MAN USCRIPT REPORTS - the IDRC Digital Library - International ...
MAN USCRIPT REPORTS - the IDRC Digital Library - International ...
MAN USCRIPT REPORTS - the IDRC Digital Library - International ...
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
ARC HIV<br />
50881<br />
<strong>International</strong> Development Research C entré<br />
<strong>MAN</strong> <strong>USCRIPT</strong> <strong>REPORTS</strong><br />
Attitudes of Primary School<br />
Children in Ghana<br />
Robert A. Ntumi<br />
February 1983<br />
:tR1OOt4i
The <strong>International</strong> Development Research Centre is a public corporation<br />
created by <strong>the</strong> Parliament of Canada in 1970 to support research designed to<br />
adapt science and technology to <strong>the</strong> needs of developing countries. The<br />
Centre's activity is concentrated in five sectors: agriculture, food and nutrition<br />
sciences; health sciences; information sciences; social sciences; and<br />
communications. <strong>IDRC</strong> is financed solely by <strong>the</strong> Government of Canada; its<br />
policies, however, are set by an international Board of Governors. The<br />
Centre's headquarters are in Ottawa, Canada. Regional offices are located in<br />
Africa, Asia, Latin America, and <strong>the</strong> Middle East.<br />
<strong>IDRC</strong> Manuscript Reports<br />
This series includes meeting documents, internal reports, and preliminary<br />
technical documents that may later form <strong>the</strong> basis of a formal publication.<br />
Manuscript Reports are given a small distribution to a highly specialized<br />
audience.
ATTITUDES OF PRIMARY SCHOOL<br />
CHILDREN IN GHANAs<br />
Family and school determinants of<br />
children's attitudes<br />
Robert A. Ntumi (Director)<br />
Curriculum Research and Development Division,<br />
Ghana Education Service, Accra<br />
<strong>IDRC</strong>iV1R72e
CHAPTER 1 THE STUDY<br />
Introduction 1<br />
Review of Earlier Studies 4<br />
Significance of <strong>the</strong> Study 7<br />
Research Design 8<br />
CHAPTER 2<br />
CHAPTER 3<br />
CONTENTS<br />
OVERALL SCHOOL EFFECTS<br />
Introduction and Descriptive Statistics 18<br />
Scale Graphs 20<br />
Multiple Classification Procedures 29<br />
Summary 32<br />
CAUSAL INTERRELATIONS<br />
Introduction 34<br />
The Influence of <strong>the</strong> Environment 34<br />
Description and Identification of Paths 40<br />
Summary 56<br />
CHAPTER 4 MAJOR FINDINGS<br />
Conclusions 57<br />
Discussion 59<br />
APPENDICES I. Ano<strong>the</strong>r View on Analytical Strategy 62<br />
Reliability and Validity of <strong>the</strong> Scales 68<br />
Prescribed Values and Attitudes Scale. 79<br />
School Resources Index Form 91<br />
REFERENCES 95
P EEFACE<br />
The research was made possible by gr&its fim ilophone Nest Africaji<br />
eg±onai Educational Research Consortium (A:ARRC), wiich i funded by Ford<br />
Founc.ation (U.S.A.) id <strong>International</strong> Development Research Centre (<strong>IDRC</strong>)<br />
1.fl Jtt awa, Ce.nwia.<br />
The study looks at how G.hanaia primary school children aCqUlX'e values<br />
and at t itucie s in two major environments, namely, <strong>the</strong> home arid school The<br />
ror focussed on <strong>the</strong> contribution made by each envirenment a a block and<br />
alSo studied <strong>the</strong> components of each block. In reporting <strong>the</strong> results we<br />
were conscious of <strong>the</strong> need of yourx researchers who wish that research<br />
should report, in some detai.l, ox. methodolor and preferences underlying<br />
'<strong>the</strong> resee'ch. This is exemplifd in chapter 3 which discusses <strong>the</strong> way<br />
in which we measured values and att itude constructs. The structure piposes<br />
a structural equat:iun rnociel which considers <strong>the</strong> <strong>the</strong>ory which guided <strong>the</strong><br />
study and a sttisticaJ, apiroach. Our readers can <strong>the</strong>refore skip chapter 3<br />
wthuut missing much.<br />
iii<br />
Robert A. Ntunil
Acknovledgelneflts<br />
Many people were involved in <strong>the</strong> study and I must begin with & expreB-<br />
sion of' ere gratitude to those Assistant Dfrectors who allowed field<br />
officers in Kpandu and Ho duoatiofl Districts in Ghana, to assist in<br />
interviewing <strong>the</strong> children. I am grateful to Dr. R lazbiaon,. Dr. 1. lUng,<br />
Dr. S. Siffer, and Dr. J. Brown for <strong>the</strong> advice and encouragement <strong>the</strong>y<br />
gave us to carry <strong>the</strong> project tWOUZe Many thanks are due to Dr. '1". Wilhiens.,<br />
Mr. Ken Ross and Dr. i. Rosier who read. through <strong>the</strong> initial version of <strong>the</strong><br />
study and offered valuable suestions and eapecally Dr. Nelly Stromquist<br />
who edited this report. Ghris Robinson helped us to get acCess to his<br />
Institution's computer. Finally ny thanks to Mr. B. K. Osei, ?viss Kuzna ildred<br />
fld Mr. Aning-G'yamti who are members of Curi'iculuni Division, Ghana education<br />
Service.<br />
iv
17 V (I\ (I (17<br />
O1JJ2 '1<br />
rr I:1<br />
. .J.<br />
In hana <strong>the</strong> school as a najor educational re'orm arfency. fiae<br />
:U2ai5fl people expect <strong>the</strong> uchoi not only to equip <strong>the</strong> younviith shills<br />
SUCil as literacy nd nuraeracy lut also to teach <strong>the</strong>m <strong>the</strong> triChtt attitudes<br />
to life: attitudes that are required fOr <strong>the</strong> community's succesafnJ. drive<br />
toward <strong>the</strong> post-traditional era.<br />
Many studies have indicated that, apart from <strong>the</strong> school, o<strong>the</strong>r<br />
fan ors shape <strong>the</strong> attitudes of children. Adult interaction with children<br />
in <strong>the</strong> home, <strong>the</strong> structural dimension o <strong>the</strong> home (for exanpie, aatcrial<br />
;osseasions of t:ie home), and aireiital education constitute <strong>the</strong> raa.jor<br />
determinants of sack charactcristic as work habits, akxaitjon to tasks,<br />
and. qualities of need achievement. Some researchers would jive ;reater<br />
vic;ht to parental interaction with <strong>the</strong> child at home than economic level<br />
o: <strong>the</strong> pa::ents, <strong>the</strong>ir level of education, or o<strong>the</strong>r status characteristics<br />
(L:aj oribonks 1 74, Willians 1 P74). All this means that tie hone and<br />
school are-ho powerful environmon Us for <strong>the</strong> development of children's<br />
caracceristles.<br />
Ly coincidence, hana's hew S'ucturc and Content of ducatioai<br />
liofrma.le, 1 P74, emphasized child i'esa'inr environments and clearly out-<br />
lined what <strong>the</strong> government exaccaed from it schools. These are to:<br />
1. Jeclou in <strong>the</strong> c:iild an appreciation of <strong>the</strong> need<br />
for che;e and adaptation;
help pupils appreciate <strong>the</strong> importance o± co-<br />
operation and tolerance and <strong>the</strong> interdependence<br />
of people of different nations and cultures;<br />
develop th pupils <strong>the</strong> habit of asking questions<br />
and a keeimess to find tiiinis out for <strong>the</strong>mselves;<br />
provide opportunities that will predispose pupils<br />
to acquire <strong>the</strong> lnowlede, skills, and prevocational<br />
experiences that will enable <strong>the</strong>m to discover <strong>the</strong>ir<br />
aptitudes and potentialities and to develop a desire<br />
for fur<strong>the</strong>r iiripr ovolient;<br />
help pupils appreciate <strong>the</strong> dinity of work and<br />
interdependence of all workers.<br />
The overnrnent's acceptance of this mandate required chanes<br />
in oxauination procedures. Unfortunately, educational evaluation in<br />
QTharna continues on a traditional basis at all pro-university levels;<br />
that is, assessment of student academic achievement alone. The<br />
examinations tap mainly <strong>the</strong> acquisition of factual haowlede, plaeini a<br />
hiGh premium on selection criteria at <strong>the</strong> expense of attitudes and value<br />
outcomes of schooliu, so that very little attention seems to be Given<br />
to those who fail to satisfy <strong>the</strong> selection criteria. iTow that fresh<br />
policy deiands have been made in <strong>the</strong> :Tow Structure and Content of<br />
Education, it i.c left to <strong>the</strong> researcher to identifj research areas Lom<br />
<strong>the</strong>se policy statements.<br />
This study, <strong>the</strong>refore, proposes to investigate <strong>the</strong> afective<br />
process as it relates to Ghctnaien children in both <strong>the</strong> rural and urban
I<br />
areas. t is belabourin <strong>the</strong> point that this aspect of education has<br />
been nc;lected in Ghma even thouifh <strong>the</strong> literature on this topic La<br />
abundant in affluent sociotic.<br />
-<br />
The study reported hero was carried out at <strong>the</strong> prinory ociiool<br />
lovel. T±e aecision to conauc uo atudy a htut level viac bacu on<br />
coveral factors. Jirst, it is <strong>the</strong> only level oP for il education that<br />
caa bc effectively enforced under <strong>the</strong> cxistin law of itdiJ A:;i aoiuiy<br />
1c)uc&f 10: (19). 'Secondly, for noct children, prade six is terminal.<br />
In <strong>the</strong> 1977 - 70 school year only 40 per cent o:Li <strong>the</strong> total frade six<br />
population vms reiatered for <strong>the</strong> first year of <strong>the</strong> :Iiddle School. Per<br />
this reason, studLri and understandiwp <strong>the</strong> process of how tue school<br />
inparts valuas end attitudes at this level will allow us to make infe-<br />
rences nlout 7 to 11 year old bos and irlo whose fornml education<br />
torithatcs at this point. Thirdly, hlooia (194) su,;:;ects that <strong>the</strong> h03;le<br />
onviru:ucn is woot influential at <strong>the</strong> tiiae of <strong>the</strong> citiLJ)c preateut<br />
duvuloinnt, nd <strong>the</strong> pre-school years are ouch a period. Thouiph this is<br />
true, <strong>the</strong> :te of transition from pri.mary to oecdary schoolinp also<br />
ioses problems for arniy children. Ideally, all levels should be studied,<br />
hut <strong>the</strong> constraints ci' funds and personnel limit <strong>the</strong> scope. This study<br />
<strong>the</strong>n explores how primary school children acquire <strong>the</strong> values and attJ.-<br />
tudes <strong>the</strong>y brinep to school and <strong>the</strong> extent to which <strong>the</strong>se value a sad.<br />
attitudes have been attained by <strong>the</strong> children; in o<strong>the</strong>r words, <strong>the</strong><br />
:.nPluuncc of <strong>the</strong> home Pu2'<strong>the</strong>raore, <strong>the</strong> study investiGates values sad<br />
atitLwes pupils brirv to <strong>the</strong> secondary school, that is, <strong>the</strong> influence<br />
of <strong>the</strong> primary school. Sjnce <strong>the</strong>se environments are not autually<br />
exclusive, it is necessary to establisa causal lmiks between une home,<br />
school, and attitude outcomes.
om what we 1aiow of our society and from statements on <strong>the</strong> role<br />
of schools in <strong>the</strong> process of inculcating values and. attitudes we realise<br />
-that our idea of "post-taditiona1 man" and Inkeles and Smitht s concept<br />
o <strong>the</strong> 'Uodern man' have several Lac-tors in common. Inkeles and Smith,<br />
(1974) described such a man as having <strong>the</strong> followin characteristics:<br />
1 Openness to new experience.<br />
Readiness or social chsn;e.<br />
Growth of opinion.<br />
Confidence in one's ability and <strong>the</strong> ability of<br />
<strong>the</strong> human being to achi.eve objectives.<br />
Belief that o<strong>the</strong>r people and institutions can<br />
be relied upon to fulfil <strong>the</strong>ir obligations and<br />
responsibilities.<br />
Valwing of technical skills.<br />
Respect for <strong>the</strong> diçnity of o<strong>the</strong>r people.<br />
3. Positive attitude towards participation in<br />
public affairs.<br />
9. Positive attitude toward work.<br />
iPTI! O ARLr.ER TU1)LS<br />
Our main objective in. reviewing <strong>the</strong> literature on home and school<br />
as cnild.-rearing environments is simply to identify those studies that<br />
have moved from predictive approaches to explanatory variables in <strong>the</strong>ir<br />
ef1oits to find out how environments affect schooling- outcomes in <strong>the</strong><br />
conitivc and affective areas. With this as bacrourid, we can <strong>the</strong>n<br />
develop a rationale for using a causal model in a study of values and<br />
attitudes outcomes of schooling in Ghana.<br />
Ierner (1956), Rogers (1969), and Dobb (1960) conceptualized
5<br />
<strong>the</strong> modern peroom in terms of certain variables ViLaLch <strong>the</strong>y claimed were<br />
I1C1UfU1 in p.'ojectin OdUcatiollal Lrations and accomplishments. 1a1il<br />
(1968) an1 especially Inkeles and Sjth (1974) went fur<strong>the</strong>r to lormu-<br />
late end tap his main characteristics. Inotrthnents for this latter studj<br />
formed <strong>the</strong> basis of <strong>the</strong> OIl scale (Short 'orn). lven though <strong>the</strong> Inkcles<br />
end Sith study assumed cumulative ro\Itla of <strong>the</strong> modern man and drew<br />
attention to <strong>the</strong> rate of chan:e over time in <strong>the</strong> acquisition of those<br />
attrizes, <strong>the</strong> investigation concerned itself with adults, not school<br />
cmildaen. Our intention, however, is to study school cin.ldren.<br />
ln'riroaiaontal correlates ol' lea nini<br />
iu'ray (1930) identified on environment accordin to <strong>the</strong> kinds<br />
of benekits or harm that it jrovided for individuals, lie arued that<br />
individuals ei<strong>the</strong>r avoided harmful environment or defended <strong>the</strong>mselves<br />
a;a:Lnst it. Jut when <strong>the</strong> environment had poontial benefits, individuals<br />
avvroachcni it arid interacted vrith it. iiurny dosL;nated this as <strong>the</strong><br />
.: t:e ovironi ient", fur<strong>the</strong>r distin8uiohed between alpha press<br />
ov Lie e:ri:Lroinent (<strong>the</strong> press that actually o::ists) end <strong>the</strong> beta press,<br />
ouch as ti:c iridividual's interpretation of tIe pmnomenon that he peroives.<br />
Several studies have since been made to find, out <strong>the</strong> direct relationship<br />
bet:een measures of <strong>the</strong> alpha press of <strong>the</strong> family environment and children's<br />
eojiitivc performance and affective cimirac tcritics.<br />
Alon.; <strong>the</strong> lines of :Iurray' s environmental press model, Bloom's<br />
discuscaon (1b4) 01 staoility and eiane in nunnn choracterisics<br />
assumes that an educational environment is influenced in part by <strong>the</strong><br />
society in which it is located, vLth <strong>the</strong> individual placed in one of<br />
several distinct environmental situations. 3loom (1964) defined <strong>the</strong>
-6-<br />
environmental situation as b(:Ln physical and social, as well as intel-<br />
lectual, and considered that all hcse formed a network of environmental<br />
forces that impinged upon <strong>the</strong> individual and influenced him. Ie<br />
suested that it was important to identify types of educational environ-<br />
nouts or sub-environments o± press variables related to <strong>the</strong> characteris-<br />
tics of <strong>the</strong> individual.<br />
ThZtouh research, Bloom's colleaue, <strong>the</strong> 'Chicago School' iden-<br />
-bified three such environmental situations as <strong>the</strong> home, school, and peer<br />
group. They vent fur<strong>the</strong>r and examined <strong>the</strong> environmental correlates of<br />
children's co,nitive and affective measures. Anong <strong>the</strong> 'Chicago School'<br />
Dave (1963) and Wolf (1964) oporationalized some of <strong>the</strong>se environmental<br />
press variables of <strong>the</strong> family and investigated <strong>the</strong> relationship between<br />
<strong>the</strong>se and acadnic achievement and intelligence respectively.<br />
Two studies (ieeves 1972, LjoribanLsI$78b) are very compre-<br />
hencive in scope and different in analytical approach in examining rela-<br />
tions between family and school environments and cognitive and affective<br />
measures. Keeves investigated <strong>the</strong> environmental correlates of children's<br />
s:ionce and ma<strong>the</strong>matics achievement and <strong>the</strong>ir attitudes towards science<br />
and ma<strong>the</strong>matics tLve attitudinal and four process variables were looked<br />
at in <strong>the</strong> family environment schedule. 1eeves (172) split family<br />
environment into three powerful dimensions, namely, structural, attitu-<br />
di.nal, and process dimensions In general, <strong>the</strong> results showed that when<br />
<strong>the</strong> attitudinal and process variables were combined, <strong>the</strong>y had moderate<br />
relations with ma<strong>the</strong>matics and science and lowser validities in relation<br />
to <strong>the</strong> attitude scores. A research on £'vri ly and school environments<br />
and affective characteristics wm conducted by LIaoribanks (1977,<br />
1978a, 1978b). In this study, he related family and school charac-
teristics to measures of children's school related affective oharacterIa-<br />
tics. He not only obtained a measure of <strong>the</strong> intensity of exposure to <strong>the</strong><br />
fRm ly environment, but also attempted to ain a measure of <strong>the</strong> cumulative<br />
nature of <strong>the</strong> environment durin <strong>the</strong> interview schedule.<br />
om <strong>the</strong> evidence of <strong>the</strong> studies cited in this report, <strong>the</strong>re is<br />
little doubt That differences in <strong>the</strong> environmental press of <strong>the</strong> home and<br />
school can be identified and linked with cbi.ldrents diference in both<br />
cognitive and affective outcomes. The onvironnental press studies have,<br />
however, been criticised for failin; to be ecplicit on whe<strong>the</strong>r <strong>the</strong>y viere<br />
simply predictin <strong>the</strong> variations ±11 children's SChOO1IW2 outcomes or<br />
whe<strong>the</strong>r <strong>the</strong>- uere attemptin also to e::piain vJiç children varied in <strong>the</strong>ir<br />
traits. In our view this criticism is not entLrelr correct because<br />
ribanJ:s (17O : and hooves (1972 : 134) uaLc tile causal (explana-<br />
tori-) notine of <strong>the</strong>ir models omlicit. The attribution v;as also apparent<br />
in t::e work of Wolf (1964), who considered environmental press in his<br />
nodal as causes of variation in cPil6ron's abilities. The implications<br />
o: <strong>the</strong> hit from prediction to causal ezplannton is Os manor interest<br />
to our study. If treater u2-lderstcndin is to be accjuired of <strong>the</strong> causes<br />
of different levels of educational outcomes between different social<br />
roups, <strong>the</strong>n environmental press characteristics should correspond with<br />
actual phenomena within family and school environments. This is to be<br />
proferrct to predictive studies, since it is entirely possible that<br />
iearmnless composite variables ua predict achievement well and may be<br />
included in <strong>the</strong> model.<br />
i-hL?ICAiCE OP TIf STUDY<br />
l-rovious investigations elsewhere have dealt with linka'es bctueen
<strong>the</strong> c1±ld' a educational environments and its con.itive, connate, and<br />
aftective traits. Efforts have also been made to identify <strong>the</strong> important<br />
dimensions of <strong>the</strong> family environments and to estimate <strong>the</strong>ir influence.<br />
Unfortunatoly very little work has been done in Ghana to assess home and<br />
uchool environments in torma Of <strong>the</strong> more sOphisticated env:Lronmontaj.<br />
indices sieated by Bloom (i 9G4).<br />
Ano<strong>the</strong>r important aspect of this study lies in <strong>the</strong> approaches<br />
adopted to assess <strong>the</strong> home and school environments and to consider <strong>the</strong>ir<br />
relationehip to <strong>the</strong> outcomes of education. A critical period occurs when<br />
pupils are about to leave school ariu must decide whe<strong>the</strong>r to enter wor: or<br />
continue with full-time education. It is likely <strong>the</strong>n that at this sta;e<br />
both school and hoxie onvironments will have i'ofound influence on <strong>the</strong><br />
choices that are made. Causal relationships between powerful ccaponents<br />
of <strong>the</strong> home and school will <strong>the</strong>refore be o value to <strong>the</strong> parent, <strong>the</strong><br />
teacher, and <strong>the</strong> adxinistz'ator as <strong>the</strong>y seek to provide <strong>the</strong> most favourable<br />
conditions for <strong>the</strong> growth and development of ti-ic children in <strong>the</strong>ir charge.<br />
In conclusion, this investigation aims to point to some ol' <strong>the</strong><br />
forces operating in <strong>the</strong> home and school, and to give some indication of<br />
<strong>the</strong>ir relative importance at <strong>the</strong> primary school level. This will pave<br />
<strong>the</strong> way for o<strong>the</strong>r studies at o<strong>the</strong>r levels.<br />
iSJMRCH I)E8IG<br />
1eoeerch Problem<br />
7e now formulate our basic research problem t1us:<br />
To investigate <strong>the</strong> nature and strength of home and school<br />
ezimtal oharautoriutics that relate to attitudes of<br />
chilcren in Ghana.
