Andrews University Digital Commons @ Andrews University
Dissertations Graduate Research
1984
Selected Educational Indicators for Monitoring the Primary-School System, Jamaica
Edith G. Marshalleck Andrews University
Follow this and additional works at: https://digitalcommons.andrews.edu/dissertations
Part of the Educational Assessment, Evaluation, and Research Commons
Recommended Citation Marshalleck, Edith G., "Selected Educational Indicators for Monitoring the Primary-School System, Jamaica" (1984). Dissertations. 559. https://digitalcommons.andrews.edu/dissertations/559
This Dissertation is brought to you for free and open access by the Graduate Research at Digital Commons @ Andrews University. It has been accepted for inclusion in Dissertations by an authorized administrator of Digital Commons @ Andrews University. For more information, please contact [email protected].
Thank you for your interest in the
Andrews University Digital Library
of Dissertations and Theses.
Please honor the copyright of this document by not duplicating or distributing additional copies in any form without the author’s express written permission. Thanks for your cooperation.
INFORMATION TO USERS
This reproduction was made from a copy of a document sent to us for microfilming. While the most advanced technology has been used to photograph and reproduce this document, the quality of the reproduction is heavily dependent upon the quality of the material submitted.
The following explanation of techniques is provided to help clarify markings or notations which may appear on this reproduction.
1.The sign or "target” for pages apparently lacking from the document photographed is "Missing Page!si”. If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting through an image and duplicating adjacent pages to assure com plete continuity.
2. When an image on the film is obliterated with a round black mark, it is an indication of either blurred copy because of movement during exposure, duplicate copy, or copyrighted materials that should not have been filmed. For blurred pages, a good image o f the page can be found in the adjacent frame. If copyrighted materials were deleted, a target note will appear listing the pages in the adjacent frame.
3. When a map. drawing or chart, etc.. is part of the material being photographed, a definite method of "sectioning” the material has been followed. It is customary to begin filming at the upper left hand comer of a large sheet and to continue from left to right in equal sections with small overlaps. If necessary, sectioning is continued again-beginning below the first row and continuing on until complete.
4. For illustrations that cannot be satisfactorily reproduced by xerographic means, photographic prints can be purchased at additional cost and inserted into your xerographic copy. These prints are available upon request from the Dissertations Customer Services Department.
5. Some pages in any document may have indistinct print. In all cases the best available copy has been filmed.
University Micrdnlrns International 300 N. ZeeO Hoad Ann Artxjr. Ml 48106
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8515553
Marshaileck, Edith Geraldine
SELECTED EDUCATIONAL INDICATORS FOR MONITORING THE PRIMARY- SCHOOL SYSTEM, JAMAICA
Andrews University Ed.D. 1984
University Microfilms International 300 N. Zeeb Road, Ann Arbor. Ml 48106
Copyright 1984 by Marshaileck, Edith Geraldine All Rights Reserved
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PLEASE NOTE:
In all cases this material has been filmed in the best possible way from the available copy. Problems encountered with this document have been identified here with a check mark v
