RESEARCH METHOD COHEN ok
RESEARCH METHOD COHEN ok RESEARCH METHOD COHEN ok
TRUE EXPERIMENTAL DESIGNS 279 textbooks/9780415368780 – Chapter 13, file 13.8. ppt). So, for example, the designs might be: Experimental 1 RO 1 X 1 O 2 Experimental 2 RO 3 X 2 O 4 Control RO 5 O 6 This can be extended to the post-test control and experimental group design and the post-test two experimental groups design, and the pretest-posttest two treatment design. The matched pairs design As the name suggests, here participants are allocated to control and experimental groups randomly, but the basis of the allocation is that one member of the control group is matched to a member of the experimental group on the several independent variables considered important for the study (e.g. those independent variables that are considered to have an influence on the dependent variable, such as sex, age, ability). So, first, pairs of participants are selected who are matched in terms of the independent variable under consideration (e.g. whose scores on a particular measure are the same or similar), and then each of the pair is randomly assigned to the control or experimental group. Randomization takes place at the pair rather than the group level. Although, as its name suggests, this ensures effective matching of control and experimental groups, in practice it may not be easy to find sufficiently close matching, particularly in a field experiment, although finding such a close match in a field experiment may increase the control of the experiment considerably. Matched pairs designs are useful if the researcher cannot be certain that individual differences will not obscure treatment effects, as it enables these individual differences to be controlled. Borg and Gall (1979: 547) set out a useful series of steps in the planning and conduct of an experiment: 1 Carryoutameasureofthedependentvariable. 2 Assign participants to matched pairs, based on the scores and measures established from Step 1. 3 Randomly assign one person from each pair to the control group and the other to the experimental group. 4 Administer the experimental treatment/ intervention to the experimental group and, if appropriate, a placebo to the control group. Ensure that the control group is not subject to the intervention. 5 Carryoutameasureofthedependentvariable with both groups and compare/measure them in order to determine the effect and its size on the dependent variable. Borg and Gall indicate that difficulties arise in the close matching of the sample of the control and experimental groups. This involves careful identification of the variables on which the matching must take place. Borg and Gall (1979: 547) suggest that matching on a number of variables that correlate with the dependent variable is more likely to reduce errors than matching on a single variable. The problem, of course, is that the greater the number of variables that have to be matched, the harder it is actually to find the sample of people who are matched. Hence the balance must be struck between having too few variables such that error can occur, and having so many variables that it is impossible to draw a sample. Instead of matched pairs, random allocation is possible, and this is discussed below. Mitchell and Jolley (1988: 103) pose three important questions that researchers need to consider when comparing two groups: Are the two groups equal at the commencement of the experiment Would the two groups have grown apart naturally, regardless of the intervention To what extent has initial measurement error of the two groups been a contributory factor in differences between scores Borg and Gall (1979) draw attention to the need to specify the degree of exactitude (or variance) of the match. For example, if the subjects were to be matched on, say, linguistic ability as measured in a standardized test, it is important to define the Chapter 13
280 EXPERIMENTS AND META-ANALYSIS limits of variability that will be used to define the matching (e.g. ± 3 points). As before, the greater the degree of precision in the matching here, the closer will be the match, but the greater the degree of precision the harder it will be to find an exactly matched sample. One way of addressing this issue is to place all the subjects in rank order on the basis of the scores or measures of the dependent variable. Then the first two subjects become one matched pair (which one is allocated to the control group and which to the experimental group is done randomly, e.g. by tossing a coin), the next two subjects become the next matched pair, then the next two subjects become the next matched pair, and so on until the sample is drawn. Here the loss of precision is counterbalanced by the avoidance of the loss of subjects. The alternative to matching that has been discussed earlier in the chapter is randomization. Smith (1991: 215) suggests that matching is most widely