Research Hypo<strong>the</strong>sis<br />
9-<br />
Prom <strong>the</strong> problem outlined above <strong>the</strong>re are several hypoi<strong>the</strong>as that could<br />
form <strong>the</strong> basis of a study into <strong>the</strong> educational environments of Ghanaian<br />
homes and schools:<br />
Hypo<strong>the</strong>sis 1 - Measures of structural variables of <strong>the</strong> home<br />
and school environment8 will be related to<br />
attitudinal outcomes of education:<br />
Ilaterial possessions of <strong>the</strong> home<br />
Parental education<br />
School resources or facilities<br />
School mansrjement<br />
School location<br />
Hypo<strong>the</strong>sis 2 - Measures of <strong>the</strong> level of schooling achieved,<br />
that is, number of years of sohoolin*, will<br />
be related to measures of attitudinal outcomes.<br />
Bach hypo<strong>the</strong>sis Wil]. be tested by decomposition o± effects in a statis-.<br />
tica]. model.<br />
jears of schooling and grade level are <strong>the</strong> same since at <strong>the</strong> elementary<br />
school level <strong>the</strong> government does not encourage children to repeat grades.
Research Plan<br />
- 10 -<br />
Our research plan requires a teat o value and attitude chane adrinia-<br />
terod uurin a si;le academi.c year to roups o ciii.luren enrolled in<br />
speciiicd primary school grades. ho suao teat will be administered also<br />
to oroups o±' children o <strong>the</strong> same a;es who have never been to school.<br />
Children who have never been to choo1 may serve as control groups to<br />
those enrolled nd will allow us to conclude with some con±'idence that<br />
any real differences between schoolers and non-schoolera may be ati-<br />
buted to <strong>the</strong> effects 0±' schooling.<br />
Variables<br />
The variables chosen or this study can be 'rouped into <strong>the</strong><br />
Jo1:Loiin cateories;<br />
Dependent Variables:<br />
The dependent variables are he v1ues and attitudes<br />
that <strong>the</strong> school is e:cpected to produce in <strong>the</strong> children.<br />
They are <strong>the</strong> child's overall score on a scale ol' desired<br />
values and attitudes, and also <strong>the</strong> child's score on <strong>the</strong><br />
subseales measuring speci±'ied components 0±' <strong>the</strong> values<br />
and attitudes that are sotv;ht by <strong>the</strong> society,<br />
attitude toward i'emales.<br />
Independent Variables:<br />
Amount o±' sohoolinr, i.e., grade or number<br />
ol' years 0±' exposure to schoqliar<br />
School resources
School size<br />
School's reliiouz tone<br />
School location<br />
c. Intez'venin Variables:<br />
Age<br />
sex composition<br />
Material possessions (cjuasisocio-ecouomlc status)<br />
Urba4/ruraJ. origin<br />
Sex<br />
Parental education<br />
The fundamental assumption that school experience produces<br />
value said attitude development implies that <strong>the</strong> more one is exposed to<br />
learning experience, <strong>the</strong> more he internalizes <strong>the</strong> content of <strong>the</strong> expe-<br />
rience. To test this assumption, <strong>the</strong> arialrsie of <strong>the</strong> impact of schooling<br />
on <strong>the</strong> acquisition of values md attitudes must also take into accotuit<br />
<strong>the</strong> intensity of schooling. The student's grade was assumed to reflect<br />
<strong>the</strong> level and intensity of schooling.<br />
School characteristics that were likely to have significant<br />
impact on <strong>the</strong> child's acquisition of values and attitudes were considered<br />
as independent variables. We assumed that <strong>the</strong> extent to which <strong>the</strong> hone<br />
envirment differed from <strong>the</strong> school environment would have a certain<br />
effect on <strong>the</strong> development of <strong>the</strong> child. Also, <strong>the</strong> extent to which one<br />
school environment differed from <strong>the</strong> envirment of ano<strong>the</strong>r school was<br />
expected to result in a differential impact of schools on <strong>the</strong> acqui.s,tion<br />
of values and attitudes. uzlity of building, library, recreational, and<br />
toilet facilities were used as indicators to measure school resources.
12 -<br />
Additional indopencLent var:Lcbioo vicre <strong>the</strong> cciiool's ainc, ccx<br />
COL1OsitiOfl, and location. The oie of <strong>the</strong> øbol wan expected to affect<br />
<strong>the</strong> acquisition of values and attitudes because <strong>the</strong> degree of interaction<br />
itbi.n small schools was considered -to be stronger than in larfe schools<br />
and thus likely to produce a more integrnted environment.<br />
LIan,y of <strong>the</strong> schools in <strong>the</strong> otudy vore operated by religious<br />
oruiizutions of different faiths, which used <strong>the</strong> schools to make con-<br />
verts for <strong>the</strong>ir respective faiths. The school's sex composition also<br />
had it unique contributions One of <strong>the</strong> correlates of learning was<br />
location of school; differences in location implied existence of diffe-<br />
rences in demographic and socio-cconouic perame ters of <strong>the</strong> school tone.<br />
We investigated ano<strong>the</strong>r set of variables which mediated <strong>the</strong><br />
impact of <strong>the</strong> independent variables on <strong>the</strong> locxning values. These we:e<br />
<strong>the</strong> student's quasi socio-economic status, urban-rural origin, and reli-<br />
cion. Ilaturation or age we regarded as not dependent on formal educatidn<br />
and tr hng. Age 11aJ seen ra<strong>the</strong>r as <strong>the</strong> natural process vihich helped<br />
tlie efficacy of educational strategies. Por <strong>the</strong> some reasons we included<br />
cluasi socio-economic status (QSLS), sex, urban-rural origin and religion.<br />
(SS was operationalized in <strong>the</strong> same way we operationalizec). school<br />
resources; that is, by taking an inventory of items suggestive of <strong>the</strong><br />
relative affluence of <strong>the</strong> child's hone, which enabled us to construct<br />
an index of Q,$$. Vie used school resources and Q$lZ indices -to estimate<br />
<strong>the</strong> extent to wftLch <strong>the</strong> child' a home differed from <strong>the</strong> eTchool:.jn terma:<br />
of omv:ronmental quality.<br />
Instruiriefltation and scale development<br />
Two instruments were adiinistered: -<strong>the</strong> values and attitudes
- 13 -<br />
interview schedule and <strong>the</strong> school resources index form. The former was<br />
used to collect data on values and attitudinal outcomes of schoolin as<br />
maitifested in primary schools, as well as among non-schoolers The<br />
school resources index was used. o take inventory of <strong>the</strong> resources ava:Ll-<br />
able in <strong>the</strong> school for <strong>the</strong> purpose of constructing an index of school<br />
quality, which was one of <strong>the</strong> independent or predictor variables Por<br />
each of <strong>the</strong> six blocks of indentified variables, a suanary measure was<br />
formed to portray <strong>the</strong> underlying property of <strong>the</strong> block.<br />
The creation of aggregate indicators of unobserved variables<br />
from many observed measures presents several problems. Some of <strong>the</strong> issues<br />
associated with composite variable formation have been investigated by<br />
Rosier (1978), Keeves (1971), and Williams (1974). Rosier derived<br />
weights from multiple regression analysis. Keevos used rosettes as<br />
carriers of regression. The latter approach involved two stases of scale<br />
refinement. During <strong>the</strong> first stage, items associated with <strong>the</strong> criterion<br />
were examined with principal component procedures in order to select items<br />
that discriminated effectively and were also consistent within <strong>the</strong> set.<br />
In step two, standardized regression weights were determined and used to<br />
calculate scores on <strong>the</strong> variable that would be strongly related to <strong>the</strong><br />
criterion measures when allowance had been made for o<strong>the</strong>r variables in<br />
<strong>the</strong> set. Thus <strong>the</strong> weightE or rosettes derived to carry <strong>the</strong> regression<br />
relationships (for instance, five measures on a given block) would be coia-<br />
puted as follows:<br />
where:<br />
= b1 A1 + b2 A2 + b3 A3 + b4 A4 + b A5<br />
b = weight (standardized)<br />
A = item value
- 14 -<br />
Unfortunately, premature collapsing on some resonse categories<br />
of <strong>the</strong> children put constraints on <strong>the</strong> use of <strong>the</strong> two techniques described<br />
above in forming composite variables for this study. In such circum-<br />
otnees, we used a simple approach in constructing OUT composite variables.<br />
It us take <strong>the</strong>m in order. :erdir; measures for <strong>the</strong> overall attitude<br />
scale and eight subscales, we adopted a sinple additive index in which all<br />
items were added toge<strong>the</strong>r to create <strong>the</strong> attitude index. This means that<br />
each of <strong>the</strong> contributing items was intically coded to contain <strong>the</strong> sane<br />
number of categories. A similar additive index was obtained for parental<br />
education with an ordinal level of Lleasurement In regard to Material<br />
Possessions of <strong>the</strong> home and school Structural Characteristics, each<br />
variable was divided by <strong>the</strong> number 0f categories it contained. Religion<br />
and School Banagement fall under <strong>the</strong> nominal level of measurement and<br />
<strong>the</strong>iefore a dummy variable coding procedure was used. Weighting presented<br />
no serious problems. It entaild scor:Lng alternative answers to each<br />
question in such a way that tdesi:ablet and 'undesirable' attitudes uere<br />
assined 1 and 2, respectively. The question of introducing a precise cut-<br />
ting point to a scale conceived as representing a continum can hardly be<br />
overlooked. But as I stated earlier, information was too quickly collapsed<br />
and thin could have presented problems for scale reliability.<br />
soale Reliability and Validity<br />
Before <strong>the</strong> start of this study concern was expressed about using<br />
tie sane scale for two distinct age groups (school children and adults).<br />
Critics have also pointed out that since <strong>the</strong> scale was used about eight<br />
years ago, and we are now in a period of rapid social change, scale<br />
reliability may be affected.
- 15 -<br />
There i also <strong>the</strong> problem of validity. Attitudes are abstractions<br />
from peoples responses. If social values fluctuate, <strong>the</strong>n <strong>the</strong> constructs<br />
we are trying to tap are unstable and <strong>the</strong> scales mar measure different<br />
constructs.<br />
In view of <strong>the</strong>se problems, principal component analyses without<br />
iteration and varimax rotations were used. ion <strong>the</strong> analyses we realised<br />
that sonic of <strong>the</strong> subscales should be refined to tal:e into account sugges-<br />
tions made elsewhere in this report. Details of this analysis are reported<br />
in Appendix 2.<br />
Sampling Procedures<br />
A total of 200 school children were interviewed. Porty-five non.-<br />
schoolers of ages comparable to those of <strong>the</strong> school children were also<br />
interviewed. The 200 school children were drav from 10 schools selected<br />
frc: urban and rural areas The criteria applied in stratifying schools<br />
for <strong>the</strong> purpose of selection showed how school location, school size, sex<br />
composition of school children, and school man.igement authority were<br />
operationalized to ensure that each specified category was adequately<br />
represented in <strong>the</strong> sample. The total sample of school children consisted<br />
of 25 children selected at random from each of <strong>the</strong> four grades (2, 3, 4<br />
and 6) included in <strong>the</strong> sample. Since all <strong>the</strong> schools viere mixed, a iro-<br />
portionate :uber of boys and girls viere randomly selected from emmch<br />
class. In <strong>the</strong> cmse of nan-schoolors, c1ii1don who viere briiii<br />
stufft from villages to sell in <strong>the</strong> market, 'acre selected and incrvic\.'cd.<br />
Their ages were comparable to those of <strong>the</strong> school childcn.
Data Analysis<br />
- 16 -<br />
A variety of approaches was utilized in analyzing <strong>the</strong> data. We<br />
started with ratior simple tabulations, such as frequency distributions,<br />
descriptive statistics, and cross-tabulations.<br />
Closer examination of discerned relationships central to <strong>the</strong><br />
main study involved multiple classification analysis uzing some of <strong>the</strong><br />
intervening variables as covariates.<br />
7e also drew upon path analyses techniques to estimate <strong>the</strong><br />
simultaneous relative power of variables in accounting for attitudinal<br />
and value outcome variance.<br />
Approach<br />
The strategy of <strong>the</strong> analyses was based on our conceptual frame-<br />
worh of <strong>the</strong> educational environment of <strong>the</strong> home, school, and peer group.<br />
G-enerally, Liultiple Classification Analysis (lICA) and path analysis pro-<br />
cedures were emplored to examine <strong>the</strong> criterion variable and <strong>the</strong> relatii-<br />
sii.ps among blocks of variables. These apuroaches were used. because we<br />
argued that <strong>the</strong> research study was <strong>the</strong> type in which prior performance<br />
vsriables served as antecedents -to predictor and criterion variables and<br />
exerted influence on both. Urless we can convincingly argue that no<br />
correlations exist between prior performance and. predictor variables,<br />
analysis of variance procedures raise serious analytical problems. How-<br />
ever, by <strong>the</strong> use of this alternative statistical tecIii.que as described<br />
in this report, meaningful results were obtaineö. and showed that <strong>the</strong><br />
techniques were appropriate and necessary for <strong>the</strong> effective examination<br />
of <strong>the</strong> complex relationships that existed among <strong>the</strong> variables. Since <strong>the</strong>
- 17 -<br />
reu1ts o± -this pilot study point to a number o± questions that should.<br />
legitimately form <strong>the</strong> subject of future investigation of our society,<br />
<strong>the</strong> tecbniques used in this study should be appropriate.
f' k C r 1T f1<br />
J LLiLI.0<br />
i:IuCTIT Ai DiSCRIPTIV2 STATISIC$<br />
In <strong>the</strong> literabore review no nade <strong>the</strong> ftdental assumption that<br />
school environnent provide school children with some aspects oi' learnin<br />
enpe:ienccs that -traditional ocioties lack, and. <strong>the</strong>se experiences in-<br />
±iuence <strong>the</strong> child to acquire desirable attitudes. On this basis, <strong>the</strong><br />
objcc-tive oi -tue various types of analysis in this chapter is to ascer-<br />
tam vihe<strong>the</strong>r it is true that school environments perform <strong>the</strong> ftmctioii of<br />
produein desirable attitudes in children -to a significant level. If<br />
chool do provide facil:Lties that help children to acquire desirable<br />
attitudes, <strong>the</strong>n children who have passed through <strong>the</strong> school system should<br />
dcoonatro-te hi;her attitudinal qualities than those who have not had this<br />
opportunity; also, <strong>the</strong> longer a child is erposed to school influence, <strong>the</strong><br />
better his attitudinal score. To es-tablisli this assumption, we compared<br />
<strong>the</strong> neaii attitude scores of cross-sectional date, for non-schoolers, and<br />
also for grades 1, 2, 4 and 6. Ii' each additional year of schooling was<br />
accompanied by a substantial gain in scores, <strong>the</strong>n we would attribute this<br />
to an overall school effect.<br />
At ano<strong>the</strong>r level o' analysis, <strong>the</strong> mean scale scores are presented<br />
in graphs shovi by grade for each subscale, nd also for <strong>the</strong> total atti-<br />
tude scale, to enable us to compare <strong>the</strong> scores for non-schoolers and<br />
schoolers Later in this and subsequent chapters, much more thorough<br />
analyses are given. We obtained simple zero-order intercorrelations and<br />
regression weights as measures of <strong>the</strong> influence of one variable on ano<strong>the</strong>r
- 19 -<br />
in a one-way causal model in order to measure <strong>the</strong> extent of <strong>the</strong> associa-<br />
tion between school experiences and <strong>the</strong> presence of desirable attitudes.<br />
The cross-seotional data are presented in table 2.1 below. The<br />
mean scale scores are reported for each of <strong>the</strong> frades sanpied. In order<br />
to teat for <strong>the</strong> level of significance of <strong>the</strong> differences observed among<br />
<strong>the</strong> mean icores for different grades, we adopted <strong>the</strong> null hypo<strong>the</strong>sis that<br />
subscales with significant levels of .10 to .001 are acceptable. 8iifi-<br />
cant levels greater th .10 allow too much risk, and significant levels<br />
of less than .001 are considered too ooiervative. Anj subscale with a<br />
signiiicance level greater than .10 was deleted from fur<strong>the</strong>r analysis.<br />
We gave, however, no lower bound for deletion, especially when <strong>the</strong> ETA<br />
squared value of a subscale is about ten percent, indicating <strong>the</strong> amount<br />
of variance being explained by it. Pollowing our coding convention,<br />
lower mean scores indicated better performance. The significance level<br />
of <strong>the</strong> overall attitude scale is below <strong>the</strong> .001 level and this may be<br />
atibuted to <strong>the</strong> influence of scale outliers.<br />
Table 2.1 SIARY 01? iIAh$ P01?. TIi2 TIiiE SUBSCALS<br />
AD POh T1 01TJAI ATTITT3 5CAI<br />
SUBSOAThS<br />
0PEifl]3SS TO iW<br />
RJi&Dfl']ES FOR<br />
SOCIAL Cl/dTGE<br />
Non-<br />
Schoolers<br />
2 3 6<br />
Significance<br />
level<br />
2 EA<br />
-<br />
1.375 1.306 1.303 1.216 1.236 .06 .06<br />
1.578 1.459 1.475 1.412 1.94 .001 .10<br />
OPINION C-11O'VTH 1.50 1.44 1.533 1.620 1.545' .001 .07
5iJDSCLIJS<br />
C '1 - L4..Lt L.tJ<br />
iTonchoolers<br />
\lhen ex i..ning tabls ama figures in tbis chapter, care must<br />
be tal:en to remember <strong>the</strong> coding convention we used to score <strong>the</strong> variables.<br />
A hi.gii Positive score on each attitude scale was assigned a low rating<br />
and conversely a lower mean score meant better performance.<br />
Openness to new experience<br />
2<br />
A six-item index for <strong>the</strong> first subscalo was a thistered to<br />
245 pulls. The scale was meant to determine how pupils would respond<br />
to op)ortwuties for new experiences, such as zna1dn friends with a new<br />
student in <strong>the</strong>ir school or travelling outside <strong>the</strong>ir immediate locality.<br />
The predictions of such predispositions go beyond <strong>the</strong> simple context in<br />
-<br />
-I<br />
- A<br />
ini<br />
ficonce<br />
level<br />
1.254 1 .293 1.306 1.133 1.247 .02 .06<br />
iIwh ) TTWST 1.304 1.01 1.329 1.13w 1.215. .001 .10<br />
DIG-IT. 1.479 1.401 1e)94 1.G6 1.329 .005 .03<br />
ATTI?UDE TO WOR3 1 .370 1 .327 1 .331 1 .238 1 .259 .001 .00<br />
VflO?JCH. 1.080 1.126 1.115 1.040 1.040 .02 .05<br />
PIhICIPATIOiT 1.514 1.558 1.597 1.566 1.538 .3 .01<br />
OviRAIa ATTIT1JDE 1.368 1.357 1.375 1.314 1.307 -<br />
6<br />
-<br />
'TA2
which <strong>the</strong> questions were sake d. It might embrace willingness to use,<br />
e.g., a new drug or new methods of farming. The results showed that on<br />
<strong>the</strong> whole pupils were willing to make new friends; and few of <strong>the</strong>m<br />
preferred staying in <strong>the</strong>ir present school to moving to a far distant<br />
school. See gure 2.1.<br />
GRADE<br />
S S<br />
Group mean<br />
.<br />
- 21 -<br />
3 4 6 0 2 3<br />
GRADE<br />
Fig. 2.1: Openness to new Fig. 2.2: Readiness for Social<br />
experiences.<br />
change.<br />
On <strong>the</strong> horizontal acLs, <strong>the</strong> 0 refers to n-schoolers, <strong>the</strong><br />
2 to grade 2, 3 to grade 3, etc.<br />
As <strong>the</strong> graph in figxe 2 1 dipped down to <strong>the</strong> right especially between<br />
grades 3 and 4, it indicated a considerable shift in response froni Grade<br />
2 to grades 3 and. 4, thus demonstrating <strong>the</strong> curiosity and desire of upper<br />
priy classes to meet and probably understand different people.