1. Glossy photographs or pages ______
2. Colored illustrations, paper or print ______
3. Photographs with dark background ______
4. Illustrations are poor copy ______
5. Pages with black marks, not original copy ______
6. Print shows through as there is text on both sides of page ______
7. Indistinct, broken or small print on several pages S
8. Print exceeds margin requirements ______
9. Tightly bound copy with print lost in spine ______
10. Computer printout pages with indistinct print _____
11. Page(s) ______lacking when material received, and not available from school or author.
12. Page(s) ______seem to be missing in numbering only as text follows.
13. Two pages numbered ______. Text follows.
14. Curling and wrinkled pages ______
15. Other ______
University Microfilms International
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Andrews University
School of Education
SELECTED EDUCATIONAL INDICATORS OR MONITORING
THE PRIMARY-SCHOOL SYSTEM, JAMAICA
A Dissertation
Presented in Partial Fulfillment
of the Requirements for the Degree
Doctor of Education
by
Edith Geraldine Marshaileck
July 1984
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SELECTED EDUCATIONAL INDICATORS FOR MONITORING
THE PRIMARY-SCHOOL SYSTEM, JAMAICA
A dissertation presented in partial fulfillment of the requirements for the degree Doctor of Education
by
Edith G. Marshaileck
APPROVAL BY THE COMMITTEE: //
Chairman: Robert J. Cruise of Education
Committ ^ F tjtc n e r 3/, n s - j Committee Me Bernard M. Lall Date Approved/'
Committee flllejnb^-: Mapon) J. Merchant
Guest Examiner fpbm School of Education: Wilfred W. Liske
External Examiner :YjCameron Lam be
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. © Copyright by Edith Geraldine Marshaileck, 1984 All Rights Reserved
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT
SELECTED EDUCATIONAL INDICATORS FOR MONITORING
THE PRIMARY-SCHOOL SYSTEM, JAMAICA
by
Edith G. Marshaileck
Chairman: Robert J. Cruise, Ph.D.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT OF GRADUATE STUDENT RESEARCH
Dissertation
Andrews University
School of Education
Title: SELECTED EDUCATIONAL INDICATORS FOR MONITORING THE PRIMARY-SCHOOL SYSTEM, JAMAICA
Name of researcher: Edith C. Marshaileck
Name and degree of faculty adviser: Robert J. Cruise, Ph.D.
Date completed: July 1984
Problem
Since the 1960s there has been increasing interest in the develop
ment of social indicators, educational indicators being an important subset
of these. Studies in the U.S. and Europe have sought to (1) establish
a conceptual framework for educational indicators, and (2) identify indi
cators useful for economic planning and monitoring the state of education.
The need for these indicators in developing countries has also been
identified.
The researcher sought by (1) a correlational/predictive study
and (2) an exploratory opinionnaire to validate a list of indicators for
monitoring primary education in Jamaica and guiding financial choices.
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2
Method
Utilizing scores from three batteries of the 1980 Common Entrance
Examination (CEE) to high schools and a profile of pupil personal and
school variables, canonical and discriminant analyses were applied to
data from a sample of 802 pupils drawn from a stratified sample of schools,
in three parishes of Jamaica. The strata were urban, rural, and most
and least successful in securing high-school places based on CEE per
formance. Opinionnaire responses were secured from 177 primary-school
teachers on a list of selected indicators.
Results
The canonical analyses identified several inputs as correlating
significantly with pupil achievement including: (1) percentage of trained
teachers over a three-year period; (2) sessions attended 1978/79; (3)
parent's socioeconomic status when introduced into the analyses; (4)
principal's administrative experience (negative relationship); (5) higher
grade at which CEE was taken; and (6-7) grade teacher's experience and
qualifications. The four highest rated indicators by the teachers were
overcrowding, trained teacher/pupil ratio, efficient principal and index
of malnutrition.
Socioeconomic status and higher grade levels in urban areas and
higher percentages of trained teachers in rural areas were chief among
the distinguishing variables identified by discriminant analyses.
