projective techniquesDSM-IV-TRdefinition, 360diagnosis parameterscross-situational consistency, 381–383developmental change in, 382diagnostic threshold, 380functioning impairment, 383onset and duration, symptoms, 381subtype stability, 380symptoms, 380–381Dyadic parent–child interaction coding system(DPICS)behavioral observation, 280categories, 287content, 287norms, 287reliability and validity, 287–288ECS. See Exner’s comprehensive system (ECS)Emic vs. etic perspectives, 76–77Empirical criterion keying, 13–14Enhance reliability, 84Erikson’s psychosocial theory, 60Ethics, peer-referenced assessmentassessor-administered exercise with unlimitedchoiceformat, 215later psychological adjustment, 213teacher-administered exercise with fixed-choiceformat, 215–216time-efficient relatively, 213Etiology, 301–303Evaluationdesigningcontexts screening, 87developmental considerations, 85–86enhance reliability, 84practical considerations, 87–88psychological domains, 86–87reasons, 88scientific approach, 85non-specifics, 81–82purpose of testing, 82–83referral problems and clarification, 82–84Event recording, 196–197Evidence-based interpretationmedical record, 72psychologists view, 70–71self monitoring questions, 71–72test standards, 69Exner’s comprehensive system (ECS)administration and scoring, 232age trends, 234evaluation, 236

subject Index507process measured, 231–232reliability, 233–234Rorschach interpretation, 231, 233–234validity and interpretations, 234–236Fairnessacculturation assessmentAfrican American measures, 78Hispanic American measures, 78–79emic vs. etic perspectives, 76–77guidelines, children assessment, 77–78symptom severity, 76Family context assessmentfunctioningconsiderations, 277, 279issues, 276–277history and differential diagnoses, 257–258influences, causal rolechildhood problems, 272–273psychological adjustment, 272interpreting information, provided by parentanxiety, 276child’s adjustment, 275depressed mothers, 275–276parental adjustment, 275marital conflictmeta-analysis, 291–292O’Leary-Porter scale (OPS), 292parental adjustmentADHD, 294–295antisocial behavior, 294anxiety, 295depression, 293family history assessment, 295–296schizophrenia, 295substance abuse, 293–294parenting stresschild abuse potential inventory (CAPI),288–289stress index-second edition (PSI), 289–291parenting styles and practicesAlabama parenting questionnaire (APQ),283–285child development, 280dyadic parent–child interaction coding system(DPICS), 286–288family environment scale-second edition (FES),280–282parenting scale (PS), 286–288social phobia, 274youth characteristics, 271Family environment scale (FES)-scale-secondeditioncontent, 281–282family environment scale, 281norms, 282reliability and validity, 282Family interaction coding system (FICS), 206Family psychiatric history, 303–305Field trials, 17Gaussian curve, 32Genograms symbols, 305Gilliam autism rating scale–2nd Edition(GARS-2), 443Hispanic American measures, 78–79History taking, child assessmentage of onset, 300content, 306–307contextual factors, 306course/prognosis, 300–301etiology, 301–303family history illustration, 304–305family psychiatric history, 303formats, 307–309genograms usage, 305nonverbal cues observation, 313previous assessment/treatment/intervention, 306role of, 299–300sample history form, 301HIT. See Holtzman inkblot technique (HIT)Holtzman inkblot technique (HIT), 231Hyperactivity, peer-referenced assessment, 221–222Individuals with disabilities education improvementact (IDEIA), 17–18, 20, 398Infant behavioral summarized evaluation (IBSE),autism, 439–440Information integrationcase study, behavior problems, 350, 351case study, language disorder and social phobia,351, 352clinical reasoningerrors, 348problem orientation, 348, 349combining rules, multiple sourcesinformation, 343cross-informant disagreementanalysis level, 340behavior types, 339different measurement techniques, 340situational variability, 340different informants use, 340discrepancies and problemsdifferential validity, 345differing motivations, 347emotional distress, 345parental adjustment and marital conflict, 347reliability, parent report, 346structured interviews and behavior ratingscales, 346multistep strategychild’s adjustment report, 347convergent findings, sources and methods,350–353critical information determination, 351–352