Readiness for social chanre<br />
- 22 -<br />
Group two subsonic items, seven :Ln nwiber, referred to pupils1<br />
acceptance of chtea in social orranizations; for example, whe<strong>the</strong>r <strong>the</strong>y<br />
considered it correct for people in authority to Share or delegate autho-<br />
rity. This was measured. by aith clues tiona framed around women' a and<br />
children's traditional roles in <strong>the</strong> houe. Some questions dealt with<br />
1eneral mobility across different laniuse groups in Ghana. Piure 2.2<br />
illustrate <strong>the</strong> results. It shows a definite response pattei. The nurfoer<br />
oi years pupils spent in school influenced <strong>the</strong>ir th1nin. Pupils in <strong>the</strong><br />
hi;1er classes felt that both parents should share domestic duties and<br />
affirned that equality of men and women was good. They also showed clear<br />
acceptance of scientific treatnent of cormaon diseases, which meant rejec-<br />
tion of sorcery.<br />
Growth of opinion<br />
The rovith in opinion subscalo souGht to discover <strong>the</strong> extent<br />
to which <strong>the</strong> pupiltS awareness of diversity of opinions and attitudes had<br />
increased beyond a General rejection of variant perceptions in o<strong>the</strong>rs.<br />
The scale was keyed to elements of tolerance and differences of opinion,<br />
while <strong>the</strong> distractors were based on autocratic and hierarchical ways of<br />
evaluatinG opinion. These ideas were woven into questions that requested<br />
<strong>the</strong> pupil to tell <strong>the</strong> interviewer whe<strong>the</strong>r he viouJ.d automatically accept<br />
<strong>the</strong> ideas of his peers and those above him in <strong>the</strong> power structure and<br />
reject <strong>the</strong> opinions of those below him. The table 2.2 and Piure 2.3<br />
show <strong>the</strong> response pattern by Grade.
Beiie±' in People and IrtstitutLons<br />
23 -<br />
Table 2.2 MEMS ID STAIDRD Di3VIJTI01T$ O1 G0WTL<br />
IN OPIIIOIT SUSCJ<br />
The results were unexpected, even though this is a Lair reflec-<br />
tion of genera]. classroom atxaosphere in Ghana. The figures tell us that<br />
younger pupils in grades 2 and 3 show greater flexibility as to whose<br />
opinion should be respected and accepted than do <strong>the</strong> higher grades, which<br />
tend to conLori arid accept orders ±'om above readily. We take consolation,<br />
however, in <strong>the</strong> fact that <strong>the</strong> sarnple under study was snail and restricted<br />
to one region. This prevents <strong>the</strong> generalization of results to o<strong>the</strong>r<br />
regions. Never<strong>the</strong>less, <strong>the</strong> present results show that lower primary grades<br />
and even non-schoolera have a more independent attitude than upper primary<br />
grades.<br />
alasa Me an Standard Deviation<br />
Non-schoolers 1 .50 0.26<br />
Grade 2 1,44 0.20<br />
Grade 3 1 .53 0.14<br />
Grade 4 1 .62 0.20<br />
Grade 6 1.55 0.21<br />
The trust and belief scale was investigated as Attitude 5 The<br />
extent to which pupils trusted taditional and government institutions -<br />
especially institutions that keep peace nd order - were important to
Fig. 2.3:<br />
Growth of opinion.<br />
24<br />
0 2 3 4 6 0 2 3 4 6<br />
GRADE<br />
GRADE<br />
Fig. 2.4: Belief in people and in<br />
institutions.<br />
<strong>the</strong> study. We exined a seven-item index that covered, anong o<strong>the</strong>r<br />
thiz;s, faith and lack of faith in <strong>the</strong>ir peer oup and in <strong>the</strong> poll ce.<br />
We are aware that students begin to show interest in school adiinisbation<br />
at <strong>the</strong> secondary school level. We have in recent years witnessed an<br />
increace in <strong>the</strong> incidence of confrontation between secondary school<br />
students and o1ice. What rniht appear to be secondary school and home<br />
effect could be present also at <strong>the</strong> primary ecliool level. The results<br />
shovied that 72 per cent of <strong>the</strong> respondents did not truz-t <strong>the</strong> police; about<br />
GO per cct would, however, trust tiie fa<strong>the</strong>rs The overall trend mdi-<br />
7<br />
6<br />
5-<br />
4-<br />
tdbj. ........<br />
2 Group mean<br />
1
- 25 -<br />
catd dcclin±n iub lie trust shovin 't<strong>the</strong> upper rades o:L' <strong>the</strong> prix<br />
school. 21e path analysis in Chapter 3 traced t7o paths leudinp Prou.<br />
-<strong>the</strong> -herial Possessions oP <strong>the</strong> home and Pareital Educa-ion variahle<br />
to this scale. The rerigths oP ti-ic pa-thu viere - .10 and .12 for<br />
lie-tend Possessions and Paren-tal :idueaion, respecively. Thin rieom-b<br />
uet pupils vino came Prom rcicuaveiy afiluen- eu uneducated homes<br />
tended to be lo on <strong>the</strong> scale. J11 thin points to <strong>the</strong> need to study<br />
student behaviour and most importantly to investigate how <strong>the</strong> interest<br />
of students in school aOitht-ti'citLon and government is related to<br />
educational aitainment and to <strong>the</strong> ot'ality of school attended. The inves-<br />
ipation should looh arefully at <strong>the</strong> rate ci? declinin trust, ii? any.<br />
See igure 2.4.<br />
Con:jdence in Oneself and O<strong>the</strong>rs<br />
:rine measures dealt with <strong>the</strong> pupi2!s confidence in himself to<br />
influence and. change hi. physical environment to his advantarpe. The<br />
i'eaul-ts are shown in figure 2.5. Questiona embraced a scientific explana-<br />
-ti on for na turaj. events (e .g. germination, causes of thunder) and covered<br />
also concepts such as a kleptomaniac'i will-power to stop his habit and<br />
<strong>the</strong> ability of a i:n to achieve a iredetermincd goal. As you may readily<br />
recogmize from <strong>the</strong> graph, <strong>the</strong> results showed that grade level, to a lare<br />
measure, seemed to have influenced achievement scores in this attitude<br />
scale In o<strong>the</strong>r words, personal dimension as defined by <strong>the</strong> scale had<br />
a strong nd positive impact on <strong>the</strong> score, with grade mean decreasing<br />
steadily from 1 47 for <strong>the</strong> non-schooling group through grade 2 to a<br />
s-batisticauy significant mean of 1.33 for class 6. (Refer to table 2.3,<br />
Attitude 4.) Given <strong>the</strong> findings outlined here, a logical flow of <strong>the</strong>
- 26 -<br />
Fig. 2.5: Confidence in oneself and<br />
o<strong>the</strong>rs.<br />
study from this initial position would have been to explore those aspects<br />
o± <strong>the</strong> classroom (teacher attitude, equipmit, tasic orientation, etc.)<br />
that influenced pupUM' confidence. A causal model in which paths lead<br />
froLa <strong>the</strong>se variables to <strong>the</strong> criterion model would have generated imich<br />
interest.<br />
0 2 3 4 6<br />
GRADE<br />
Attitude toward Work and Technical k1l1<br />
Pinally, two <strong>the</strong>mes, attitudes toward work commitment and<br />
techaical skills were dealt with. The two <strong>the</strong>mes are closely interrelated<br />
because -<strong>the</strong>y both involve manipulative sld.11s at <strong>the</strong> behavioural level.<br />
Thus a pupil described as havi.ig <strong>the</strong> right attitu3.e toward work must be<br />
observed to be interested in wor1dg. After a pupil had declared his<br />
position on each scale du.Dirg <strong>the</strong> interview he was fur<strong>the</strong>r probed for
- 27 -<br />
inforn2ation and evidence regarding his behaviour. Pupils' perorrnance on<br />
<strong>the</strong> two scales are shown by rap1is. ]gure 2.7 and 2.8 illusated that<br />
each step up <strong>the</strong> education ladder brought with it a £airly regular and sub-<br />
stantial increment in score which could be described as good attitude<br />
toward work and technical skills. We hesitate to sake any Lur<strong>the</strong>r conmen-bs<br />
0 2 3 4 6 0 2 3<br />
GRADE GRADE<br />
Fig. 2.7: Attitude toward work. Fig. 2.8: Valuing of technical skills.<br />
on <strong>the</strong> two scales because <strong>the</strong> <strong>the</strong>oretical bases l'or interpreting <strong>the</strong><br />
attitude off a pupil low on <strong>the</strong> scale a±1 to explain variances in terms<br />
o learning <strong>the</strong>ories.
- 23 -<br />
Table 2.3 SiRY OP LAI$ SUBSCAIS OVM?JJ SCAI<br />
Scalea<br />
Non-<br />
Sohoolers<br />
Grade<br />
2<br />
Lie an<br />
Grade<br />
3<br />
Fig. 2.9: Overall school influence.<br />
Grade<br />
4<br />
Grade<br />
6<br />
1 .375 1 .306 1 .03 1 .216 1 .236<br />
1UDIISS 1.578 1.459 1.475 1.412 1.394<br />
OPINIOIT 1.503 1.44 1.533* 1.620 1.545<br />
OTiNCE 1 .479 1.401 1.394 1.366 1.329<br />
1.304 1.301 1..29 1.186 1.215<br />
B2SPE0T 1.254 1.298 1.,06 1.183 1.247<br />
WORK 1 .370 1.327 1.3)1 1.2)8 1.259<br />
TEOiTICAI4 1 .080 1.126 1.115 1.040 1.040<br />
ITTITUJ<br />
1.368 1.357 1.375 1.314 1.307
Pigure 2.9 illustrates clearly how years of schooling influenced clas<br />
performance and, in fact, shows that a definite change occurs between<br />
grades 3 said 4. Even though no definite interpre-batii at this stage<br />
can be made, <strong>the</strong> general 'nei1ozt is that certain aspects of<br />
school environments foster desirable attitude formation. We hesitate<br />
to be specific here because <strong>the</strong> results could be attributed to maturity,<br />
which is one of <strong>the</strong> subjects for investigation in <strong>the</strong> following section.<br />
LW1TIPI CLkSSIPICATION PROCEDURES<br />
Ano<strong>the</strong>r approach to <strong>the</strong>se results is to use multiple classifi-<br />
cation analysis (IcA). One o± our initial predictions in <strong>the</strong> study was<br />
that <strong>the</strong> length of exposure to education would have a great impact (<br />
attitude formation end attitude change. To fur<strong>the</strong>r verify this, <strong>the</strong><br />
is:uc of maturation as competing with o<strong>the</strong>r school Thctors in children's<br />
attitude formation is examined here by using LICA. Since both <strong>the</strong> non-<br />
school and school children live in <strong>the</strong> sane geographical area, we<br />
obtained a sample of non-school children whose mean age and family socio-<br />
economic status were to a large measure equivalent to <strong>the</strong> age and socio-<br />
economic status of <strong>the</strong> school children sanipe. The nen-school sample<br />
was made of 45 children selected on :nrket days in tiie villages where<br />
<strong>the</strong> schools were located. The instrument was administered to each child<br />
individually and in private.<br />
Among <strong>the</strong> school factors <strong>the</strong> focus was on school quality. 'That<br />
is implied here is that good structural ±'ucilities influence <strong>the</strong> child's<br />
attitude in such a manner <strong>the</strong> child's attitude score becomes higher with<br />
each successive year in school, and that children who have never been
-to school and <strong>the</strong>refore live in poorer environnents should have lower<br />
ttitudo score The procedure for anaiysi -this situation is simple.<br />
ivon that both home and school environments operate in a complex manrie::,<br />
it is desirable to Liovi <strong>the</strong> no effect of each variable when differences<br />
in one variable are resioved or controlled. i.noe our criterion variable<br />
was measured on both school and non-school pupils and we wish to hold age<br />
ccaamt, regard measures taken on non-schoolers as base line, and <strong>the</strong>n<br />
e:ine <strong>the</strong> effect of year of schooiinf, iatui'ity and non-schoolirLf have<br />
become two important aspects of <strong>the</strong> analysis and ve must demonstrate that<br />
<strong>the</strong>y o not seriously cast doubt on <strong>the</strong> pattern shovrn in Taile 2.1.<br />
The data obtained in this study were a minture of ordinal,<br />
nominal as well as interval predictors in which <strong>the</strong> last one was obtained<br />
ma class intervals Vie explained elsewhere in <strong>the</strong> study that <strong>the</strong> crite-<br />
rica scores wore made up of addit:Lve conpoacntc for each cater;ory of e<br />
:rcdictor variables. 3'urvev data in fenoral have inter-correlated nrc-<br />
Lotors, so tha effects noticed dLath.': analysis of data could be cantri-<br />
bucd by a predictor variablctc cori'elcitioii with o<strong>the</strong>r predictors.<br />
3eeua of this, <strong>the</strong> usual analysis of variance procedure seem to enable<br />
us estimate parmeters of this analysis. hultiple Classification Analysis<br />
(::CA) aat±fios our desiji. Andreas et al.(173) suested <strong>the</strong> foilov;in.';<br />
cnus:tjon for <strong>the</strong> model:<br />
where<br />
Y = I:+<br />
Y = <strong>the</strong> score on <strong>the</strong> dependent variable<br />
= frond mean on <strong>the</strong> dependent varioble<br />
a1 = <strong>the</strong> effect of cated'ory 1 of predictor Tat<br />
= <strong>the</strong> effect of cLttofory 2 on predictor 'b'<br />
e = error term.
- 31 -<br />
In this approach, <strong>the</strong> :inin-effeet coefficients are expressed as<br />
deviations from <strong>the</strong> grand mean, sirailca' to <strong>the</strong> least squares procedures<br />
in which <strong>the</strong> constant term is restricted to <strong>the</strong> grand mean. The :;CA<br />
niodel fits a set of coefficients to obtain nicasures for a1 b2 in <strong>the</strong><br />
niodel. By this approach we are able to shov' <strong>the</strong> effects of any explana-<br />
tory variable both before and after takiii into account <strong>the</strong> effects of<br />
all o<strong>the</strong>r predictors. The fit of <strong>the</strong> model as a whole can be judged by<br />
<strong>the</strong> proportion of variance in <strong>the</strong> criterion explained by <strong>the</strong> predictors.<br />
With <strong>the</strong>se explanrtions as bac1round we set to examine <strong>the</strong><br />
pattern of chong in <strong>the</strong> effects of years o± schooling as we control for<br />
age., The results are presented in Table 2.4.<br />
Table 2.4 LTt31TIPI 0LA&IFICATI0N &IjYZIS<br />
N<br />
TJNJbD )JUS.TD GRiiD I.AIT<br />
(1.39)<br />
PITJ TA<br />
ETA ETA<br />
Non-sohoolera 40 .06 .06 1.45<br />
Grade 2 54 .03 .03 1.42<br />
Grad.e 3 47 .03 .03 1.42<br />
Grade 4 50 -.06 -.06 1.33<br />
Grade 6 54 - .05 -.04 1.34<br />
Eta Value<br />
Multiple R2 - .23<br />
IIultiple fl .48<br />
.47 .45
- 32 -<br />
Table 2.4 shows effects by grade. irst, eonaeeted with <strong>the</strong><br />
set of tu2adjusted effects for years of schooling (reported at tiie bottoxi<br />
of colunn 3) is an eta value of .47. B-ba is simply <strong>the</strong> correlation ratsLo<br />
and eta squared indicates <strong>the</strong> propotion of variance in <strong>the</strong> attitude<br />
scale e::plcined by <strong>the</strong> factors :Lu <strong>the</strong> sna:Lysiz; in this case, ae and<br />
years of schooling. Second, associated with <strong>the</strong> adjusted cateory<br />
effects of years of ehoolin on attitude is a partial correlation ratio<br />
of .45. This is shown imdex' <strong>the</strong> colunri labelled beta. Beta can be<br />
redarded as standardized reression coefficient.<br />
A comparison between <strong>the</strong> unadjusted etc and <strong>the</strong> partial beta<br />
yielded a valuable result. The squared 'betas' decreased from e221 to<br />
.203, a difference of an insifnificant value of .018 as we cantrolled<br />
for ace. That is, ae contributed about 2 per cent of <strong>the</strong> variance.<br />
2hds shows that while a very small part of <strong>the</strong> association between<br />
attitude and years of schoolini is due to iiatirity of pupils, educatian<br />
and o<strong>the</strong>r factors not yet partialled out accounted for nearly all <strong>the</strong><br />
variance. Iinally, <strong>the</strong> of .23 shows that 23 per cent of <strong>the</strong> varia-<br />
-tion in <strong>the</strong> attitude scale is explained by <strong>the</strong> additive effects of all<br />
facdors including <strong>the</strong> covariate, and coluirn 5 headed 'grand moan plus<br />
ola' ivos a simple pattern of <strong>the</strong> results. ecallingt1ie coding pr000-<br />
dare, we now have three distinct groups upper priaary, lower primary end<br />
ion-cchoolers, in that order of merit. This demonstrates <strong>the</strong> intensity<br />
Oi' school exposure, upon student attitudes.<br />
SUMMARY<br />
At <strong>the</strong> various stages of <strong>the</strong> analyses in this chapter <strong>the</strong>
- 33 -<br />
emphasis was on <strong>the</strong> general quality o± <strong>the</strong> school environment, as compared<br />
with what we may tern traditional influences such as A±'rican religion.<br />
The explanatory variables wor1d.n as a iitary force on <strong>the</strong> subscales and<br />
also on <strong>the</strong> composition scale o <strong>the</strong> criterion were explored. We have<br />
produced some evidence to show that <strong>the</strong> school environment can<br />
have positive iiffluence and that <strong>the</strong> school' a influence on children<br />
become greater with each successive year in school. By conaparin school<br />
and non-school results, we realized that school ciiildren demonstrate much<br />
more desirable attitude than non-schoolers.