Conclusions
The researcher proposed a number of indicators set in a systems/
economics/time perspective. These included: (1) percentage of male/
female pupils in the age cohort attending 90 percent of sessions shown
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3
teachers; (3) standardized test results; (4) primary-school completion
expectancy rate; and (5) enrollment projections. Other indicators were
discussed, e.g., overcrowding and teacher/pupil ratios. Recommenda
tions were made.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS
ILLUSTRATIONS ...... vii LIST OF TA B LE S ...... v iii ACKNOWLEDGMENTS ...... xi
C h ap ter I. INTRODUCTION ...... 1
Social Indicators ...... 1 Educational Indicators ...... 3 Educational Indicators—Further N eeds ...... 4 Statement of the P ro b lem ...... 6 Purposes of the S tu d y ...... 6 Significance of the S tu d y ...... 7 H y p o th e s e s ...... 8 List of Indicators ...... 9 Definition of Term s ...... iG Eleven Plus (11+) Exam ination ...... 10 In d ic a t o r s ...... 10 Primary-School System ...... 12 Secondary-School System ...... 12 S y s te m s ...... 13 Triennial Examinations— Primary and All-Age Schools, Jamaica ...... 13 Delim itations ...... 14 Assumptions ...... 14 Organization of the S tu d y ...... 15
II. REVIEW OF RELATED LITERATURE ...... 16
Indicators—Other than Educational ...... 16 B a c k g r o u n d ...... 16 Economic In d icato rs ...... 17 Demographic Indicators ...... 19 Social In d icato rs ...... 20 Educational Indicators ...... 22 Studies Aimed at Identifying Indicators: Input/Output Educational Studies ...... 25 Outcome In d icato rs ...... 31 Criteria for Selection of In d icato rs ...... 31 Economic Considerations for Educational In d ic a to r s ...... 34 Educational Indicators: A Comprehensive W o r k ...... 40
iii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Literature Related to the Variables under S t u d y ...... 49 Sex, Age, Achievement ...... 49 Numbers of Years in S chool ...... 52 T e a c h e r /Schools Differential Effects ...... 54 Educational Televisio n ...... 56 School Libraries ...... 58 Socioeconomic Status: Classification I n d e x ...... 58 Literature on Jamaican Education Relevant to the S tu d y ...... 59 The Common Entrance Exam ination ...... 59 Correlates of Achievement in a Common Entrance-Level Examination ...... 64 Smaller Studies ...... 71 Towards Indicators in the Jamaican Educational S y s te m ...... 73 Quantitative Inform ation ...... 73 Educational Television: Jam aica ...... 75 School Libraries: Jamaica ...... 76 S u m m a r y ...... 77
III. METHODOLOGY ...... 80
O v e r v ie w ...... 80 The Correlational/Predictive S tu d y ...... 80 Selection of Sample S ch o o ls ...... 81 Description of Sample S chools ...... 83 Sampling of P u p ils ...... 88 Data Cathering— Correlational Study ...... 91 Obtaining the Quantitative Indices: C rite rio n V a ria b le s 1 - 4 ...... 92 Obtaining the Quantitative Indices: P re d ic to r V a ria b le s 1 - 1 6 ...... 92 Qualitative D ata ...... 94 Categorization of the Predictor V a r i a b l e s ...... 95 Rationale for Selection of the V a r i a b l e s ...... 95 Environmental Factors to Be Considered ...... 96 Strength, Relevance, and Degree of Differen tiation of Certain Variables of the S tu d y ...... 96 The Exploratory S tu d y ...... 97 Hypotheses and Statistical Analyses U sed ...... 98 Hypotheses 1(a)-1 ( c ) ...... 100 Hypotheses 2(a) — 2 (d ) ...... 100 Hypothesis 3 ...... 101 Hypothesis 4 ...... 102 Hypothesis 5 ...... 102 Hypothesis 6 ...... 103 Hypothesis 7 ...... 103 Hypothesis 8 ...... 103
iv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Analytical Methodology ...... 103 Pearson Product-Moment C o efficien t ...... 104 Analysis of V arian ce ...... 104 Canonical C orrelation ...... 105 Discriminant A nalysis ...... 106 Analysis of Teacher Responses by the Categorical Scalings M o d e l ...... 107 Data A nalysis ...... 108 The Family of Indicators ...... 108
IV. FINDINGS OF THE STUDY ...... 109