used in quasi-experimental and nonexperimental research, and is a far inferior means of ruling out alternative causal explanations than randomization. The factorial design In an experiment there may be two or more independent variables acting on the dependent variable. For example, performance in an examination may be a consequence of availability of resources (independent variable one: limited availability, moderate availability, high availability) and motivation for the subject studied (independent variable two: little motivation, moderate motivation, high motivation). Each independent variable is studied at each of its levels (in the example here it is three levels for each independent variable) (see http://www.routledge.com/ textbooks/9780415368780 – Chapter 13, file 13.9. ppt). Participants are randomly assigned to groups that cover all the possible combinations of levels of each independent variable, as shown in the model. INDEPEN- DENT VARIABLE Availability of resources Motivation for the subject studied LEVEL ONE Limited availability (1) Little motivation (4) LEVEL TWO Moderate availability (2) Moderate motivation (5) LEVEL THREE High availability (3) High motivation (6) Here the possible combinations are: 1 + 4, 1 + 5, 1 + 6, 2 + 4, 2 + 5, 2 + 6, 3 + 4, 3 + 5and 3 + 6. This yields 9 groups (3 × 3combinations). Pretests and post-tests or post-tests only can be conducted. It might show, for example, that limited availability of resources and little motivation had a statistically significant influence on examination performance, whereas moderate and high availability of resources did not, or that high availability and high motivation had a statistically significant effect on performance, whereas high motivation and limited availability did not, and so on. This example assumes that there are the same number of levels for each independent variable; this may not be the case. One variable may have, say, two levels, another three levels, and another four levels. Here the possible combinations are 2 × 3 × 4 = 24 levels and, therefore, 24 experimental groups. One can see that factorial designs quickly generate several groups of participants. A common example is a 2 × 2design,inwhichtwo independent variables each have two values (i.e. four groups). Here experimental group 1 receives the intervention with independent variable 1 at level 1 and independent variable 2 at level 1; experimental group 2 receives the intervention with independent variable 1 at level 1 and independent variable 2 at level 2; experimental group 3 receives the intervention with independent variable 1 at level 2 and independent variable 2 at level 1; experimental group 4 receives the intervention with independent variable 1 at level 2 and independent variable 2 at level 2.
- Page 248 and 249: INTERNET-BASED SURVEYS 229 instruc
- Page 250 and 251: INTERNET-BASED SURVEYS 231 Box 10.1
- Page 252 and 253: INTERNET-BASED SURVEYS 233 Box 10.1
- Page 254 and 255: INTERNET-BASED SURVEYS 235 Box 10.1
- Page 256 and 257: INTERNET-BASED SURVEYS 237 Witte et
- Page 258 and 259: INTERNET-BASED EXPERIMENTS 239 requ
- Page 260 and 261: INTERNET-BASED INTERVIEWS 241 ‘ne
- Page 262 and 263: SEARCHING FOR RESEARCH MATERIALS ON
- Page 264 and 265: COMPUTER SIMULATIONS 245 autho
- Page 266 and 267: COMPUTER SIMULATIONS 247 computer s
- Page 268 and 269: COMPUTER SIMULATIONS 249 On the oth
- Page 270 and 271: GEOGRAPHICAL INFORMATION SYSTEMS 25
- Page 272 and 273: 11 Case studies What is a case stud
- Page 274 and 275: WHAT IS A CASE STUDY 255 (providing
- Page 276 and 277: WHAT IS A CASE STUDY 257 argue that
- Page 278 and 279: EXAMPLES OF KINDS OF CASE STUDY 259
- Page 280 and 281: PLANNING A CASE STUDY 261 accounts
- Page 282 and 283: CONCLUSION 263 In the narrativ
- Page 284 and 285: CO-RELATIONAL AND CRITERION GROUPS
- Page 286 and 287: CHARACTERISTICS OF EX POST FACTO RE
- Page 288 and 289: DESIGNING AN EX POST FACTO INVESTIG
- Page 290 and 291: PROCEDURES IN EX POST FACTO RESEARC
- Page 292 and 293: INTRODUCTION 273 Box 13.1 Independe
- Page 294 and 295: TRUE EXPERIMENTAL DESIGNS 275 motor
- Page 296 and 297: TRUE EXPERIMENTAL DESIGNS 277 2 Sub
- Page 300 and 301: TRUE EXPERIMENTAL DESIGNS 281 Facto
- Page 302 and 303: A QUASI-EXPERIMENTAL DESIGN: THE NO
- Page 304 and 305: PROCEDURES IN CONDUCTING EXPERIMENT
- Page 306 and 307: EXAMPLES FROM EDUCATIONAL RESEARCH
- Page 308 and 309: EVIDENCE-BASED EDUCATIONAL RESEARCH
- Page 310 and 311: EVIDENCE-BASED EDUCATIONAL RESEARCH
- Page 312 and 313: EVIDENCE-BASED EDUCATIONAL RESEARCH
- Page 314 and 315: EVIDENCE-BASED EDUCATIONAL RESEARCH
- Page 316 and 317: 14 Action research Introduction Act
- Page 318 and 319: PRINCIPLES AND CHARACTERISTICS OF A
- Page 320 and 321: PRINCIPLES AND CHARACTERISTICS OF A
- Page 322 and 323: ACTION RESEARCH AS CRITICAL PRAXIS
- Page 324 and 325: PROCEDURES FOR ACTION RESEARCH 305
- Page 326 and 327: PROCEDURES FOR ACTION RESEARCH 307