L 2ODUCTION<br />
- 34 -<br />
CiIAI2 3<br />
OAUZAI Ii:rEL&TIOITS<br />
It has been established in <strong>the</strong> previous chapters that <strong>the</strong>re is<br />
a nosjtjvc association between formal schoolinr' and attitudinal values<br />
in children. In <strong>the</strong> preliminary alyses, including' ICA, <strong>the</strong> school vias<br />
re:Tded as a 5±nfle unit, so that school effects were shown as net of<br />
o<strong>the</strong>r coupe tins . However, <strong>the</strong> teeimiq.ues used could not enable<br />
us to delineate -<strong>the</strong> special £eatrcs of <strong>the</strong> school that influenced<br />
attitudes Ji'ui<strong>the</strong>rriore, <strong>the</strong> results we obtained did not o far enough<br />
to c::1:lain fully <strong>the</strong> relationship between <strong>the</strong> two environmental situa-<br />
tious (home end school) and !C criterion variable.<br />
Some arproaches have bean developed in recent studies for deal-<br />
in viith this kind of problem. Oxe of <strong>the</strong>se, <strong>the</strong> causal model, is used<br />
here to explain <strong>the</strong> linka:es between <strong>the</strong> home, <strong>the</strong> school and <strong>the</strong> crite-<br />
rion. The purpose of this chapter, <strong>the</strong>n, is to investiate selected<br />
aspects of <strong>the</strong> school cnvironraent which we thinic influence attitudes and<br />
o fur<strong>the</strong>r to examine <strong>the</strong> relationship between <strong>the</strong>se and home factors,<br />
sLnce i:revious studies have ouestod that home factors have direct<br />
effects on attitude.<br />
IiTLTTCE OP Tfl ETIION.IITT<br />
ieeves (1972) considered hone and school as dimensions
- 35 -.<br />
of <strong>the</strong> educational environment. They fur<strong>the</strong>r suggested that <strong>the</strong> school<br />
diiinsjons constituted forces that operated upon <strong>the</strong> child to bring about:<br />
changes associated with experiences in <strong>the</strong> home, <strong>the</strong> school and <strong>the</strong> child's<br />
class or peer group as social organizations.<br />
Williams (1976) presented a modified. model and indicated that<br />
environmental dimension or press models (such as that of Dave, 1963) could<br />
be flawed on <strong>the</strong> roinids that<strong>the</strong>yl) omitted parental abilities from <strong>the</strong><br />
models, and 2) focussed <strong>the</strong> analyses around <strong>the</strong> proportion of variance<br />
accounted for He said:<br />
Family environment influence on cognitive development<br />
must be seen within <strong>the</strong> context of a model that allows<br />
for <strong>the</strong> influence of parental abilities on children's<br />
directly, and indIrectly via <strong>the</strong> social and economic<br />
attainments of parents, via family environment, and<br />
through <strong>the</strong> socio-coonomic status - family environment<br />
linkage itself ... parental abilities seem. to<br />
be clear-cut antecedents to all vciables within such<br />
models.<br />
In this sense it is somewhat parado:cal to find an<br />
elaborate multidimensional <strong>the</strong>oretical model operationalized<br />
empirically by a single statistic measuring<br />
<strong>the</strong> combined effect of <strong>the</strong> several environmental<br />
dimensions.<br />
This suzwey does not intend to examine <strong>the</strong> merits and demerits<br />
of <strong>the</strong> two aspects of <strong>the</strong> model hero but to show that several forces<br />
interplay in a ra<strong>the</strong>r complex way. Any efforts to use <strong>the</strong> causal model<br />
to explain how <strong>the</strong>se forces influence a cognitive and attitude<br />
learning should bear in mind <strong>the</strong> issues of nature and nurture These<br />
are discussed in <strong>the</strong> following paragraphs.
I rome IJnvjronrnent<br />
Two r)C5 Of ileajuros o: hIC :lOfllC eyLvi:ponmellt v;ere ei)lorcd in<br />
<strong>the</strong> study: iioasureo concerned viitli iuvl;orial possessions of <strong>the</strong> home end<br />
pulon ml education. We a<strong>the</strong>red inforuution from <strong>the</strong> children on a nunbor<br />
of related socio-cultural aspects of <strong>the</strong> home. Put problems of multi-<br />
colinoority in rereseion analysis forced us to stick to three variables:<br />
fa<strong>the</strong>r's education, mo<strong>the</strong>rts education, od nunber of books in <strong>the</strong> home<br />
which were in turn combined to form a composite index to assess <strong>the</strong> socio-<br />
cultural level of <strong>the</strong> home.<br />
The second type of measures of home environment dealt speci±'i-<br />
cally with some economic items of <strong>the</strong> home and not parental behaviour.<br />
\7e feel that <strong>the</strong>se play important roles in living conditions. Vernon's<br />
(19G5 cross-cultural study on <strong>the</strong> relationship between environ-<br />
mont and ability used eight categories which included housing, equipment<br />
in <strong>the</strong> home and books. The Piowden Report (i 9G7) subjected 80 variables<br />
to factor analysis and identified 14 variables including physical ameni-<br />
ties and literacy of <strong>the</strong> home. The two studiea were significant in many<br />
ropects but not clear in worhin; out <strong>the</strong> linkages among <strong>the</strong> dimensions.<br />
&e have included items like <strong>the</strong> refrigerator, gas/electric cooker, and<br />
sanitary facilities to trace <strong>the</strong>ir path to <strong>the</strong> criterion. We believe<br />
that as <strong>the</strong> home atmosphere varies for tl'ie growing child so varies <strong>the</strong><br />
stimuli <strong>the</strong>y provide for learning attitudes and values.<br />
School Environment<br />
There are several dimensions to <strong>the</strong> concept of school environ-<br />
ment, such as learning environment, classroom behaviour of <strong>the</strong> teacher,
- 37 -<br />
teacher's attitude towards teaching, and physical setting of <strong>the</strong> school.<br />
The study was concerned with <strong>the</strong> physical environment, organizational<br />
structure of achool, arid classroom characteristics. These three were<br />
selected 8imply because prinvry school buildings and school equipment iii<br />
Ghana are in deplorable conditicir and we wish to draw attention to <strong>the</strong>n.<br />
We think <strong>the</strong> presence or absence of' such facilities as school toilets,<br />
desks, recreational areas, playground, equipment, textbooks and library<br />
may have a inaj or role in attitude formation and change. After several<br />
visits to schools in <strong>the</strong> sample, we prepared a school resource index<br />
form containing 11 items This form was completed by a trained inter-<br />
viewer for each school and according to a ixrccoded scheme. Items were<br />
combined to reflect <strong>the</strong> underlying property of each variable. The<br />
resulting values were recorded for each child, depending on which school<br />
he was aasociated with.<br />
By school management, we mean <strong>the</strong> organization responsible for<br />
adiiinistering <strong>the</strong> schools1 The government of Ghana, for several years,<br />
took over <strong>the</strong> administration o± all first cycle schools in <strong>the</strong> public<br />
system and shared financing of <strong>the</strong>se schools with local councils.<br />
Government, however, reconsidered its position recenly and came out with<br />
a new policy allowing religious bodies to build and administer <strong>the</strong>ir<br />
schools. Since <strong>the</strong>n, religious bodies have funded and managed schools.<br />
It was thought interesting to find out <strong>the</strong> impact of schools run by<br />
local councils and religious bodies on student attitudes.<br />
Classroopeer group<br />
Researchers such. as Bronfembrenner et al. (1965) and Holsinger
(172) have reported results thd±catii. <strong>the</strong> socialization influence of<br />
<strong>the</strong> classroom. The ±'oriner found that a teacher' a behaviour towards his<br />
pupil indicated a hiker relation to id values than <strong>the</strong> pupil's report<br />
of his parents. Tha cbi1d's c].aBsroOn behvioui, hi attitude to hiB work and<br />
hiz attitide of his soøtyre <strong>the</strong> ix4ioators: aelsinger oornbied. into, a<br />
sinle surrry measu.re of classroom 'ivality and tried to estimate <strong>the</strong><br />
strent1i of <strong>the</strong> relationship between attitude and classroom environment.<br />
Vie abinstered Inheles and Smith' a scale (Short JTorni 2) containing 13<br />
itens to teachers.<br />
Lost urban schools in <strong>the</strong> experiment practised team teaching.<br />
Thus it was difficult to associate a part:Lcular teacher with a class<br />
din coding sessions. We were loft with one al-ternative, that is, we<br />
used peer group as a r;lcasure of <strong>the</strong> contention that years of schooling<br />
increases <strong>the</strong> in:;ensity of school exposure to school learning. Vie 110w<br />
turn to see how <strong>the</strong>se variables related in a causal sequence. The<br />
variables are reported in Table 3.1.<br />
- 36 -<br />
TABM 3 1 $tiAIY OP VIABLiS LI :li iIODL<br />
Variable name<br />
1 Parental dueation<br />
Description<br />
Ordinal measure based on La<strong>the</strong>r's and<br />
mo<strong>the</strong>r's levels of education respectively.<br />
4 rcpresenbs university education,<br />
3 diploma<br />
2 elementary school,<br />
1 no forial education.<br />
These were coded from children's interview<br />
response sheets.<br />
(ii) ilurrfocr of books in <strong>the</strong> home.
6. Classroom<br />
Variable name Description<br />
Material possessions<br />
in <strong>the</strong> home<br />
Physical structure<br />
of <strong>the</strong> sohoo]/school<br />
characteristics<br />
School location<br />
School miagenient/<br />
Proprietorship<br />
The interviewer checked which of 11 items<br />
<strong>the</strong> respondent reported to have at home.<br />
A code of 'i' represents more than 5 of<br />
<strong>the</strong>se items and '2' from 0-5 items.<br />
School sanitation - three-point<br />
scale.<br />
School bui.l&Ln - nine-point scale.<br />
Playi..ni &rounds - two-point scale.<br />
Recreational equipment - six-point<br />
scale.<br />
This referred to whe<strong>the</strong>r a school was<br />
located in a rural or urban environment.<br />
Rural environment is defined by <strong>the</strong> folloiinc:<br />
Number of times <strong>the</strong> village received<br />
natIonal newspapers a week.<br />
Presence of electricity.<br />
Presence of pipe-borne water,<br />
Presence of health clinic.<br />
Number of lorries that come to <strong>the</strong><br />
place daily.<br />
Presence of a cinema house.<br />
3 By a church<br />
2 By local authority<br />
1 Arabic school type<br />
(These were recorded using dummies for<br />
each child.)<br />
Bac1çround influence of teachers<br />
such as - age, sex, professional<br />
and. academic qualifications, and<br />
experience. These were precoded<br />
on <strong>the</strong> questionnaire.<br />
A thirteen-item attitude scale<br />
based on Inkeles' modernity<br />
scale, Short Form 2. These vere<br />
cod.e With '1 * for desirable<br />
attitudes and '2' for undesirable<br />
attitudes.
7.<br />
ex<br />
8. Religion<br />
Variable name<br />
Se:: was coded:<br />
DLSCIIIPTIOJ AITD IYE1TTIPiC.TIO1T OP PAT1IS<br />
1 ior female<br />
0 for male<br />
Consistent with <strong>the</strong> intention of taking into account hone end<br />
scbool environments, a chain model wa docigned. To do this effectively,<br />
five blocks of explanatory variables were formed and included in <strong>the</strong>' study.<br />
Por ouch block of variables, a single composite variable was formed, and<br />
;ie re3.ationbips between <strong>the</strong> blocks and <strong>the</strong> criterion were described in<br />
eruz of a causal sequence. The whole procedure was a long process of<br />
constructing variables which would assist in <strong>the</strong> explanation of variance<br />
in <strong>the</strong> criterion, provided we succeeded in organizing tese variable into<br />
logical blocks. Two blocks of factors described family environments of<br />
<strong>the</strong> child in terms of parental education and material possessions of <strong>the</strong><br />
hone. Both blocks constituted <strong>the</strong> structural dimension of <strong>the</strong> home. The<br />
third, fourth, end fifth blocks of variables reflected school environ-<br />
menta. These were management of school proprietorship, physical facili-<br />
ties in <strong>the</strong> school, and classroom or peer group.<br />
Description<br />
3 for Christianity<br />
2 for African religion<br />
1 for :iuslim religion<br />
In addition to <strong>the</strong> five ooQite blocks of variables, sex,<br />
religion en location of school were used in <strong>the</strong> estimation of <strong>the</strong>
- 41 -<br />
mainitude of <strong>the</strong> relationships between <strong>the</strong> components of <strong>the</strong> causal model.<br />
One of <strong>the</strong> main constraints of <strong>the</strong> causal model is <strong>the</strong> concern<br />
for sequence (gure 3,1). Temporal sequence determines how <strong>the</strong> variables<br />
enter <strong>the</strong> causal model. Accordinrly <strong>the</strong> variables Parental Education and<br />
1aterial Possessions of <strong>the</strong> home come first in this sequence. Since no<br />
arrows lead to both variables any variabilities in <strong>the</strong>m may, for this<br />
reason, be attributed to influences outside <strong>the</strong> causal model. Variables<br />
that are not influenced by o<strong>the</strong>r variables in <strong>the</strong> model are said to be<br />
exoenoua t)hidimensional lines with arrow heads flowini from variables<br />
(x1) and (x2) to variables School LInnarçerrient type, School Structural<br />
Facilities and Classroom Characteristics depict <strong>the</strong> second set of variables<br />
in <strong>the</strong> sequence. The ±i.nal variable (L1.) is in turn influenced by all<br />
variables preceding it. The arrowheads act like valves to demonstrate<br />
that. variables cannot perform two functions of being <strong>the</strong> cause and<br />
effect of o<strong>the</strong>r variables Thus <strong>the</strong> variables (x3) to (x7) are caned<br />
endoanous, that is, <strong>the</strong>y are influenced by o<strong>the</strong>r antecedent variables<br />
and in <strong>the</strong> circumstances <strong>the</strong>ir variability can be explained, to sonic<br />
extent, by both exogeno.w and o<strong>the</strong>r endogenous variables in <strong>the</strong> model.<br />
It is also possible that some variables may play mediating roles and<br />
will <strong>the</strong>refore influence o<strong>the</strong>r variables indirectly. No survey studies<br />
have ever claimed o explain one hundred per cent variance All un-<br />
explained vsrianoea or error equations are shovi by (u), (m), (v), and<br />
(w). We assume that <strong>the</strong> errors correlate nei<strong>the</strong>r with <strong>the</strong>mselves nor<br />
with independent arid dependent variables. Where <strong>the</strong> model shows bi-<br />
directional arrows <strong>the</strong>y are meant to indicate non-causal correlations<br />
between variables.