Results—Correlational/Predictive S t u d y ...... 109 Hypotheses 1 (a )-1 (c ) ...... 110 Hypotheses 2(a )-2 (d ) ...... 111 Hypothesis 3 ...... 115 Hypothesis 4 ...... 122 Hypothesis 5 ...... 124 Hypothesis 6 ...... 126 Hypothesis 7 ...... 127 Hypothesis 8 ...... 129 The Exploratory Study ...... 131 Results of the Opinionnaire ...... 133 Additional Indicators Provided by Teachers ...... 137 Factors Contributing to CEE Success as S u p p lie d by P r in c ip a ls ...... 138 Summary of the Findings ...... 138
V. SUMMARY. CONCLUSIONS, AND RECOMMENDATIONS ...... 142
S u m m a r y ...... 142 B a c k g r o u n d ...... 142 Methodology— Correlational Predictive A nalyses ...... 145 Survey Methodology ...... 147 F i n d i n g s ...... 147 Conclusions ...... 149 Im plications ...... 156 C o m m en ts ...... 157 Family of Indicators ...... 158 Indicator 1— A Context/Leading In d ic a to r ...... 159 Indicator 2— A Process/Contributing In d ic a to r ...... 160 Indicator 3— A Output/Lagging In d ic a to r ...... 161 Other Indicators ...... 162 Recommendations ...... 164
v
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX A—Sample Pages from 1984 Economic In d ic a to rs ...... 169 APPENDIX B—Johnstone (1981) Selected F o r m u la s ...... 171 APPENDIX C—McGuire-White Social Class I n d e x ...... 180 APPENDIX D— Performance of All Schools in C E E ,T h re e P arishes 1979 183 APPENDIX E— Performance of All Schools in C E E .T h re e Parishes 1980 189 A P P E N D IX F— CEE A w ard s 1979 and 1980 as Percentage of Pupils Enrolled in Grades 4-6—Ten Best Performing Urban and Rural Schools ...... 195 APPENDIX C— Details of School Sample Drawn, Urban (Most and Fewest A w a r d s ) ...... 196 APPENDIX H— Details of School Sample Drawn, Rural (Most and Fewest A w a r d s ) ...... 1S8 APPENDIX I—Common Entrance Score Card ...... 200 APPENDIX J—Socioeconomic Status— Categories 1-3 201 APPENDIX K—Examination Class List ...... 204 APPENDIX L—Sample of Questionnaire with Accompanying L e tte r ...... 205 APPENDIX M—Additional Indicators as Suggested by Teachers ...... 210 APPENDIX N— Principals' Observations on Factors Influencing Performance in the Common Entrance Examination ...... 213
BIBLIOGRAPHY ...... 216
vi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ILLUSTRATIONS
Figure 1. Educational Outputs and Examples ...... 35 2. A Framework for the Development of Indicators of Education Systems ...... 43 3. Schema for Indicators Developed by W rite r ...... 165 4. Lorenz Curves Measuring Inequality of Distri bution of Enrollments Across Grades at the First-Level— El Salvador and Nicaragua, 1965 ...... 172
vii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES
1. Categories of Educational Indicators ...... 24 2. Major Studies of Teacher Quality and Class S iz e ...... 26 3. Major Studies of Teacher Quality and Class Size— Frequency Count of Input Variables' E ffe c ts ...... 29 4. Selected Outcome Indicators and Possible Approaches for Measurement of these Outcome Indicators ...... 32 5. Results of Educational Production Function Studies for Poor C o u n tries ...... 41 6. Summary of Education Variables Used in Selected Comparative Education Studies ...... 45 7. Summary of Core Educational Input Indicators ...... 46 8. Summary of Core Educational Process Indicators ...... 47 9. Summary of Core Educational Output Indicators ...... 50 10. Core Indicators System— Societal Congruence ...... 51 11. Mean Quotients in Moray House Battery (1959 - Total Sample Jamaica) ...... 51 12. Mean Quotients—Partial Sample—Moray House Battery 1959 - Jamaica (Primary School Entrants O n ly ) ...... 52 13. Social Class and Test Performance—Moray House Tests (Jamaica) 1959 ...... 63 14. The Significance of Differences Between the Means Obtained by Different Social Groups ...... 63 15. Regression Weights of Promising V a ria b le s ...... 68 16. ETV Utilization Survey— 1978/79 R esults ...... 76 17. Schools with Bookstocks, Summer 1979 ...... 77 18. Number of Schools in the Parishes of Kingston, St. Andrew, and St. Thomas Classified as Urban (City/Town) or Rural/Remote by Ministry of Education (1976) ...... 82 19. Ten Best Performing Urban Schools (Grades 1-6) ...... 84 20. Ten Best Performing Rural Schools (Grades 1 -6 ) ...... 86 21. Urban Schools with Fewest Awards (Grades 1 -6 ) ...... 87 22. Rural Schools with No Awards (Grades 1 -6 ) ...... 88 23. Distribution of the Four Types of Schools in the Stratified S am ple ...... 89 24. 1980 Common Entrance Exam ination ...... 90 25. Sixteen Indicators Presented for Rating to the Sample of T each ers ...... 99 26. Correlational Analysis for Forty Schools—Obtained Correlations ...... m 27. Correlational Analysis for Forty Schools—Means and Standard D eviations ...... 111