- Page 328 and 329: PROCEDURES FOR ACTION RESEARCH 309
- Page 330 and 331: SOME PRACTICAL AND THEORETICAL MATT
- Page 332: CONCLUSION 313 3 Actionresearchreso
- Page 336 and 337: 15 Questionnaires Introduction The
- Page 338 and 339: APPROACHING THE PLANNING OF A QUEST
- Page 340 and 341: TYPES OF QUESTIONNAIRE ITEMS 321 If
- Page 342 and 343: TYPES OF QUESTIONNAIRE ITEMS 323 de
- Page 344 and 345: TYPES OF QUESTIONNAIRE ITEMS 325 Ra
- Page 346 and 347: TYPES OF QUESTIONNAIRE ITEMS 327 Ve
TRUE EXPERIMENTAL DESIGNS 279<br />
textbo<strong>ok</strong>s/9780415368780 – Chapter 13, file 13.8.<br />
ppt). So, for example, the designs might be:<br />
Experimental 1 RO 1 X 1 O 2<br />
Experimental 2 RO 3 X 2 O 4<br />
Control RO 5 O 6<br />
This can be extended to the post-test control and<br />
experimental group design and the post-test two<br />
experimental groups design, and the pretest-posttest<br />
two treatment design.<br />
The matched pairs design<br />
As the name suggests, here participants are<br />
allocated to control and experimental groups<br />
randomly, but the basis of the allocation is that<br />
one member of the control group is matched to a<br />
member of the experimental group on the several<br />
independent variables considered important for<br />
the study (e.g. those independent variables that are<br />
considered to have an influence on the dependent<br />
variable, such as sex, age, ability). So, first, pairs of<br />
participants are selected who are matched in terms<br />
of the independent variable under consideration<br />
(e.g. whose scores on a particular measure are the<br />
same or similar), and then each of the pair is<br />
randomly assigned to the control or experimental<br />
group. Randomization takes place at the pair<br />
rather than the group level. Although, as its name<br />
suggests, this ensures effective matching of control<br />
and experimental groups, in practice it may not be<br />
easy to find sufficiently close matching, particularly<br />
in a field experiment, although finding such a<br />
close match in a field experiment may increase the<br />
control of the experiment considerably. Matched<br />
pairs designs are useful if the researcher cannot be<br />
certain that individual differences will not obscure<br />
treatment effects, as it enables these individual<br />
differences to be controlled.<br />
Borg and Gall (1979: 547) set out a useful<br />
series of steps in the planning and conduct of an<br />
experiment:<br />
1 Carryoutameasureofthedependentvariable.<br />
2 Assign participants to matched pairs, based<br />
on the scores and measures established from<br />
Step 1.<br />
3 Randomly assign one person from each pair<br />
to the control group and the other to the<br />
experimental group.<br />
4 Administer the experimental treatment/<br />
intervention to the experimental group and,<br />
if appropriate, a placebo to the control group.<br />
Ensure that the control group is not subject<br />
to the intervention.<br />
5 Carryoutameasureofthedependentvariable<br />
with both groups and compare/measure them<br />
in order to determine the effect and its size on<br />
the dependent variable.<br />
Borg and Gall indicate that difficulties arise<br />
in the close matching of the sample of the<br />
control and experimental groups. This involves<br />
careful identification of the variables on which<br />
the matching must take place. Borg and Gall<br />
(1979: 547) suggest that matching on a number<br />
of variables that correlate with the dependent<br />
variable is more likely to reduce errors than<br />
matching on a single variable. The problem, of<br />
course, is that the greater the number of variables<br />
that have to be matched, the harder it is actually<br />
to find the sample of people who are matched.<br />
Hence the balance must be struck between having<br />
too few variables such that error can occur, and<br />
having so many variables that it is impossible<br />
to draw a sample. Instead of matched pairs,<br />
random allocation is possible, and this is discussed<br />
below.<br />
Mitchell and Jolley (1988: 103) pose three<br />
important questions that researchers need to<br />
consider when comparing two groups:<br />
<br />
<br />
<br />
Are the two groups equal at the commencement<br />
of the experiment<br />
Would the two groups have grown apart<br />
naturally, regardless of the intervention<br />
To what extent has initial measurement error<br />
of the two groups been a contributory factor in<br />
differences between scores<br />
Borg and Gall (1979) draw attention to the need<br />
to specify the degree of exactitude (or variance)<br />
of the match. For example, if the subjects were to<br />
be matched on, say, linguistic ability as measured<br />
in a standardized test, it is important to define the<br />
Chapter 13