Dir X<br />
III SEX<br />
X3<br />
1il'<br />
SCHPROP<br />
X 5<br />
PARENDU<br />
cuss<br />
PEERS<br />
A14k<br />
.<br />
MATTPOSS<br />
ATT ITUL]<br />
RE LI G ION<br />
Lure 3' 1 T0TDSIZ'D PATHS .O1.I iau IwI:aO11ffNT TOtTGiI<br />
S 011001 TO ATTITUDE<br />
1Piure 3.1 represents paths of influence from home environ-<br />
ment tbrouh school setting end subsequent lthaes leadi finally<br />
to <strong>the</strong> criterion. The lines show <strong>the</strong> causal iinacees we intend to<br />
invei;ate, xsciinin <strong>the</strong> diagram om left to riht we see that home<br />
environment affects school rnanaenient, structural characteristics of<br />
<strong>the</strong> school and it also influences classroom or peer roups as vell as<br />
<strong>the</strong> criterion. There are two different sets of independent variables.<br />
The first set ha two separate but related components and em.bracos<br />
those variables that describe <strong>the</strong> home environment. One ol' <strong>the</strong>se two<br />
components consists of four levels of parental education, <strong>the</strong> raeasures<br />
of which were later on combined into an indcc. The o<strong>the</strong>r component<br />
contains measures on material possessions o:L' <strong>the</strong> home. Our second set
43 -<br />
of variables describes school management by type (church or local autho-<br />
rity) structural characteristics of <strong>the</strong> school and classroom or peer<br />
group. All <strong>the</strong>se are taken to compete with home environment variables<br />
in forxnin and changing children1s attitude.<br />
Since several measures were taken on each of <strong>the</strong> components,<br />
single indices were formed from <strong>the</strong> many observed nieasu.res relating to<br />
<strong>the</strong>m and by this we formed six bTàTèki iaiià1Iisüáeiy naI<br />
education, material possessions of <strong>the</strong> home, school management, school<br />
structural characteristics, classroom or peer group and <strong>the</strong> criterion<br />
that isattitude scale. In addition to <strong>the</strong>se we took age, sex and<br />
school location as given and associated with our variables. We also<br />
added religion because it has much to do with <strong>the</strong> value system of <strong>the</strong><br />
G.hpnni a.<br />
The hypo<strong>the</strong>sized paths traced in figure 3.1 carry negative and<br />
positive signs based on assumptions described here. Children of educated<br />
parents in general pare highly motivated and have greater ability than<br />
o<strong>the</strong>rs. (iven parental education arid material wealth of <strong>the</strong> home, parents<br />
may look for goo4 schoola with regard to management, curriculum arid physi-<br />
cal enviraruuent. They may even obtain special student services such as<br />
enrichment lessons for <strong>the</strong>m. Children from poor homes tend to be less<br />
motivated and <strong>the</strong>refore less able, unless <strong>the</strong>y are provided efficient<br />
teachers and professional help such as remedial toaching. In this way<br />
two measures - parental education and material possesions of <strong>the</strong> hone .<br />
serve as measures of both parental and community preferences. Based on<br />
<strong>the</strong>se, we should find positive direct effects on school niariagernent arid<br />
school facilities, including staffing. The lvel of material resources<br />
of <strong>the</strong> home should, however, be related negatively to instructional
- 44 -<br />
effectivetiea with parental ethuution having a small direct effect on it.<br />
'he location of school will affect ngatively school physical facilities<br />
bix should nei<strong>the</strong>r effect management nor classroom instruction. Religion,<br />
age and sex should have negative effect on <strong>the</strong> interveaing variables.<br />
What shuld be <strong>the</strong> effect of each of <strong>the</strong> variables in our model<br />
on pupils? If what we ostulate is correct, <strong>the</strong>n all <strong>the</strong> exogenous<br />
variables, except religion, should not have direct effects on attitude<br />
outcome. Whe<strong>the</strong>r any of <strong>the</strong> school enviroirnental factors (a set o±<br />
organizational attributes involving functions such as school management<br />
and fiscal support t'or progranmles) has direct conseq.uence for attitude<br />
achieveraent is moot. If <strong>the</strong>re are any effects <strong>the</strong>n attainment should<br />
be influenced by management facilities in <strong>the</strong> school through teacher-<br />
pupil classroom relations. The gains from manegernent and facilities on<br />
attitude should be very small, while we expect large positive effects<br />
from peer groups or class on attitude.<br />
Identification of paths<br />
The following set of equations are used to identify <strong>the</strong> various paths in<br />
<strong>the</strong> causal model o± <strong>the</strong> attitude scale:<br />
x =p<br />
3 32 2 + 3l l + 3O X0 + P X<br />
3u U<br />
x =p x +p x +p x +p x<br />
4 433 422 411 4im<br />
x5=P53x3+P x +P X +P X +P X +P X<br />
52 2 54 4 50 0 51 1 5v
where:<br />
= Sex<br />
- 45 -<br />
X1 = Material possession of <strong>the</strong> home (TTPoss)<br />
= Parental education (PAiEiDu)<br />
X3 = School nvngenient type (scia'iwp)<br />
= Physical aoilitiea of <strong>the</strong> school (SCHcART)<br />
= Classroom characteristics (aIA.SS/PEE1S)<br />
= School location (ScnocAT)<br />
= Overall attitude (cIuRIor)<br />
=<br />
= Religion<br />
QuantifyiEg <strong>the</strong> paths<br />
= Error term for school msnement type (x3)<br />
= Error term for classroom characteristics (x5)<br />
= Error term for overall attitude (x7)<br />
In order to pick out and establish inportant paths we explored all<br />
hypo<strong>the</strong>sized linkaees by forrninr <strong>the</strong> following set of 'normal equationst<br />
from <strong>the</strong> basic causal model.<br />
Ti? P12<br />
123<br />
P32 + P31 1'12 + P30 T02<br />
r =P r +P +P... r<br />
13 32 12 31 ;O 01
- 46 -<br />
= P32 r + 230<br />
02 )1<br />
- P43 + P42 + P41 r17<br />
r =P,r+P +P r<br />
24 4 2 42 41 12<br />
114<br />
= P43<br />
= 53<br />
i3<br />
P53 T23 +<br />
+ P42<br />
12 +<br />
Pr +<br />
52 T29 + 5O O3<br />
52<br />
r45 P53 r34 + P52<br />
+ P54 T24<br />
24<br />
+ P54 + P50 T04 +<br />
P51 T13<br />
O2 i2<br />
T05 P5 + P b02 + P54 3?04 P50 + P51 r01<br />
3 0 52<br />
= F5? 1 + I I<br />
P75 + P72<br />
13 2 12<br />
25<br />
127 = P57 + 73 r 23<br />
To7 = P97 T05 P72<br />
1: +P<br />
57 53 72<br />
., =P r +P<br />
79 57 59 72<br />
+ P4 + P50 TO1 + P51<br />
14<br />
+ P70 r05 + P73 r58 P79 r59<br />
28<br />
+ + P70 r2 + P73 r28 + P79 T29<br />
+ P70 + P78 r08..+ P79 r09<br />
T28 + P70 r03 + P73 + P79 r89<br />
29 + P70 109 + P?8 r9 + P79<br />
Zero-order correlatians between each of <strong>the</strong> variables in <strong>the</strong> basic causal<br />
uodel were obtained so that in P1. <strong>the</strong> relationship between an indepen-.<br />
dent variable X. cnd a dependent viable X1 could be expressed as <strong>the</strong><br />
o <strong>the</strong> direct effect of X. on X., <strong>the</strong> indirect effect of X. on X.<br />
1<br />
mediated by <strong>the</strong> relationship of X. with o<strong>the</strong>r independent variables in<br />
<strong>the</strong> raodel, and any causally unonalysed component.
- 47 -<br />
Table 3.2 ZO 0RDfl CRMTI0ll5<br />
1 Overall scale (iTiw) I<br />
2. Parental education (PIDu)<br />
3. Literial possession (TTPoss)<br />
4. Classroom (CiA.sS/Is)<br />
5. Physical facilities (SOHCART)<br />
6. iIanagement (soop)<br />
7. Sex<br />
0. Jge<br />
). Religion<br />
C!)<br />
a)<br />
r1 d P<br />
d C!) -P<br />
4-' ca r4 )<br />
a) H<br />
C) O<br />
::i Cl) 0 b<br />
rcj 0<br />
a) 2.i<br />
H H 0<br />
r<br />
C) H<br />
u rl 0<br />
r1 a) a) Cl) Cl)<br />
a)ca<br />
0<br />
-! -P c - < a)<br />
0 P4 C12 t'<br />
H . 0 a) Ma)<br />
These simple correlations enable us to examine <strong>the</strong> variables<br />
not only when <strong>the</strong>y tnd alone but also as <strong>the</strong>y enter into a more coraulcx<br />
competition with each o<strong>the</strong>r. Six variables ShOW moderate correlation with<br />
<strong>the</strong> criterion (i.e., <strong>the</strong> overall scale) and out o <strong>the</strong>se, classroom effect<br />
come up as <strong>the</strong> highest WIth .24. In rank order, this result does no-b<br />
differ from <strong>the</strong> .59 reported by Inkeles and Smith (1974).<br />
ie disadvonte about correlational interpretation is that in<br />
a simple correlational situation, one can hnvdly explain how much an,r<br />
variable borrows its manifest atrenth from ano<strong>the</strong>r with which it is<br />
closely linked. In this reaa'd, multiple regression analysis is con-<br />
sidered more efficient.<br />
.08<br />
.05<br />
.24<br />
.12<br />
.09<br />
.04<br />
.10<br />
.18<br />
.59<br />
.11<br />
.21<br />
.38<br />
.O<br />
-.09<br />
.2<br />
.18<br />
.45<br />
.56<br />
.07<br />
-.12<br />
.16<br />
.26<br />
.40<br />
.01<br />
.19<br />
.29<br />
.38<br />
.02<br />
.06<br />
.25<br />
.00<br />
-5<br />
.01<br />
.05<br />
.07 .02
One basic i terest in <strong>the</strong> path analysis procedure is to estimate<br />
<strong>the</strong> strength of <strong>the</strong> linkages between blocks of variables and use <strong>the</strong> magni-<br />
tude and sign of <strong>the</strong> coefficients to explain <strong>the</strong> underlying causal model.<br />
To obtain a solution we can ei<strong>the</strong>r substitute <strong>the</strong> correlations we have in<br />
Table 3.1 into <strong>the</strong> normal equations we generated or simply regress each<br />
endogenous variable on those variables that directly influence it, using<br />
unstandardized regression weights as path estimates. Blalock (1964)<br />
cautioned on <strong>the</strong> use of correlations and standardized regression weihts<br />
and au';fosted <strong>the</strong> use of wistandardized regression weights instead. He<br />
The unstandardized reTession coefficient is<br />
suggested as acre appropriate for <strong>the</strong> statement<br />
of causal laws than <strong>the</strong> correlation coefficient<br />
or <strong>the</strong> standardized regression coefficient since<br />
<strong>the</strong> untsndardizod coefficient is relatively<br />
stable across subsanpies of a population where<br />
<strong>the</strong> standardized coefficients may vary significantly,<br />
as a function Of <strong>the</strong> standard deviations.<br />
The argument against <strong>the</strong> use of standardized regression weights<br />
nd correlations, as indicated in <strong>the</strong> above quotation, is that <strong>the</strong>y are<br />
unduly influenced by <strong>the</strong> particular distribution of variables involved<br />
in a study. Coleman (1975) recommended <strong>the</strong> use of <strong>the</strong> standardized<br />
regression coefficient or, where meaningful, <strong>the</strong> unstandardized regression<br />
coefficient. The latter has greater power in generalizing from one study<br />
to <strong>the</strong> ne:t. We, however, used <strong>the</strong> standardized regression weights<br />
because of <strong>the</strong> ahortcomings of <strong>the</strong> scales on which our predictor variables<br />
.eeo ucaaircd. W.th iit yaianeo .aid zero tdnd5rd devation it wm<br />
rossible for us to use beta coefficients to estimate <strong>the</strong> change in <strong>the</strong><br />
criterion score vthi.ch was related to a standard unit variation in <strong>the</strong><br />
predictor. Path coefficients were <strong>the</strong>refore identified by beta weichts
derivod from multiple re ression ooedures viith airwise deletion 2cr<br />
rnissin data. The formula vias used to determine residual path<br />
coefficients.<br />
Table 3.3 SAi CRITI0NS AIJ) ESTILU3D PATI CPPICTT$<br />
0R TI CAUSkL :ioL<br />
Sample Correlations standardized Path Coefficients<br />
"12<br />
"23<br />
"13<br />
"03<br />
"34<br />
124<br />
"14<br />
.59 P32 .11<br />
.38 P12 .59<br />
.56 .51<br />
.01 P30 - .08<br />
.38<br />
.21<br />
r-. .40<br />
"25<br />
"15<br />
"05<br />
"57<br />
"37<br />
"27<br />
"17<br />
P.,<br />
5;)<br />
P<br />
52<br />
.40<br />
- .00<br />
.45 P. .17<br />
.11<br />
50<br />
1<br />
.01<br />
-.12<br />
.18 P43 .19<br />
.26 P42 - .09<br />
.01 P41 .39<br />
,24 P75 .19<br />
.09<br />
7)<br />
-.00<br />
.12 P74 .06<br />
.08 P72 .05<br />
.05 P71 - .04
SaLlplO Correlations<br />
07<br />
r79<br />
.02<br />
.04<br />
r78 .10<br />
.18<br />
- 50 -<br />
Standardized Path Coefficients<br />
:oi' ou' purioe we rcided to use <strong>the</strong> followir as our criteria:<br />
P76<br />
P70<br />
P79<br />
P78<br />
(±) Stron path equols coefficient Teater than 0.25<br />
Llodera-1;e jath equils 0.11 to 0.24<br />
1eaL path equals 0.05 to 0.10<br />
Table 3 4 :a.T;. iEIGHTS :an 0V$;ALL LD ZTCJLES<br />
.00<br />
.05<br />
.11<br />
.07<br />
0<br />
ca p a<br />
Q c4 a) H<br />
a 0 a) a> + ci<br />
a a<br />
Cl) P 0-P<br />
Cl) + t r4 0 C) -P<br />
H - H P rI<br />
H<br />
Cl) Or<br />
9<br />
(l) rj<br />
C) Cl) rQ a>0 -P0 a><br />
p4 th 0 a p4 - ' -<br />
0irsroora aotoro .19 -.05 .06 .20 .16 .16 -.05 .10 .11<br />
Sch. nnaernent -.00 .18 .07 -.06 .00 .00 .06 .00 -.11<br />
eh. 'acilities .08 -.13 .12 .05 .08 -.07 .20 .00 .00<br />
Pi'entaJ.. education .05 -.00 .00 .08 .06 - -.05 .12 .07<br />
L±ly iaaterial poss. -.04 -.06 .00 -.13 .00 - .13 -.10 .06<br />
$ o:: .05 .00 .14 .00 .00 .00 .08 .00 -.06<br />
Ae<br />
n'7<br />
*'..j<br />
rv\ JJ - 'J'J rr *LJ_ rr<br />
.0r\r\<br />
.J.i r ,v7 ¼J<br />
Ieliion as a whole .11 -.00 .08 -.11 -.05 .00 .24 .06 .38<br />
C1rirtizm reli$ion .00 -<br />
JO tv'
Ancestral worship<br />
lIe slexa religion<br />
School location<br />
School by nianaement<br />
Ohurch<br />
Local authority<br />
Sub-scales<br />
The figures reported in table 3.4 for <strong>the</strong> overall scale indicated<br />
that classroom factors emered as <strong>the</strong> niajor influence in <strong>the</strong> total sezaple<br />
and ain cane out significantly first in three o<strong>the</strong>r sub-scales. School<br />
facilities, school management and parental educot..on followed next in<br />
that order. This outcome clearly demonstrates that classroom or peer<br />
group effects, toge<strong>the</strong>r with o<strong>the</strong>r school related factors, make con-<br />
vinci,g and independent contribution in having children acquire desirable<br />
attitudes.<br />
.02 -<br />
Built into <strong>the</strong> path ualysis method are <strong>the</strong> techniclues for<br />
measuring <strong>the</strong> direct and indi'ect effects that one variable has on<br />
ano<strong>the</strong>r, and a mechanism that helps us to decompose <strong>the</strong> correlation<br />
between any two variables into a sum of single and compound paths.<br />
1J Sewall Vrig1it (1964) method, a compound path is computed by <strong>the</strong><br />
product of <strong>the</strong> simple patll9 comprie:Ui it.<br />
0<br />
Cl) + C))<br />
C) C) -P<br />
.00 -.09 -.06 - .00 .07 .07 -<br />
OH<br />
o<br />
O-f 0-P<br />
O -P00 -I<br />
-P .-1 rl CiH<br />
C)) .-4 -1 -P H<br />
0r1 ri 00 +'O<br />
r4 W -'F-4 HC)<br />
i-i Md £-C))
1:iiioinii <strong>the</strong> Paths<br />
- 52 -<br />
Figure 3.2 MODEL SHOWING PATH COECIENT8<br />
Oonsidering <strong>the</strong> values of <strong>the</strong> path coefficients, classroom or<br />
peer group characteritios Eain had <strong>the</strong> strongest direct effect on <strong>the</strong><br />
crie'ion. Thia indicates that, in general terms, <strong>the</strong> higher <strong>the</strong> class<br />
level of <strong>the</strong> child, <strong>the</strong> better he performed. on <strong>the</strong> atti1de scale. This<br />
finding confirms <strong>the</strong> results of variance and multiple classification<br />
analysis already reported. O<strong>the</strong>r variables like religion, age, end school<br />
structwl facilities in that order, had mod.erate paths. The paths froLl<br />
parental education and material possessions of <strong>the</strong> home to <strong>the</strong> criterion<br />
indicate <strong>the</strong> weakest links among <strong>the</strong> ix'ect paths The negative sign of<br />
<strong>the</strong> latter's path to CLIU3S and <strong>the</strong> weac regression coefficients reported<br />
for Ilaterial Possession (LTTPOs$) nd Pareta1 iducation (PiJDT) seemed
- 53 -<br />
to surggest that children from good homes performed poorly on <strong>the</strong> attitude<br />
scale and. that physical quality of <strong>the</strong> school d±d not influence <strong>the</strong>m as<br />
much as it did. to children from relatively poor homes in acquiring desi-<br />
rable attitudes. The edeting links from School LianagerLlent (scHcP) and<br />
School Structural Facilities (scIIaT) to CLtSSiWOii characteristics were<br />
expected. PAPJi}DU had a near to nil path to CISSROOLi while i.TOS led<br />
a strong negative direct path to CIkSSROOLI. However, both PAEITDIJ and<br />
LL1TPCSS had strong links with SCEPROP and SCH(ThRT. These observations<br />
show that quality o±' school has a tremendous influence on children, as<br />
far as attitudes are concerned; hence, ve have strong direct paths from<br />
bC1OP azie. SCHCT to ClASS. ven though <strong>the</strong> home (i.e., i-.eJrJ,JtT and<br />
fTPa) does not seem to be as eliective as tue scnool, its eiect 15<br />
seen in its strong links with ittRC and SCIICX??. Probably, parents with<br />
good educational end financial bac1ron obtain quality schools for <strong>the</strong>*<br />
children.<br />
Instead of reportiig cigni:Licanee levels for tile affects<br />
observed as usuaJ.ly done through UT(VA, vie adop od <strong>the</strong> decomposition<br />
Ltethod to explain how much confidence we can put on psth-Lua.lyticai<br />
results.<br />
hi0000atio: of :Lnto D:iro c arid Inai.LOCL Cc: IOrantc<br />
geuerr.l method for dccorpoaing total efPQctE :h::o <strong>the</strong>:L C C.:<br />
here.<br />
and indiret effects suggoeted by uior - (i 97) ir
Table 3.5<br />
Influencing<br />
Dependent Variables<br />
Total<br />
Effect<br />
Indirect Effect<br />
X3 X4 X5<br />
Direct<br />
Effect<br />
X1 .51 .51<br />
X2 .1i .11<br />
- -<br />
X .10 .10<br />
X1 .39 .097 .293<br />
x6<br />
-.09 .021 -.111<br />
x9<br />
X2<br />
x6<br />
Xc?<br />
- - - -<br />
.19 .19<br />
-.12 .2014 .016 -.34<br />
-.005 .0/44 .004 .05<br />
.40 .40<br />
.17 .17<br />
X7 X1 -.04 -.005 .006 .003 -.03<br />
.049 -.001 .001 .007 .04<br />
X6 .005 - - - -<br />
X9 .11 .11<br />
X3 .001 .001<br />
.06 .06<br />
Xr .19 .19
- 55 -<br />
The table 35 illustrates <strong>the</strong> cf:'ccts of seven ranjor variable.<br />
us exani.ne decomposition effects of each in <strong>the</strong> i.iodel:<br />
(rIlTTPoss) has an effect of .39 on X4 ($ciiaT) of<br />
which .097 or 25 is contributed by or trans:xtitted thouh<br />
X3 (scjitpaop).<br />
x2 (.iwIu) has an effect of -.09 on of which .021<br />
or 22 is tranuitted via X,. You uay notice that in<br />
<strong>the</strong> next two variables (in Table 35 X. and X7, direct<br />
and indirect effects) counteracted esch o<strong>the</strong>r. These<br />
are clear cases o± supressor effects so that <strong>the</strong> total<br />
effect is less than <strong>the</strong> sum of <strong>the</strong> absolute effects.<br />
Also, some components appeared to be larger than <strong>the</strong><br />
total effect. A solution to this kind of situation<br />
(wiere components ore laxer than <strong>the</strong> total effects) is<br />
Siflij)ly to express <strong>the</strong> various couponents as proportions<br />
of <strong>the</strong> sum of <strong>the</strong>ir absolute.<br />
In <strong>the</strong> light of IP.ie explan: tion above, X1 may nay; be<br />
interpreted as hvin its effect on Xr (aLss/Px) via<br />
X3 and X4. In <strong>the</strong> process, .204 or is contributed<br />
by or trans1itteCi via X7 and .016 or 4.7 via OflO.<br />
finally .12 or 35.3 is ur.mediotcd in o<strong>the</strong>r words, this<br />
is <strong>the</strong> real effect of AT?0Z on 0IS.<br />
In <strong>the</strong> sane vein, X2 has on <strong>the</strong> whole very little<br />
direct effect on and in doinG so .044 or 8 is<br />
transmitted via X... and .004 or 8 via X.<br />
p<br />
9.