viii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28. Comparison of Quotients in Manley's Study— 1959 Data; Present Study— 1980 D ata ...... 112 29. Means of Ten Best Performing Urban Schools— Total B a tte ry of 1980 CEE S c o r e s ...... 113 30. Analysis of Variance on School Means of Total Battery of CEE Scores—Ten Best Performing Urban S chools ...... 113 31. Means of Ten Best Performing Urban Schools— Mathematics Test of 1980 CEE Scores ...... 114 32. Analysis of Variance on School Means of Mathe matics Test— Ten Best Performing Urban S chools ...... 114 33. Means of Ten Best Performing Urban Schools— English T e s t of 1980 CEE S c o r e s ...... 116 34. Analysis of Variance on School Means of English Test—Ten Best Performing Urban Schools ...... 116 35. Means of Ten Best Performing Urban Schools— Intelligence Test of 1980 CEE S cores ...... 117 36. Analysis of Variance on School Means of Intelli gence Test— Ten Best Performing Urban S c h o o ls...... 117 37. Matrix of Differences of Means (Rounded)—Ten Best Performing Urban Schools— Intelligence T e s t ...... 118 38. Canonical Correlation and Standardized Coefficients for Each Canonical Function of Intelligence, Mathematics, and English Test Scores with Thirteen Personal and School Input Variables (N = 8 0 2 } ...... 120 39. Canonical Correlation and Standardized Coefficients for Each Canonical Function of Intelligence, Mathematics and English Test Scores with Twelve Personal and School Input Variables (N = 8 0 2 ) ...... 123 40. Discriminant Functions—Upper 33 Percent Compared with Lower 33 Percent of Urban Students— Twelve Variables ...... 125 41. Discriminant Functions— Upper 33 Percent Compared with Lower 33 Percent of Urban Students— Thirteen Variables ...... 128 42. Discriminant Functions— Upper 25 Percent Compared with Lower 25 Percent of Rural Students-- Twelve Variaoles ...... 130 43. Discriminant Functions— Upper 25 Percent Compared with Low er 25 Percent of. R u ra l S tu d en ts— Thirteen Variables ...... 132 44. Rank Ordering of Indicators Based on Teacher R a tin g s ...... 134 45. O rd erin g o f S ixteen In d ic a to r s ...... 136 46. Additional Indicators Provided by T e a c h e rs ...... 137 47. Ranking of Findings for Six Statistical Tests and Teacher Survey Opinionnaire ...... 139 48. Summary Ranking of Personal and School Variables (Canonical Correlations) ...... 150
ix
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 49. Summary Ranking of Personal and School Variables— Urban (Discriminant Analysis) ...... 151 50. Summary Ranking of Personal and School Variables— Rural (Discriminant Analysis) ...... 152 51. Summary Rankings of Teachers' Opinions on Sixteen Variables ...... 153 52. Percentage Distribution of First-Level Students by Grade—El Salvador and Nicaragua, 1965 172 53. Farrell's Scalogram of Educational Structural Differentiation: 49 Nations, I960 177 54. McGuire-White Social Class Index— Occupations ...... 182 55. CEE A w ard s 1979 and 1980 as P ercentage o f Pupils Enrolled in Grades 4-6—Ten Best Performing Urban Schools ...... 195 56. CEE Awards 1979 and 1980 as Percentage of Pupils Enrolled in Grades 4-6—Ten Best Performing Rural Schools ...... 195 57. Details of School Sample Drawn, U rb an (Most Awards) . 196 58. Details of School Sample Drawn, Urban (Fewest Awards) 197 59. Details of School Sample Drawn, Rural (Most Awards) • 198 60. Details of School Sample Drawn, Rural (Fewest Awards) 199
x
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGMENTS
The prolonged and intense effort involved in the production of this
dissertation means that more than usual thanks are due to a number of
persons who were instrumental in assisting me in bringing the project to
closure. My most grateful thanks are due (1) to the chairman of my committee.