508 subject IndexInformation integration (Continued)discrepancy explanation, 352strengths and weaknesses report, 353Informed consent, 90Inkblot techniquesExner’s comprehensive system (ECS)administration and scoring, 232evaluation, 236process measured, 231–232reliability, 233–234validity and interpretations, 234–236Holtzman inkblot technique (HIT), 231Integrating assessment information. See InformationintegrationInternal consistency coefficients, 35, 175–176Internalizing and externalizing dimensions, 15–16Interpreting assessment information. SeeInformation integrationLanguage disorder, interpretation, 349–350Mean score differences, 73Measurement issues. See Personality testsMedical models, 52Minnesota multiphasic personality inventory(MMPI)original scales, 14personality inventory for children (PIC), 14–15rating scales, 15Minnesota multiphasic personality inventoryadolescent(MMPI-A)administration and scoring, 119–120case study, MMPI-A, 129–130interpretation, 128norming, 121–122reliability, 122–125scale content, 117–119strengths and weakness, 128validity, 125–127validity scales, 121MMPI-A. See Minnesota multiphasic personalityinventory-adolescent (MMPI-A)Multiple source information, combiningrules, 341Multivariate approaches, 52–53NIMH diagnostic interview schedule for children(DISC-IV)administration, 266–267development, 265reliability, 267–268structure and content, 265–266test–retest and internal consistencyestimates, 267validity, 268–269Normal curves. See Gaussian curveNorm developmentage-based, 32clinical, 32–34gender-based, 31sampling, 30–31Norm-referenced scores, 25Objective tests, 12ObservationsADHD, 190aggressive behaviorobservation study, 195preschool children, 195–196ASEBA, 200–201assessment techniques, 191basics of, 192BAT, 203–204behavioral assessment system, childrendefinition, 202clinical assessor, 191conflict negotiation task, 204–205data collectiondodge study, social competence, 197duration recording, 197event recording, 196–197momentary recording, 198time sampling, 197whole/partial interval, 198direct observation form (DOF), 200event recording, 196–197examples of, 200–207FICS, 206–207intrapsychic variables, 191limitations, 190naturalistic observations, 194, 195overt behavior, 191peer interaction, 205–206personal digital assistants (PDA), 198–199reactivity, 190samples, 201selection of observer, 172–174settings, 191, 196, 198simple antecedent-behavior-consequence(A-B-C), 193SOAPS, 207systemsanalog, 196analysis level, 192basics of, 192naturalistic observation, 194, 195several behavioral domains, 193simple A-B-C, 193subjective and idiosyncratic definition, 192target behavior, 192, 205, 206TAB, 206test observation form (TOF), 200total problems score, 201Observer selectionself monitoringadvantages of, 198

subject Index509PDA, 198–199steps in, 198Observing nonverbal cues, 313Obsessive-compulsive disorder (OCD)anxiety disorder, 429O’Leary-Porter scale (OPS), 292Omnibus personality inventories. See Self-reportinventoriesOPS. See O’Leary-Porter scale (OPS)Parental adjustmentADHD, 294–295antisocial behavior, 294anxiety, 295depression, 293family history assessment, 295–296schizophrenia, 295substance abuse, 293–294Parenting scale (PS)attitudes and beliefs about discipline, 286content, 286norms, 286reliability and validity, 286Parenting stress index (PSI)-second editioncontent, 289–291norms, 291reliability and validity, 291Parent interview for autism (PIA), 440Parent rating scaleschildren evaluation, 141–142factors that influence, 142omnibus scales, 144overview of, 144–145Parent report PIC-2, 174–177Peer nomination inventory for depression(PNID), 422Peer nominations, 214–217Peer-referenced assessmentclassroom participation levels, 220ethicsassessor-administered exercise with unlimitedchoiceformat, 215later psychological adjustment, 213teacher-administered exercise with fixed-choiceformat, 215–216time-efficient relatively, 213intuitive and empirical basis, 211sociometric techniques, 211types ofaggression, 221depression, 222–223hyperactivity, 221–222paired comparison technique, 223peer nominations, 214–217sociometrics, 217–221Peer relationships vs. later psychologicaladjustment, 213Percentile ranks, 29Personality assessmentdistinguishing features, 5history, 6projective techniquesassociation, 8–9for children, 11–12Rorschach’s approach, 9–10thematic apperception test (TAT), 9traits, 4Personality inventory for children-2 (PIC-2),174–177Personality inventory for children (PIC), 14Personality inventory for youth (PIY)abbreviated form, 135administration and scoring, 134–135interpretation, 136–137norming, 135reliability, 135–136scale content, 133–134strengths and weakness, 137–138validity scales, 134, 136Personality testsdefinition, 22diagnostic schedulesidentification, 22limitation, 23rating