(v) Finally, REIIGIoII nd CIE3S have stronç direct influence<br />
on ehilclzen1 $ attitude The indirect influences are<br />
negligible.<br />
2he purpose of this chapter was to cain in ins:LCht into seine<br />
jecific aspects of <strong>the</strong> school environment that would enable us to make<br />
<strong>the</strong> assertion that school children who have had more exposure to <strong>the</strong>se<br />
idenbiied school qualities acquired more desirable attitudes. In this<br />
attem)t, we focussed on school adninistration and structural facilities<br />
both irwide <strong>the</strong> classroom arid in <strong>the</strong> school as a whole to measure <strong>the</strong><br />
s ehoolin environment The firidin supported our hrpo<strong>the</strong>sis that struc-<br />
tu.ral features of schools contribute substantially to school ehildrents<br />
attitude in Ghana. In addition, evidcnco vao produced to show that <strong>the</strong><br />
rclationn;ap between quality of school environment and desirable attitude<br />
orientation is lar;ely independent of several almrxiate environmental<br />
)ras tilat also influence a.tuci.ea.
COiTCI1USIOTS<br />
With <strong>the</strong> goal of assessing <strong>the</strong> impact oL' schooling on children's<br />
attitudes, our research has investigated some environmental properties of<br />
schools which to our mind pose problems for policy makers and for child-<br />
ren who pass through <strong>the</strong> school system. In doing so our basic interest<br />
was on <strong>the</strong> structural dimensions of elementary schools; but we also<br />
examined end assessed some home factors in order to weigh <strong>the</strong> relative<br />
contribution of both environments of ter <strong>the</strong> ini'lucnce of o<strong>the</strong>r competing<br />
elements such as urbaWrural origin, sex, end religion had been considered.<br />
By doing this, vie were able to ascertain <strong>the</strong> independent effect of school-<br />
ing emperience in order to give support to our <strong>the</strong>oretical stand that<br />
schooling affects in a definite, positive direction <strong>the</strong> attitudinal<br />
values of children and that this is quite independent of o<strong>the</strong>r generally<br />
accepted actore.<br />
The results emphasized, though by inference, <strong>the</strong> importance of<br />
school quality (apart from normal curriculum effects) in exerting in-<br />
fluence on <strong>the</strong> formation o± desirable attitudes:<br />
1.<br />
Elementary school children attain higher attitude levels<br />
than non-school children of comparable socio-eeonoraic<br />
status and age. Differences in <strong>the</strong> mean scores between<br />
school and non-school groups is significant.
-<br />
r<br />
2 The means of <strong>the</strong> various Tades show that <strong>the</strong> longer<br />
citLidren have been exposed to schoci influence <strong>the</strong> better<br />
<strong>the</strong>ir attitude scores. Thin finding is supported by<br />
results from PICA and path analysis, thus discounting any<br />
clairna that significant diforences observed could be due<br />
-to chance elements,<br />
The quality of physical structures of <strong>the</strong> child's school<br />
environment correlates with <strong>the</strong> child's attitude score.<br />
The positive association between schooling and individual<br />
attitude attainment cannot be explained or interpreted<br />
entirely by o<strong>the</strong>r alternative forces such as <strong>the</strong> home A<br />
causal model was used for <strong>the</strong> analysis, with home influence,<br />
sex, religion, age, and urhaWrural ori4n, controlled<br />
separately. School vaiabls were foiaid to maintain <strong>the</strong>ir<br />
independent association in <strong>the</strong> Lace of all competing<br />
factors.<br />
In affluent societies, <strong>the</strong> home environment is comonly<br />
associated with -<strong>the</strong> child's process of acquiring desirable<br />
attitudes and is found to exert greater influence on <strong>the</strong><br />
child than are o<strong>the</strong>r environmental factors. This study,<br />
however, could not support this finding. What we<br />
observed from <strong>the</strong> path analysis was that parental educa-<br />
tion and wealth had little direct effect on individual<br />
attitudes. But <strong>the</strong> path analysis put -this finding in a<br />
different perspective by showing that home influence was
DISCUESIQt<br />
- 59 -<br />
manifested in <strong>the</strong> choice of schools by parents. Also<br />
observed was that children who were high on hone items<br />
were equally high on school variables.<br />
The research has raised quite a number of issues, <strong>the</strong> implica-<br />
tions of which should concern policy makers as well as educational<br />
planners.<br />
Our stand on ga<strong>the</strong>ring information from young children was<br />
made clear in <strong>the</strong> rationale for <strong>the</strong> study. We have succeeded in<br />
collecting useful data from children between six and twelve years of age,<br />
although o<strong>the</strong>r difficulties have emerged in <strong>the</strong> process, compounded by<br />
<strong>the</strong> current national debate on how many years of free and compulsory<br />
edueati <strong>the</strong> state should provide. Some of <strong>the</strong>se problems are: o<br />
what degree would children o1 ages six to twelve be able to learn and<br />
practise what <strong>the</strong> schools have to offer with regard to attitude? Will<br />
passage of time between <strong>the</strong> £irt terminal point in schooling and adult<br />
life produce retention loss to such an extent that <strong>the</strong> validity of<br />
formal basic education could be qusstianed?<br />
In recent years, three committees have reviewed <strong>the</strong> structure<br />
and content of education in G-iiana. Bach eommitee (1rmiissah, Dzobo,<br />
.&rapene) recommended nine years of basic education, which is <strong>the</strong> asme as<br />
<strong>the</strong> functional literacy level in UiTBSCO terminology. O<strong>the</strong>r organiza-<br />
tions have recommended seven years. The reports from <strong>the</strong>se eommittee<br />
adequately defined basic education in terms of policy and content, but
- GO -<br />
were ra<strong>the</strong>r vague on prescribing what school outcomes <strong>the</strong> s-bate expects<br />
its educational institutions to achieve at each terminal point. (kie<br />
could <strong>the</strong>refore challenge <strong>the</strong> basis on which <strong>the</strong>se bodies fixed <strong>the</strong><br />
first terminal point at seven or nine years of schooling.<br />
The results reported in <strong>the</strong>ce pages showed that improvement in<br />
-i-bUC1O scores was observed in every successive year of schooling for<br />
all grades. We fur<strong>the</strong>r observed that attitudinal influence of <strong>the</strong> home<br />
is less effective than school influence.<br />
In light of this evidence it is obvious that <strong>the</strong> unfortunate<br />
child whose education terminates at class six or sevens when learning<br />
has not ye taken root, will be thrown back into a less challenging<br />
environment where lessons learned in school will be subject to <strong>the</strong> inter-<br />
Lerence of non-controlled influences. Retention loss due to many inter-<br />
actin: fo:ccs is possible. Linc ycors free education is laudable, but<br />
if <strong>the</strong> rcsent state of <strong>the</strong> economy ceuiot sustain free and compulsory<br />
education beyond six or seven years of schooling, <strong>the</strong>n a possible alter-<br />
native is to persuade institutions of production such as <strong>the</strong> Volta<br />
Aluminium Company (vAIoo) and <strong>the</strong> Cocoa Liarketing Board (CLrn) to perform<br />
part of <strong>the</strong> educational functions.<br />
Ano<strong>the</strong>r finding showed that home influence is mediated througla<br />
<strong>the</strong> school, that is, parents are willing to create good. educational<br />
environments for <strong>the</strong>ir wards. This is indicated by <strong>the</strong> fact that child-<br />
ren with high home environment scores were ecually high on school quality<br />
responses. This finding brings into focus issues of <strong>the</strong> boarding school<br />
system in Ghana, where tuition is free and boarding fees are heavily
- 61 -<br />
subsidized by <strong>the</strong> State. Since <strong>the</strong>re is enough evidence that <strong>the</strong><br />
affluent in <strong>the</strong> society are williug to provide quality schools for <strong>the</strong>ir<br />
wards we believe that boarding costs should become <strong>the</strong> responsibility of<br />
parents. This would, of course, limit access by poorer funilies to <strong>the</strong>se<br />
schools, and many people have expressed concern that such a policy would<br />
benefit only children of <strong>the</strong> elite and, in turn, retard democratization<br />
of education in Ghana. We, however, differ in our aesessuent. The<br />
ratio between boarding costs and funds for textbooks is six to one.<br />
Savings derived from <strong>the</strong> boarding system could be diverted to provide<br />
classrooms and badly needed textbooks for children who would o<strong>the</strong>rwise<br />
have been deprived of educational opportunities.<br />
Two major factors have emerged: that physical environment is<br />
an important aspect of <strong>the</strong> school and. in need of reform, and that<br />
improvement in attitude performance is eonnulative and <strong>the</strong>refore one<br />
should look at <strong>the</strong> total schooling effect ra<strong>the</strong>r than at <strong>the</strong> less<br />
important yearly gains. With this we conclude that our attempts have<br />
clarified, <strong>the</strong> role of foxnal elementary education in exerting a positive<br />
influence on children's attitudes.
ATOTIThR VIEW 011 AITALYTIOAIJ SLAGY<br />
Kahi (1963) and Inicciec and Smith (1966, 1974) defined what<br />
<strong>the</strong>y termed 'Individual Liodernity' as a set of attitudes and values that<br />
formed a cohesive syndrome. They presented several <strong>the</strong>mes which pre-<br />
supposed composite scales. Since we adopted some of <strong>the</strong> scales and<br />
adapted o<strong>the</strong>rs to depict a school child having what we described as<br />
desirable attitude qualities in our culture, it was necessary to treat<br />
each subs cale independently and <strong>the</strong>n orm an index of <strong>the</strong> components to<br />
measure pupils on <strong>the</strong>m. The issue of scale homogeneity is examined in<br />
Appendix II, where we deal with realiability and validity of Inkeles and<br />
Smith's 0115 scale in <strong>the</strong> Ganalan cultural background. The purpose of<br />
this section <strong>the</strong>n, is to examine come methodological and analytical<br />
issues and, on tie basis of <strong>the</strong>se, outline our data analysis procedures.<br />
Preliminary results of simple descriptive statistics axe also reported.<br />
It is intended that this pilot study will be followed by cross-<br />
sectional and longitudinal investigations. For <strong>the</strong> three studies a con-<br />
to1 roup (schoolers and non-aehoolera) deci was adopted. With re6ard<br />
to <strong>the</strong> present study, sampling was done with varying degrees of random-<br />
ness. Pumils were randomly selected at <strong>the</strong> classroom level while<br />
teachers remained intact with <strong>the</strong>ir classes. Stepwise regression pro-<br />
cedures and analysis of variance and covaxianco were intended to be used<br />
in processing <strong>the</strong> data. At <strong>the</strong> initial stages, tb present data were<br />
in fact, subjected to variance and covariance analysis but <strong>the</strong>se were<br />
later abandoned when we discovered <strong>the</strong> existence of some methodolo-<br />
I ncnciI RhAyHI 4 - iicnn steni.ea re bound to emere
glaringly when we embark on cross-sectional end longitudinal studies.<br />
But methodological and analytical approaches should be consistent in all<br />
<strong>the</strong> studies, since we envisage a more comprehensive investigation con-<br />
cerned with determining relationships among home environment variables,<br />
classroom variables, teacher characteristics, end curriculum outcomes<br />
in terms o achievement and attitude during <strong>the</strong> second and third phases<br />
o <strong>the</strong> study.<br />
Unortuna-bely, this is not going to be so i we stick to our<br />
origi.nal strategy. During our recent interview sessions, pupils were<br />
randomly selected from classroom to classroom while <strong>the</strong>ir teachers, who<br />
also formed an integral part o± <strong>the</strong> experiment, were not randomized.<br />
This and many o<strong>the</strong>r site observations raised issues of methodology end<br />
analysis. We wondered whe<strong>the</strong>r <strong>the</strong> investigations should adhere to<br />
methodological controls and be carried out under traditional experiment<br />
conditions or whe<strong>the</strong>r we should carry out <strong>the</strong> studies under natural con-<br />
ditions using intact classrooms with <strong>the</strong> sane teachers in charge of<br />
classes as before, and make no attempt in zubseç.uent studies to shift<br />
s-udnts to different learning conditions in order to satisfy traJ.itional<br />
experimental conditions.<br />
Keeves end Iwis (1979) provided good reasons to conduct<br />
studies using intact classrooms. ke first is ropresentativoness. Tke<br />
argued that it might not be advisable to use oxeriuental situations of<br />
randomly assigning teachers and ngils to cxpermcntal groups or iii an:;<br />
way mazipulate teacher behaviour, because such situations have little in<br />
eomiion with complex natural environments.<br />
The; sug,;ested that intact<br />
classrooms in real situations should be aJ2..oied to covery<br />
<strong>the</strong>y do in<br />
<strong>the</strong> natural setting. This approach, <strong>the</strong>y felt , iight simplify data. collec-
- G4 -<br />
The Choice of intact clairooms, however, raised three major<br />
.osuco, namely: appropriate procedure or data analysis, si.nificonce<br />
tes-m.n, and ucii- c analysis. .Any ansvicr to tno last asuo must cc<br />
baiod on t1e type of inferance to be Oavin about lcarnin and teaebin;<br />
conditions ander ctrvcy. That is, vihe<strong>the</strong>r concern is about indidual<br />
or 'roupa.<br />
Prou <strong>the</strong> wsr we carried out sannlin of pupils and involved<br />
toachero Oil <strong>the</strong> measures, vie could not claimi to have carried out random<br />
sanj1na: to i-es logical end. It a not pocsble now, mowrng conditions<br />
in Ghana, to set up a fully doained oporinent on a lare scale, in<br />
vthich all <strong>the</strong> participants are e1ec-tcd at random from a population and<br />
in which teachers ore assiGIled at random, liaving first been randomly<br />
oclocted from a defined populatien. To jus-ti±'y <strong>the</strong> use of analysis of<br />
variance, <strong>the</strong>refore, is difficult; and since rye did. not randomly assign<br />
pupils -to treatment groups (ti-mt is, to <strong>the</strong> various predictors of <strong>the</strong><br />
study), it was likely that <strong>the</strong>re were differences between groups before<br />
<strong>the</strong> application of <strong>the</strong> treatment. This situation would happen in intact<br />
classroom setting as viell as ecperimental situations. Under <strong>the</strong>se circum-<br />
stances <strong>the</strong> use of simple analysis of variance procedures will clearly<br />
produce unsound results. An alternative to this procedure could be to<br />
obtain seas measure of <strong>the</strong> level of performance of <strong>the</strong> ioups prior to<br />
<strong>the</strong> administration of <strong>the</strong> treatment and to adjust <strong>the</strong> criterion scores<br />
by analysis of covcmriance procedure - a method likely to be used in <strong>the</strong><br />
lonitudinal study. Again, <strong>the</strong> use of <strong>the</strong> covariance analysis in this<br />
way wan attacked by 3U.ashoff (1969) and by Cronbach and Purby (1970),<br />
because <strong>the</strong> assumptions tuiderlying <strong>the</strong> use of covariance analysis would<br />
be contravened.
- 65 -<br />
The problem still i'eiIains of how boat to analyse <strong>the</strong> data in<br />
order to obtain some estimate of <strong>the</strong> effects of <strong>the</strong> predictors of our<br />
study or estimate of change in performance between two testing periods<br />
for <strong>the</strong> loiitudinal study. Linn and Kent (1069) sug:e a ted four<br />
approaches:<br />
1 The partial correlation between (A) rand (N) with <strong>the</strong> prior<br />
performance variable (R) controlled /ui.PJ (Fig. A-i).<br />
2, Part correlation between (A) and (N) with <strong>the</strong> effects of<br />
<strong>the</strong> prior performance variable (R) removed only om (A)<br />
[A/)NJ. Tgs is <strong>the</strong> square root of <strong>the</strong> marginal<br />
variance or unique variance of (u) explained by (A) in a<br />
stepwise regression analyse with both (H) anó. (A) entered<br />
as predictor variables and (N) as criterion variable. (Fig. A-2).<br />
Part correlation between (A) and (kr) with <strong>the</strong> effects of<br />
<strong>the</strong> prior performance variable (H) removed only from (u)<br />
This is <strong>the</strong> analysis of covariance generally<br />
uaed in studies of teaclthig behaviour (Fig. A-3).<br />
The standardized partial regression weight of (II) on (A)<br />
obtained from <strong>the</strong> regression equation in which <strong>the</strong> rredic-<br />
tor variable (A) and <strong>the</strong> prior performance variable (H)<br />
are both included in <strong>the</strong> regressIon equation with (r) as<br />
<strong>the</strong> criterion variable(NA.R3 (Fig. A-4).<br />
It now remains to state <strong>the</strong> form of our problem to tal:e into<br />
account <strong>the</strong> pilot study, as well as strateiies for <strong>the</strong> two studies to
Fig. A-i: Procedure 1<br />
Fig. A-3: Procedure 3<br />
Fig. A-2: Procedure 2<br />
Fig. A-4: Procedure 4<br />
ollcvi. Tue problem restated is that we iaave aisdo independent or predic-<br />
lor v b1es (A) associated with <strong>the</strong> home, school, teacher characteristics,<br />
and teacPinr behaviour prior in time sequence to <strong>the</strong> criterion variable<br />
(::), viith a measurement taken on a variable to assess previous per±ormancc<br />
(n), viiich is in turn antecedent to (A) and (IT). This accepted, <strong>the</strong> reso-<br />
ltLon o <strong>the</strong> problem rests with indi appropriate procedure that will<br />
establish causal relationships ith prior Ie'orLance (R) and both <strong>the</strong><br />
rcdictor variable (A) and tlie criterion variable (IT). 7e <strong>the</strong>n crue that<br />
<strong>the</strong> principle oi sequence operates in <strong>the</strong> war vie sot up <strong>the</strong> study and that<br />
prior percrmance () exerts impact on <strong>the</strong> predictor variable (A) and that<br />
(h) has influence on <strong>the</strong> criterion variblc (::), and conclude that <strong>the</strong><br />
outh rrocedure is <strong>the</strong> most aT)propriate to use See iLure 4.
- 67 -<br />
Procedure 2 is unsuitabl because it takes no account of <strong>the</strong><br />
influence of <strong>the</strong> prior performance on <strong>the</strong> criterion when considering <strong>the</strong><br />
relationship between <strong>the</strong> criterion and predictor. In. any naturalistic<br />
investigutions, where nei<strong>the</strong>r individual students nor intact classrooms<br />
have been randomly assigned to treatment grouns, we cannot say with any<br />
certainty that prior performance does not influence teacher characteris-<br />
tics in <strong>the</strong> classroom on <strong>the</strong> behaviour of teacher and his pupils in <strong>the</strong><br />
classrooms; and from <strong>the</strong> way our schoOls function <strong>the</strong>re is bound to be a<br />
relationship betveen <strong>the</strong> prior performance variable (R) and tue predictor<br />
varia'ole (A), which we call treatment, so that <strong>the</strong> cova.riance analysis<br />
of procedure is inappropriate. What this, in. effect, means is that,<br />
unless <strong>the</strong> correlation between prior performance variable (R) an.d pre-<br />
dictor variable (A) is zero, <strong>the</strong> use of covariance is questionable.<br />
In Procedures I and 4 <strong>the</strong> effect of prior performance has been<br />
partialled out from predictor variable (i) and <strong>the</strong> criterion variable (ir).<br />
Regression analysis (though not stepwise) is preferable because <strong>the</strong><br />
results are more readily understood.<br />
Suiimary<br />
In summing up we have argued that <strong>the</strong> research study i <strong>the</strong><br />
type in which <strong>the</strong> prior performance variable (it) serves as antecedent<br />
to <strong>the</strong> predictor variable (A) and <strong>the</strong> criterion variable (N) and exerts<br />
influence on both. In pursuit of this causal model our analysis will<br />
involve obtaining descriptive statistics a1 <strong>the</strong> initial staos followed<br />
by IICA and path analysis procedures.
- 60 -<br />
APP11TDIX II<br />
RIIMABILITY Aim VALIDITY 02 Ti $CAIES<br />
ih-t attitude scales were used in this study. Some of <strong>the</strong><br />
iteiis we used vero taken directly from Inhales and smith' s 0113 scale;<br />
end o<strong>the</strong>rs were adaptations of it in order to fit cultural differences.<br />
There is no doubt that <strong>the</strong> oriinal 011 scale was refined to<br />
obtain Iaifh reliability end validity. We should, however, remember that<br />
<strong>the</strong> scale was used about 1972 and this means en cifht to ten years diffe-<br />
rence between <strong>the</strong> study referred to end ours, which is yet to start on<br />
national basise It is true also that both research studies are located<br />
in less affluent societies and <strong>the</strong> Ghanaisn one is tahin place during<br />
a decade of rapid social chone In those circumstances, concern has<br />
been oxressed about using <strong>the</strong> same scale for two distinct ae croups<br />
(ciiool children and working adults) even t1iouh <strong>the</strong> literature revealed<br />
that <strong>the</strong> eale had been used succeosfally on children elsewhere. This,<br />
to critics pointed out, mi1t affect <strong>the</strong> reliability Of t1e scales.<br />
iccordinfly, it is desirable to ensure constancy of <strong>the</strong> scales<br />
over tme Per internal consitcncy, <strong>the</strong> alpha eoeidcient - <strong>the</strong> most<br />
poiuiar internal consistency measuro - was surjesued and in .act, used<br />
at one atafe in <strong>the</strong> analysis. But this poses dificultics. Jovic anu<br />
Iviis (1967) have demonstrated that alpha is not eciul to <strong>the</strong> reliability<br />
of cc osite scores, but instead is a lower bound to it. They fue<strong>the</strong>r<br />
provad that if <strong>the</strong> items ore tau-equixalent, that is, <strong>the</strong> true scores oa<br />
one item differ from <strong>the</strong> true scores of ano<strong>the</strong>r item, just by a constant,
- G9 -<br />
<strong>the</strong>n alpha is an exact estimate oJ' <strong>the</strong> reliability. Heise and Bohrnctedt<br />
(1970) derived <strong>the</strong> statistic omega and proved that whenever items are<br />
tau-eciuivalent alpha and omega should equal each o<strong>the</strong>r Under this con-<br />
dition inter-item correlations are equal and item variances arc also equal,<br />
o<strong>the</strong>rwise omega will be larger than alpha. We have subscales and <strong>the</strong>re-<br />
Lore decided to use omega statistic in estimating reliability of our<br />
subscales.<br />
Apart from reliability issues <strong>the</strong>re is also <strong>the</strong> riroblem of<br />
validity. We expect that differences in children's score, to a large<br />
extent, represent differences in <strong>the</strong> attitude we are investigating. But<br />
attitudes are abstractions we ixner from case responses and if, as we<br />
claim, social values fluctuate, <strong>the</strong>n it is lihely that <strong>the</strong> constructs we<br />
are trying to elicit are unstable said for that matter <strong>the</strong> scales pr<br />
measure different constructs.<br />
In view of all this, vie have decided to look at validity of <strong>the</strong><br />
eubscales also. We used a very simple strategy:<br />
1. We factor analysed a set of twenty-eight items to determine<br />
<strong>the</strong> relationship of items underlying factors or causes.<br />
ctors which are strong indicators (based on factor 'load-h<br />
i*e)Lormed linear composites,. The nun<strong>the</strong>r of items used<br />
agreed with Nunallys auigection to use items, when<br />
<strong>the</strong>y arc to be factored simultaneously (iTunally. 1 9G7:<br />
257).<br />
2. Reliability and validity indices of <strong>the</strong> composites were<br />
estimated by using Ileize and Bobrnstedt (1970) derivations.
:si Data<br />
- 70 -<br />
There was consensus amoi intorvLcwers that <strong>the</strong> oriGinal over-<br />
all ocule was too lon., and ebildren became tired during <strong>the</strong> interview<br />
sessions. Vie <strong>the</strong>refore conputed estimates for half <strong>the</strong> total number o±<br />
ienc. niosri data on triese ienc were relatively small ranin appro-<br />
:Laatoly between zero and three per cent e::cept on item twelve where <strong>the</strong><br />
perce11ta:e for missi data was hiGh.<br />
2hL.CILtL c0: :.JT0 £ALiIS<br />
Analysis<br />
incipal components analysis without iteration and varima:<br />
rotations wore used. Tha number 0±' factors was selected to correspond<br />
with <strong>the</strong> nu bcr oL' scales into which <strong>the</strong> items had originally becn put.<br />
7on <strong>the</strong> results of <strong>the</strong> factor analysis, a revised clustorini for each<br />
set of items was carried out. It would have been better if principal<br />
couponent analysis were aain carried out on <strong>the</strong> revised clustering<br />
for any aisplaced items. This was not done The number of subfactors<br />
in each scale was determined simply by <strong>the</strong> highest positive value load-<br />
Qenerally, <strong>the</strong> number of sub±'actors viitithi. each scale is deter-<br />
mined by <strong>the</strong> iiser criterion of acceptinG principal components with<br />
latent roots Greater than one.<br />
The factor analysis revealed that <strong>the</strong> number o± principal<br />
couponents with latent roots Greater than one were more than <strong>the</strong><br />
:nuired number of factors. Probably, this was due to factor analysinG<br />
at <strong>the</strong> item level. The table 4.0 shows <strong>the</strong> factor structure of <strong>the</strong> item.
Attitude 1<br />
1-6<br />
Number o± principal cornponents with lctent roots 1 12<br />
Percentaie of variance explained by <strong>the</strong>m 63.0<br />
Number of principal compients with latent roots 2 2<br />
Itent root of <strong>the</strong> irst principal component 2.33<br />
Original item<br />
clusterinç<br />
- 71 -<br />
Table 4.0 FACTOR SUCTL OF ITiii<br />
Table 4.1 INCIPA1 OOI0NENT ANALi OP IT<br />
Pacto loadinjs without value, restricti-is<br />
1 2 3 4 5 6<br />
Factor Factor Factor Factor Factor, Factor<br />
1 .478 .24<br />
2 .409<br />
3 -.22 .587 .213<br />
4 .262 .655<br />
5 .521 .227<br />
6 .25 -.28<br />
Attitude 2<br />
7 - 13<br />
7 .296 .330<br />
8 .401 -.223 .'13 .292<br />
9 .217 .511<br />
10 .304 .43<br />
11<br />
.269<br />
12 .233 .265 -.367 .39<br />
13<br />
.457
iinal item<br />
c1utering<br />
Attitude 3<br />
14 - 19<br />
- 72 -<br />
Factor loQdins without value i'estriction<br />
1 2 3 4 5 6<br />
Factor Factor Factor Factor Factor Factor<br />
14 -.421 -.354 .224 .215<br />
15 .17 .483<br />
16 -.453 .233<br />
17<br />
16 -.28 .631 .567<br />
19 .733 .332<br />
Attitude 4<br />
-<br />
20 -.97 .263 .429<br />
21 .374 .244 -.205<br />
22 -.19 .402<br />
23 .747<br />
24 .735<br />
25 .164 .536<br />
26<br />
27 .11 .495 .534<br />
2 -.197 .19 .557<br />
The analysis incateO. that all he aubscales could be etic1od<br />
heterogeneous. This observation does not preclude co: posite scalco
- 7$ -<br />
from being a measure of some general factor. ieterogeneity, however,<br />
prevents <strong>the</strong> use of alpha, a substantial level of vihich shows horl1oeneity<br />
and high correlation among aubscales. This situation emphasizes our<br />
decision to treat each subscale thdependently. With <strong>the</strong> nunber of f:ctora<br />
restricted to to solutions are shown in tne aole 4.1. ily loauiigs<br />
greater than 01 ecluaJ. to .11 are reported.<br />
Scale Revision<br />
Based on <strong>the</strong> factor analysis, tue attitude scale is revised as below;<br />
Table 4.2 PRLTCIPAL 0OiL0IITT SOLJTIO1TS iR FACT 1 (o:uuris)<br />
Item Factor structure Factor score coefficient<br />
1 .473 .221<br />
5 .521 .257<br />
(_J . .177<br />
13 .457 .220<br />
25 .164 .039<br />
The first factor here is Openness to New Experience. ly two of <strong>the</strong><br />
original six items loaded on this scale. The remaining three items<br />
were a nixture of items from Attitudes 2 and 4 of tixe original subcales.<br />
A departure from <strong>the</strong> scoring specifications is shown by <strong>the</strong> negative<br />
loadings on items 14, 16 and 20, indicating that pupils with favourable<br />
attitudes to o<strong>the</strong>r items in this scale tended to disa.gree with <strong>the</strong>se<br />
three. It is possible that <strong>the</strong> key scorings are inappropriate. Despite<br />
-<strong>the</strong> high absolute loadings on <strong>the</strong>m, <strong>the</strong>y are deleted from <strong>the</strong> present<br />
computations.
- 74 -<br />
Table 4.3 IINCAL C0L2OflLNT SOLUTION P 1?ACT 2 (REDIImss<br />
P OCIAI ciLkia.E)<br />
I tern Factor structure Pactor score coefficient<br />
7 .330 .192<br />
10 .304 .159<br />
11 .269 .144<br />
20 .263 .110<br />
21 .374 .203<br />
2 .747 .399<br />
24 .735 .385<br />
The second factor loaded on four items of <strong>the</strong> oriina1 scale<br />
as well as on four items from <strong>the</strong> oriinal Attitude 4 (Confidence in<br />
One's Ability and <strong>the</strong> Ability o± in ...). Children appeared to have<br />
uced different criteria in responding to <strong>the</strong> scale items. To Inkeles<br />
nd Smith, <strong>the</strong>se are efficacy questions end refer to man's potential<br />
for mastery over nature and <strong>the</strong> sense that one can effectively do some-<br />
t1thv in concert with o<strong>the</strong>r men to br about chances.<br />
Let us examine <strong>the</strong> items in question. Item 21: 'If you were<br />
president what would you do?' Children tended to answer this question<br />
by indicating changes <strong>the</strong>y would like to see take place in <strong>the</strong>ir corn-<br />
aunity, end seemed to have avoided our idea that a favourable response<br />
meant personal confidence.<br />
Item 23:<br />
'Do you think that we can widerstand <strong>the</strong> causes of
- 75 -<br />
thunder, rainfall and lighting?' and Item 24: 'Do you think we can<br />
w'iderstand how a seed turns into a plant?'<br />
WhAt children seemed to have done to <strong>the</strong>se questions was to<br />
respond to <strong>the</strong> first part - - that changes were boimd to come through<br />
man - - but did not see <strong>the</strong>moelvea, at <strong>the</strong>ir tender age, as part of<br />
<strong>the</strong> solution.<br />
Table 4.4 IIIPAL C01T SOLUTION PW FACTOR 3 (GR1TH<br />
op oPINION)<br />
Item Factor structure Factor score<br />
2 .409 .229<br />
3 .567 .327<br />
4 .360 .36<br />
6 .278 .278<br />
12 .265 .138<br />
15 .17 .11<br />
27 .11 .05<br />
Factor 3 derived as many as four items from <strong>the</strong> original<br />
Attitude 1 and two o<strong>the</strong>r items from original Attitudes 3 and 4. The<br />
items that switched from Attitude 1 were investigated. They all<br />
involved willingness on <strong>the</strong> part ci' <strong>the</strong> respondent to move from <strong>the</strong>ir<br />
present comniunity to ano<strong>the</strong>r place far away. The items failed to con-<br />
sider one crucial element, and that was <strong>the</strong> reasons for changing <strong>the</strong>ir<br />
present locations. It is felt that if interviewers had made it abun-
- 76 -<br />
dantly clear that rnovin to a new environment promised better life, <strong>the</strong><br />
expected trait might have beeii tapped. Consequently, readiness for <strong>the</strong><br />
new experience which we were see1d..n should depend on chances 01' improved<br />
tandard of living.<br />
Table 4.5 INCIPAL C0UT S0LtJI0IT P PACT 4 (EFFIcACY)<br />
Item Factor structure Pactor score<br />
9 .511 .285<br />
18 .631 .360<br />
Factor 4 seemed not to have loaded on any items from <strong>the</strong><br />
oi;iiactl scale But a close exathiation of <strong>the</strong> efficacy construct<br />
rovalc two components: Confidence in one1s ability and confidence in<br />
<strong>the</strong> ability of man. Naturally, <strong>the</strong> factor analysis split <strong>the</strong> items<br />
noatly to account for <strong>the</strong> two factors. For <strong>the</strong> purposes of this<br />
exercise, however, only one factor is used.<br />
.733 .409<br />
28 .185 .18<br />
We present <strong>the</strong> reliability and validity results as followa:
etora<br />
- 77 -<br />
Table 4.6 8TJ.IMARY TABI POR (UGINAL SCAIE<br />
Attitude 1 .41 .65 .25<br />
Attitude 2 .21 .63 .07<br />
Attitude 3 .29 .60 .33<br />
Attitude 4 .43 .65 .06<br />
Table 4.7 StJDffARY TABE POR IiVIS SOAIE<br />
ug'eStion<br />
Pactors<br />
Pactor 1 .56 .74<br />
Pactor 2 .66 .68<br />
Pactor 3 .63 .78<br />
Pactor 4 .65 .84<br />
Since Cronbach's alpha is an exact estimate of reliability only under<br />
tau-equivalent assumptions of equal intercorrela-bions and equal<br />
variances, <strong>the</strong> use of Heise and Bohrnstedt's omega is preferable.<br />
Table 4.6 compares alpha and omega reliabilities and <strong>the</strong> latter esti-<br />
mates exceed alpha as expected.<br />
Reliability Validity<br />
Alpha Omega Pta<br />
Reliability Validity<br />
Omega Pta
- 78 -<br />
Evidence from <strong>the</strong> factor analytical results shows that <strong>the</strong><br />
traits are correlated, arid that sonic i-art of <strong>the</strong> correlation may be<br />
due to <strong>the</strong> influence of o<strong>the</strong>r traits that are correlated with that of<br />
interest. We feel, however, that this is not <strong>the</strong> case with some of<br />
<strong>the</strong> items singled out earlier We think that children have not used<br />
ti-ic acne criteria that adults would in constructing <strong>the</strong> scales.<br />
i-actor loadings chow <strong>the</strong> cxtont to which latent traits influence<br />
scores d if factors are out of alinont with latent traits, factor<br />
loadings have little meaning. It is <strong>the</strong>refore desirable to take a<br />
fresh look at item clustering to ensure that factors and latent traits<br />
are adequately aligned.