scales, 23Post-traumatic stress disorder (PTSD), 428Predictive validity bias, 75Process-oriented approach, 59, 61Projective techniquesfor children, 11–12drawing techniquesdraw-a-person technique, 246house–tree–person (HPT), 248–249kinetic family drawing, 249Koppitz emotional indicators, 247nonverbal, 245psychometric cautions, 249–250inkblot techniquesExner’s comprehensive system (ECS),231–236Holtzman inkblot technique (HIT), 231projection/behavioral sampledemand characteristics, 230hypothesis, 230interpretation types, 230perceptual test, 230stimulus type selection, 230projective controversy, 226projective debate, 226psychometric testapproaches and limitations inherent, 228clinical impressions, 228person’s unique qualities, 228reliability and validity, 227standardization efforts problem, 227traditional measurement theory, 227

510 subject IndexProjective techniques (Continued)Rorschach’s approach, 9–10sentence completion techniques (SCT)administration, 244content, 243–244evaluation, 245interpretation, 244–245interpretive approaches, 246thematic apperception test (TAT), 9thematic story-telling techniquesgeneral interpretation, 237–238roberts apperception technique, 238–243word association, 8PSI-second edition. See Parenting stress index(PSI)-second editionPurpose of testing, 82–83Rapport buildingexplaining testingchild, 93–94parent, 94–96teachers, 96importance and recommendations, 89informed consent, 90–91strategies, 95RATC. See Roberts apperception test for children(RATC)Rating scales, 15Raw score, 25Readability, 43Referral reasonbackground information, 325behavioral observations, 326previous evaluation, 326test results and interpretationadaptive behavior, 323communication, 323daily living skills, 323psychoeducational assessment, 326Reliabilityaffecting variables, 35internal consistency coefficients, 35reliable specific variance, 35–36standard error of measurement (SEM), 36test–retest method, 34Reliable specific variance, 35–36Report writingabbreviations and acronyms, 365adaptation, audience and setting, 368difficult word reduction, 366editing, 366evidence-based assessment, 357examples usage, 366grammar and spelling check, 368heading and list usage, 367instruments description, 366length reduction, 367, 368oral communication, resultschild feedback, 375parent conferences, 373, 374teacher conferences, 374pertinent information, 365pitfallsconsumer’s view of, 362failure, address referral questions, 362faulty interpretation, 361number obsession, 361report length, 361vocabulary problems, 360problems and challengesassessment result presentation, 357improper interpretation, 358score check, 368sectionsassessment procedures, 369assessment results and interpretation, 370background information, 369behavioral observations, 370diagnostic considerations, 371information identification, 369psychometric summary, 372recommendations, 371referral questions, 369self-test, 372summary and signatures, 371word—number emphasize, 365Revised children’s manifest anxiety scale (RCMAS)administration and scoring, 430norming and reliability, 430scale content, 429–430strengths and weaknesses, 431validity, 430Reynolds adolescent depression scale, 2nd edition(RADS-2)administration and scoring, 420norming and reliability, 420scale content, 420strengths and weaknesses, 421validity, 420Reynolds child depression scale (RCDS)administration and scoring, 421norming and reliability, 421scale content, 421strengths and weaknesses, 422validity, 421–422Roberts apperception test for children (RATC),238–240Rorschach’s approach, 9–10Sample impairment-oriented scaleschild global assessment scale (CGAS), 184–185home situations questionnaire (HSQ), 184school situations questionnaire (SSQ), 184Scales of independent behavior-R (SIB-R)administration and scoring, 329content, 328

subject Index511interpretation, 329–331norming, 329reliability, 329strengths and weaknesses, 330validity, 329School apperception method, 237Scoresnormal distribution, 28, 29norm-referenced, 25percentile ranks, 29raw, 25standardconversion table, 26–27T-scores, 28–29uniform T-scores, 29–30Self-report inventoriesConners, 3rd edition, self-report (Conners-3 SR)administration and scoring, 131–132case sample, 129–130interpretation, 133reliability, 132scale content, 131strengths and weakness, 133validity, 132–133Minnesota multiphasic personality inventoryadolescent(MMPI-A)administration and scoring, 119–120interpretation, 128norming, 121–122reliability, 122–125scale content, 117–119strengths and weakness, 128validity, 125–126validity scales, 121omnibus personality, 101–102overview of, 103–104personality inventory for children (PIC), 101personality inventory for youth (PIY)abbreviated form, 135administration and scoring, 134–135interpretation, 136–137norming, 135reliability, 135–136scale content, 133–134strengths and weakness, 137–138validity scales, 134, 136self-report of personality (SRP), 102–112single construct personality, 138youth self-report (YSR)administration and scoring, 113–114interpretation, 116norming, 114reliability, 114scale content, 113strengths and weakness, 116validity, 114–116Self-report of personality (SRP)administration and scoring, 105age, 103interpretation, 110norming, 108reliability, 108–109scale content, 102–105strengths and weakness, 110–112validity scales, 105–107Sensitivity and specificity, 42Sentence completion techniques (SCT), 243–245administration, 244content, 243–244evaluation, 245interpretation, 244–245interpretive approaches, 246SIB-R. See Scales of independent behavior-R(SIB-R)Simple antecedent-behavior-consequence(A-B-C), 192Situational stability, 61–62Social competence, dodge study, 197Social phobia, 428Social phobia, interpretation, 351–352Social skills rating system (SSRS)administration and scoring, 329content, 330norming, 331reliability, 331validity, 329Sociometric nominations social statusdetermination, 218SSRS. See Social skills rating system (SSRS)Standard error of measurement (SEM), 36StandardsAmerican psychological association (APA), 68–69ethics and self-examination checklist, 68evidence-based interpretationmedical record, 72psychologists view, 70–71self monitoring questions, 71test standards, 69psychological assessment practice, 67use and misuse, 68Storytelling approach. See Thematic story-tellingtechniquesStructured diagnostic interviewscommonalities across interview schedulescharacteristics, 255DISC-IV question format, 254computer-administered versions, 255evaluationadvantages, 259–260disadvantages, 260–261K-SADS, 258, 259NIMH diagnostic interview schedule for children(DISC-IV)administration, 266–267development, 265reliability, 267–268

512 subject IndexStructured diagnostic interviews (Continued)structure and content, 265–266test–retest and internal consistency estimates, 267validity, 268–269organization and content, 266recommendationsadminister diagnostic interviews, 264attention-deficit hyperactivity disorder, 262pictorial interviews, 263–264symptoms, 264strengths and weaknesses, 260variation across interview schedules, 258–259verbal instructions, 266Structured observation of academic and play settings(SOAPS), 207Student behavior survey (SBS), 178–182parent report PIC-2, 174, 176–177teacher reportadministration and scoring, 178–179content, 178interpretation, 181norms, 179reliability, 179–180sample case, 182–183strength and weakness, 181–182validity, 180–181TAT. See Thematic apperception technique (TAT)Teacher-administered exercise with fixed-choiceformat, 215–216Teacher rating scaleschildren evaluation, 143factors that influence, 144overview of, 144–145Teacher report form (TRF), 162–164Tell-me-a-story technique (TEMAS), 237TEMAS. See Tell-me-a-story technique(TEMAS)Test bias, 73Test observation form (TOF), 201–202individual standardised test, 201narrow band scale, 201Test–retest and internal consistency estimates, 267Test–retest method, 34Test standards, 69Thematic apperception technique (TAT), 237Thematic story-telling techniquesgeneral interpretation, 237–238Roberts apperception techniqueadministration and scoring, 238–240content, 238evaluation, 243indicators, 241–242norming, 240reliability, 240stimulus cards depictions, 239validity and interpretations, 240tell-me-a-story technique (TEMAS), 237thematic apperception technique (TAT), 237Traits, 4T-scores, 26–29Uniform T-scores, 29–30Validityconstruct, 37content, 37–38convergent/discriminant, 39criterion-relatedconcurrent, 38correlations, 39predictive, 37–38factor analysiscluster, 41–42confirmatory factor, 40–41MMPI-A, 40threatsguarding against, 44readability and response set, 43utility, 44–45Variation across interview schedules, major sources,258–259Vineland adaptive behavior scales-2. See alsoAdaptive behavior scalesadaptive behavior construct, 322administration and scoring, 323content, 322–323norming, 323reliability, 324strengths and weaknesses, 324–325validity, 324Whine, 194Woodworth personal data sheet, 7Word association techniques, 8–9Youth self-report (YSR), 113–116administration and scoring, 113–114interpretation, 116norming, 114reliability, 114scale content, 117strengths and weakness, 116validity, 114–116