ITotes to <strong>the</strong> Class Teacher<br />
Name<br />
- 79 -<br />
APPiDIX III<br />
PCRIBID VALtY2S AID ATTITUDES SCALE<br />
Purpose: The purpose o± -<strong>the</strong> questions and statements in <strong>the</strong>se<br />
paies is to collect from some of your pupils information which<br />
can be used to improve <strong>the</strong> quality of education in our primary<br />
schools. The exercise wi.l1 in no way affect adversely <strong>the</strong><br />
performance of <strong>the</strong> pupils in <strong>the</strong>ir school work ei<strong>the</strong>r now or<br />
later.<br />
Request: We are requesting that you help us acheinister very<br />
carefully <strong>the</strong> instrument to <strong>the</strong> pupils concerned niakin sure<br />
<strong>the</strong>y tderstand each item before responding to it. It will<br />
take only about 30 rinutes to administer <strong>the</strong> instrument to<br />
each of <strong>the</strong> pupils.<br />
Confidentiality: The names of <strong>the</strong> pupils and <strong>the</strong> schools<br />
participating in this ecercise will remain strictly confidential.<br />
The pupils' individual responses will not be made<br />
public at any time.<br />
Age (in years)<br />
SECTION I: BJ.CKGR0IJ1D ILTORNATION
3'<br />
Parents<br />
Pa<strong>the</strong>r<br />
Mo thor<br />
- 80 -<br />
Sex (Put an "x" in tho appropriate space)<br />
(a) Male Female<br />
Place of birth (?lrito <strong>the</strong> name of <strong>the</strong> village or town in <strong>the</strong> appro-<br />
priate space.<br />
(a) Village (b) Town<br />
Name of school presently attending<br />
Prcoent class in school (Circle as applicable)<br />
Grade I II III Iv V VI<br />
No. of years so far spent in school including <strong>the</strong> present school<br />
year<br />
2.eligion (Put an "X" against <strong>the</strong> pupil's response)<br />
(a) Cbristianity (b) Islam (c) O<strong>the</strong>rs<br />
9, Soojo-econornic status<br />
(a) Parent's level of education (Put an 'IX" in <strong>the</strong> appropriate space)<br />
No Schooling<br />
At AU<br />
Fa<strong>the</strong>r's Occupation<br />
Mo<strong>the</strong>r' a Occupation<br />
LCompleted 'Completed Completed<br />
Primnry Secondary Post-Sec. Don't Xnovi<br />
ducatjon Education Education
I OP2iflISS TO<br />
(d) Parents' affluence (Put an "X" siathat <strong>the</strong> pupil's response<br />
indjoatjri <strong>the</strong> items which are available in his home among <strong>the</strong><br />
following)<br />
i. Oar ii. Television iii. Radio, Player<br />
or Stereo iv. Gas Cooker v. Refrierator<br />
(friöe) vi. Beds<br />
None<br />
Wooden<br />
Iron<br />
vii Books (apart from <strong>the</strong> Bible or <strong>the</strong> Koran)<br />
Toilet facilities<br />
none_____ Palm rond Iterine<br />
Plush (Water system) toilet<br />
The type of houae in which <strong>the</strong> pupil and his parents live<br />
mud house plastered (cemented) house<br />
a furnished flat<br />
SECTION II: RESEARCH DATA<br />
(Suppose) xnaxmy strangers visit your town/village each year.<br />
Do you think this is a good thing?<br />
Good, Bad, Not Sure, Don't Know, O<strong>the</strong>rs.<br />
Suppose your parents ask to choose between going on holidays<br />
with <strong>the</strong>m to (nearby, town) and<br />
going on holidays with <strong>the</strong>m to (dist. town),<br />
which would you choose? Nearby Town, Distant Town
- 82 -<br />
Would you be willing to move from your present school 'to<br />
a school in a far away place where <strong>the</strong> people eat food<br />
and wear clo<strong>the</strong>s that are different from yours?<br />
1Iove, Stay, Don t Know, O<strong>the</strong>r<br />
Some people want to live in town or cities while o<strong>the</strong>rs<br />
want to live in villages. Where do you wont to live?<br />
City, Village<br />
You have your friends at school. A mcvi boy/girl comes to<br />
your school. Would you make hinVher your friend?<br />
Liake friends, Don't make friends, Don't Know<br />
If you could live anywhere you wished, where would you<br />
prefer to live?<br />
Why?<br />
lie R1ADIiE8S POR SOCIAL ChANCE<br />
Suppose someone in your class makes higher marks than you,<br />
what would you say is <strong>the</strong> cause?<br />
- The child wor1 harder than. me - The child ha good luck<br />
- The teacher likes him better - The child has joy in his<br />
work<br />
- Don't Know - O<strong>the</strong>r<br />
Suppose a child your class becomes very ill, where<br />
should <strong>the</strong> child be taken for treatment?<br />
- hospital - churc1/mosque/prophet/jujumari<br />
- herbalist - don't 1ow<br />
- o<strong>the</strong>r
III 0'JTII 0i 0PINIC1<br />
- 83 -<br />
Suppose pour La<strong>the</strong>r said something that you laww is not<br />
true would you correct him?<br />
Yes No Don't iaiow O<strong>the</strong>r<br />
Some people say mummy should cook all <strong>the</strong> time O<strong>the</strong>r<br />
people think daddy should help to cook at times. Vlhat do<br />
you think?<br />
- Mummy cooks all <strong>the</strong> time - Daddy cooks at times<br />
- Daddy cooks all <strong>the</strong> tine - Don't ]rnow<br />
Some people say boys are better th girls. Do you agree?<br />
Interviewer:<br />
No Yes Don' t laaow<br />
IL no, what do you think?<br />
Girls are better than boys<br />
girls and boys are <strong>the</strong> same.<br />
Ii' you score a high marl: on your test in school, what would<br />
you say is <strong>the</strong> cauce?<br />
I work hard, <strong>the</strong> test was easy, <strong>the</strong> Teacher likes me<br />
I don' t 1iow, O<strong>the</strong>r<br />
Then members ol' your c(:e oup disaroe with you, what<br />
do you do?<br />
I ch-e my iiii, I jiold on to my opinion,<br />
Don't 1movi O<strong>the</strong>r
- 34 -<br />
25. :Iot o± <strong>the</strong> time oioi' ieol10 decide on everythfli.<br />
Do you think this ic riiht?<br />
Yes, Don't iaioi, O<strong>the</strong>r<br />
Then peJ?sonc younror than you dicsrec with you,<br />
what do you do?<br />
I ehane ry ithid, I hold on to my o)inion,<br />
Don' t Iaiow, O<strong>the</strong>r<br />
When persona older than you diaaree with you, what do<br />
you do?<br />
I chano my mind, I hod on to my opinion,<br />
t laiow, O<strong>the</strong>r<br />
23 Do you always cay viiict you think?<br />
Yes, 110, Don't .siow, O<strong>the</strong>r<br />
27. If you have somethinr to say do you say it?<br />
Yes, 110, Don't know, O<strong>the</strong>r<br />
V. C0LLUICB LI CUE'S AI3ILITY AiiD A$ILITY<br />
02 i.11 TO ACHI2VE I:IS OBJCTIVE<br />
23. Bad tnins happen. Por example a child ±'allin down and<br />
broaid.n his 1CC while runninr. Is it possible Lor a<br />
child like you to atop such bsd thinCa?<br />
Yes, :To, t know, O<strong>the</strong>r<br />
29. (a) 1±' you were resident/iIeaô. o±' State what would you do?
- 35 -<br />
(b) Why would you do such a thing?<br />
30. Some people are bo poor. Do you believe <strong>the</strong>se poor<br />
people can become rich?<br />
Yes, No, Don t irjiow, O<strong>the</strong>r<br />
31 Do you think thAt we can understand <strong>the</strong> causes o± things<br />
like rainfall, lightning sad thunder?<br />
Yes, No, Dontt 1-iow, O<strong>the</strong>r<br />
Do you think we can understand how a seed turns into<br />
a plant?<br />
Yes, No, Don't haow, O<strong>the</strong>r<br />
(a) Some people say children who don't go to school are<br />
(b) Why<br />
better off than children who go to school. What<br />
do you say?<br />
Children in school are better off<br />
Children out of school are better off<br />
Suppose a Thrnily has child, who steals. Do you think<br />
that family can make that child stop stealing?<br />
Yes, No, Don't haow, O<strong>the</strong>r<br />
Some people believe human beings (Ghenaiaas, higerians<br />
or Liberians or Sierra Leoneans) can do whatever <strong>the</strong>y<br />
decide to do. Do you believe this?<br />
Yes, No, Don't imow, O<strong>the</strong>r
- 06 -<br />
V. i3TJ1' TIIAT OT IOPI AIID IIJSTITLTTIOiS CAIf PIE ILI<br />
132011 TO ?UIJIL TF3EIR OBLIGATIONS AITD EZP0IT$IBIMT]S<br />
Suppose people are fighting nearby. n unlmowri<br />
policeman offers to take you away from <strong>the</strong> area.<br />
Will you follow him?<br />
Yes, Ilo, Dont t iow, O<strong>the</strong>r<br />
(a) When your parents promise to buy t1iins, do you<br />
expect <strong>the</strong>m to do so?<br />
Yea, No, Don' t imow, O<strong>the</strong>r<br />
(b) Interviewer: If yes, probe for frequency,i.e.,<br />
how often?<br />
If a man steals or does something bad, do you<br />
believe he will be caught and punished by <strong>the</strong><br />
government?<br />
Yea, lIe, Don' t laiow, O<strong>the</strong>r<br />
70 Suppose your teacher says your class will pay a<br />
visit to <strong>the</strong> post office (or any place of interest),<br />
would you start making plans for <strong>the</strong> trip?<br />
Yes, No, Don' -t know, O<strong>the</strong>r<br />
40. (a) When members of your ae group promise you things,<br />
do you expect <strong>the</strong>m to fulfil <strong>the</strong>ir promises?<br />
Yes No, Don' t know, O<strong>the</strong>r<br />
(b) Interviewer: If yo5 probe for frequencyi.e.,<br />
how often?
VI. VALUIIIU- OP TECHNICAL SKII<br />
- 87 -<br />
Suppose your dad says he will give you 1 (Dollar, Taira,<br />
Leone, Cedi) today, but if you Wait. fOr 10 days he will<br />
give you 5 (Dollars, ....). That would you do?<br />
Take one dollar, Take five dollars later,<br />
Don't know, O<strong>the</strong>r<br />
You and some o<strong>the</strong>r children have volunteered to clean<br />
dirty roads in your tov,/villa(;e. Do you think <strong>the</strong><br />
otiter children will do <strong>the</strong>ir part of <strong>the</strong> work?<br />
Yea, No, Don't know, O<strong>the</strong>r<br />
Supose you are a farmer.<br />
This year your crop is not crowing well. What would<br />
you do?<br />
Do eomethin oci-uücal, As:: for technical advice,<br />
Do something non-technical, Do nothing, Don't know,<br />
O<strong>the</strong>r<br />
Some people think it is important for a child to know<br />
how to make his or'n toys such as (boats and wire cars).<br />
That do you think?<br />
It is important, It is not important, Don't know,<br />
O<strong>the</strong>r<br />
E*Lxaniple should natch sex of child, e .G. boys<br />
wire cars; girls<br />
baby dolls.<br />
See **2 under Readiness for Social chcne
- 83 -<br />
A lorry has broken down and <strong>the</strong> owner con' t make it run.<br />
Call a mechanic, abandon <strong>the</strong> lorry, pray about it,<br />
don't laiow, O<strong>the</strong>r<br />
Suppose your .ilwo/tovy'city has nei<strong>the</strong>r hospital nor<br />
churcmosque. Your villae/ton/city has money to<br />
build only one of <strong>the</strong>se. Which one would you like to<br />
be built?<br />
Hospital, Ohurcli1osque, Dons t thow, 0<strong>the</strong>r<br />
Interviewer: Probe why?<br />
43. That would you lihc to be when you 3row up?<br />
VIL IhSPL'CT iCi OTI- PL0PI'S DIGNITY<br />
Suppose you did something won in class and <strong>the</strong> teacher<br />
decides to punish you. Where would you prefer to be<br />
punished? In front of your class]nates or in <strong>the</strong><br />
Prmncipalts Office?<br />
In front of classmates, In Principal's Office<br />
Don't ia.ow, O<strong>the</strong>r<br />
Suppose a classmate did. something wroi in class and tile<br />
teacher decides to punish biiy'her. Where would you<br />
prefer he/she to be iuni.shed? In front o± his/her<br />
classmates or in <strong>the</strong> Principal's Offiec?<br />
In front o.f his/her classmates, In <strong>the</strong> Principal's<br />
Office, Done t 1aow, O<strong>the</strong>r
VIII. ATTITuDE TO 1X<br />
-<br />
51 Suppose you did soiaethin vrong in (a public<br />
place, ., lorry park). Should your mo<strong>the</strong>r pimish you<br />
<strong>the</strong>re, or should she wait and punish you at home?<br />
Pt.iish me <strong>the</strong>re, Wait until we Get home,<br />
Don't 1iow, O<strong>the</strong>r<br />
Do you think a school boy/girl ii: you should do work<br />
at home?<br />
Yes, No, Don't lc2low, O<strong>the</strong>r<br />
Suppose a man has many children and sends <strong>the</strong>m all to<br />
school. Should <strong>the</strong> children come home and help with<br />
<strong>the</strong> house work or should <strong>the</strong>ir ±'a<strong>the</strong>r Get a houseboy?<br />
Children should help, Pa<strong>the</strong>r should Get a<br />
houseboy, Don't 1now, O<strong>the</strong>r<br />
Some people think that a person who haa gone to ached<br />
should not cut grass or chop wood. Vlhat do you think?<br />
He should cut grass and chop wood,<br />
He should not cut grass or chop wood,<br />
Don't IcIOW<br />
O<strong>the</strong>r:<br />
55, Do you like to do work at home?<br />
Yes<br />
No<br />
Don' t imow<br />
O<strong>the</strong>r
- 90 -<br />
56. Do you enjoy doii or £Lin things with your hands?<br />
Yes<br />
Don't 1Qww<br />
O<strong>the</strong>r
.1<br />
-<br />
APiTDIX IV<br />
SCHOOI PLSOURCIS I}C BORM<br />
(SCIJ 001 iWIR0llIENT)<br />
:Tote to <strong>the</strong> Resoarcher/iesearch Assistant<br />
A number of statements are presented on <strong>the</strong>se paces. You are<br />
to us <strong>the</strong> statements to check certain conditions and resources available<br />
in some of our primary schools and record <strong>the</strong>m exactly as you find <strong>the</strong>m.<br />
Por a few of <strong>the</strong> items you will find it necessary to interview <strong>the</strong> head-<br />
masters of <strong>the</strong> schools.<br />
It is absolutely necessary to seek <strong>the</strong> ft lest co-operation of<br />
each headmaster in order to obtain all <strong>the</strong> information required. You may<br />
<strong>the</strong>refore have to let each headmaster lmow that <strong>the</strong> purpose of <strong>the</strong> exercise<br />
is to collect information which can be used to irijrove <strong>the</strong> conditions of<br />
our primary schools. It has nothing to do with <strong>the</strong> performance of <strong>the</strong><br />
school or <strong>the</strong> efficiency of <strong>the</strong> headmaster and all information collected<br />
will remain strictly confidential.<br />
T.B. Please put an X" as appropriate against items 3, 4, 5, 6, 7, 8<br />
and 11.<br />
The name of this school is:<br />
The mailing address of this school is
Thic school is located in:<br />
- 92 -<br />
(a) a rural area (b) an urban area<br />
4. Proprietor$hip (or Religious tone) of <strong>the</strong> school<br />
(This school is):<br />
a mission (church) school<br />
a moslem school<br />
a government school<br />
composition of <strong>the</strong> school (This school is):<br />
an all boys school<br />
an all girls school<br />
a mixed school<br />
6. Population of <strong>the</strong> school (The nwnber of pupils in this school is):<br />
Less than 250<br />
above 250<br />
7. Tot o± <strong>the</strong> buildings in this school<br />
i. have nd walls<br />
ii. have concrete (cement block) walls<br />
i. have unceniented floors<br />
have concrete (cemented) floors<br />
(C) i have no ceilings<br />
:11. have mat ceilings<br />
have high quality ceilings<br />
Idbrary facilities available in this school<br />
none<br />
only a reading corner in each class<br />
central (or common) library
- 93 -<br />
9. Recreational facilities available in this school<br />
(a) Sports for which <strong>the</strong>re are separate playi'ounds<br />
1<br />
2<br />
)<br />
4<br />
5<br />
(b) Materials available for sports ni games<br />
1.<br />
2.<br />
3.<br />
4.<br />
5.<br />
Names of materials<br />
10. ilanes of schoola with which this school has played matches and axaes<br />
in <strong>the</strong> last one year<br />
1.<br />
2.<br />
7<br />
4.<br />
5.<br />
Sports and games for<br />
which <strong>the</strong> materials are used<br />
iTames of Schools Ilatches and Games Played
- 94 -<br />
1. Toilet acilitie available in tbi ehool<br />
none<br />
palm ±'rid/pit latrine<br />
(C) flush (water system) toilet<br />
This is <strong>the</strong> end oi <strong>the</strong> inctrurnent but<br />
that every item has been checked end recorded accurately.<br />
please, o over to aake cure
95<br />
Andreas,F.M.; Morgan, J.N.; Sonquist, J.A.; Klein, L. 1973.<br />
Multiple classification analysis: A report on computer<br />
program for multiple regression using categorical<br />
predictors (2nd ed.). Ann Arbor, MI, USA: Institute<br />
for Social Research, University of Michigan.<br />
Blalock, H.M. 1964.Casual reference in non-experimental<br />
research. Chapel Hill, NC, USA: University of North<br />
Carolina Press.<br />
Bloom, B.S. 1964Stability and change in human characteristics.<br />
New York, NY, USA: John Wiley. Pp. 73-190.<br />
Bronfenibrenner, U. et al. 1965. Adults and peers as a source of<br />
conformity and autonomy. Ithaca, NY, USA: Cornell University,<br />
unpublished study.<br />
Coleman, J.S. 1975. Methods and results in <strong>the</strong> lEA studies of<br />
School on learning. Review of Educational Research,<br />
45: 335-386<br />
Cronbach, L.J.; Furby, L. 1970. How we shcxtld measure Ichaflge<br />
or should we? Psychological Bulletin, 7l 68-80.<br />
Dave, R.H. 1963. The identification and measurement of<br />
environmental process variables that are related to<br />
educational achievement. Chicago, IL, USA: University of<br />
Chicago, PhD dissertation.<br />
Elashoff, J.D. 1969. Analysis of covariance A delicate<br />
instrument. American Educational Research Journal, 6(3):<br />
383- 401.<br />
Hauser, RN. 1975. The decomposition of effects in path<br />
analysis. American<br />
Sociological Review, 40: 37-47.<br />
Heise, D.; Bohrnstedt, G.W. 1970. Validity, invalidity and<br />
reliability. In E.F. Borgotta and G.W. Bohrnstedt (Eds.).<br />
Sociological methodology. San Francisco, CA, USA: Jossev Bass.<br />
Pp. 104-127<br />
Holsinger, D. 1972. The elementary school as an early socializer<br />
of modern values. Stanford, CA, USA: Stanford<br />
University, PhD. dissertation. Pp. 102-184.<br />
Inkeles, A.; Smith, DH. 1974. Becoming modern. Cambridge,<br />
MA, USA: Harvard University Press.
96<br />
Inkeles, A.; Smith, D.H. 1966. A comparative socio-.psychological<br />
measure of individual modernity. Readings on Modern Sociology.<br />
Kahi, J.A. 1968. The measurement of modernism A study ot<br />
values in Brazil and Mexico. Austin , TX, USA:<br />
University of Texas Press.<br />
Keeves, J.P. 1971. The home, <strong>the</strong> school and educational achievement.<br />
Canberra, Australia: Australian National University, unpublished Ph.D.<br />
Keevea, J.P. 1972. Educational environment and student<br />
achievement. Stockholm, Sweden; Almquist and Vikeell<br />
<strong>International</strong>. P. i34.<br />
Keeves, J.P.; Lewis, R. 1979. Classroom, pupils and classroom<br />
outcomes. Hawthorn, Australia: Australian Council for Educational Research.<br />
Linn, R.L.; Werts, C.E. 1969. Assumptions in making causal<br />
differences from part correlations, partial regression<br />
coefficients. L'rinceton, NJ, USA: Educational Testing Service,<br />
Research .Bulletion RB-69-6.<br />
Majoribanks, K. 1974. Ano<strong>the</strong>r view of <strong>the</strong> relation of<br />
environment to mental abilities. Journal of Educational<br />
Psychology, 66: 460-3.<br />
Majoribanks, K. 1977e Affective and environmental correlates<br />
of cognitive perfoxmance. Journal of Educational Research,<br />
io: 3...8<br />
Majoribanks, K. 1978a. Birth orders, age spacing between<br />
siblings and cognitive performance. Psychological Reports,<br />
42: 11-123.<br />
Majoribanks, K. i978b. Family and school environmental<br />
correlates of school related affective characteristics.<br />
Journal of Social Psychology, in Press.<br />
Murray, M.A0 1938. Explorations in personality. New York, NY, USA:<br />
Oxford University Press. Pp. 16-124.<br />
Noyjk, M.R. and Lewis C. 1967. Coefficient alpha and <strong>the</strong><br />
reliability of composite measurements. Psychometrica, 32:<br />
1-13.
I<br />
97<br />
Nunally, J.C. 1967. Psychometric <strong>the</strong>ory. New York, NY, USA:<br />
McGraw-Hill, p. 186-257.<br />
Plowden, B, 1967. Children and <strong>the</strong>ir primary schools. Report<br />
of <strong>the</strong> Central Advisory Council for Education<br />
London, UK: HMSO.<br />
Rosier, N.J. 1978. Early school leavers in Australia. Stockholm,<br />
Sweden: Almquist and Wiksell <strong>International</strong>.<br />
Vernon, D.E. 1965. Ability factors and environmental influence.<br />
American Psychologist, 20: 723-733.<br />
William, T.H. 1976. Abilities and environment: Schooling<br />
and achievement in American Society. New York, NY, USA:<br />
Academic Press. Pp. 60-101.<br />
William, T,H. 1974. Competence dimensions of family<br />
environments. Paper presented at American Educational<br />
Research Association Annual Meeting, Chicago.<br />
Wolf, R. 1964. The identification and measurement of<br />
environmental press variables related to intel]igence.<br />
Chicago, IL, USA University of Chicago,Ph.D.<br />
dissertation.<br />
Wright, S. 1964. The methods of path coefficient. Annals<br />
of Ma<strong>the</strong>matical Statistics, 5 161-215.