Dr. Robert J. Cruise, whose optimism, inspiration, faith, and advice kept
me going over the long haul; (2) to the other members of my committee:
Drs. Futcher, Lall, and Merchant, and particularly Dr. Futcher who
assisted greatly in the statistical areas; (3) to Dr. Neville Ying, my local
(Jamaican) adviser; (4) to Dean G. H. Akers who so kindly made allowances
for my peculiar circumstances; (5) to Dr. Gertrude Jordan who unstintingly
provided material and moral support in numerous ways including long
distance computer/library work and editing; (6) to the staff of the Andrews
University computer center and to Mr. Bill Cash who interestedly ran the
analyses; (7) to Ken Harris-Stubbs, Devon Galbraith, and Mrs. Bruce Jones
who assisted me with data gathering; (8) to Eleanor Pommells-Clark and
Suzy Wilk who typed the first draft and final paper, respectively; (9) to the
school principals and teachers who cooperated in supplying the data without
which the study would not have been possible; (10) to the Ministry of
Education and the Ministry of Finance and Planning, Jamaica, for a grant
towards the project and leave-of-absence to complete it, respectively; and
(11) to Dr. Rudolph Klimes who launched me on the doctoral program.
Special thanks are also due to my superiors and colleagues in the
xi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ministry of Finance and Planning for their help and support.
I am indebted to my brother; Dr. L. M. Mullings, and sister, Mrs.
Edna Crarey, my friend, Cem Fitz-Henley, and others who encouraged and
generously assisted me financially.
My husband, Alton, and children, Sheryl and Hans, and my father,
patiently shared in the sacrifice of time which this study necessitated and
supported me fully to the end. May they share my joy.
Friends arose at the last minute to help and to them I am also very
g r a te fu l.
Finally, above all, I am deeply grateful for the Providential guidance
and assistance which have at all times attended me.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C H A PTER I
INTRODUCTION
Social Indicators
Introduced in the mid 1960s, the term "social indicators" has
excited the interest of professionals and lay people who have
interested themselves in studying the evolution of the social system.
Labelled also "social accounts," "social report," "social book-keeping,"
"social intelligence," and "monitors of social change," these indicators
were intended to identify and quantify those societal variables which
significantly affect the quality of life.
Proponents of social indicators did not overlook the already
well-established economic and demographic indicators. In the
published discussions on the subject. Land (1971, p. 323) stated:
I propose that the term social indicators refer to social statistics that (1) are components in a social system model (including sociopsychological, economic, demographic, and ecological) or of some particular segment or process thereof, (2) can be collected and analyzed at various times and accumu lated into a time-series, and (3) can be aggregated or dis aggregated to levels appropriate to the specifications of the model. . . . The important point is that the criterion for classifying a social statistic as a social indicator is its informa tive value which derives from its empirically verified nexus in a conceptualization of a social process.
Public debate over the types and potential uses of social indi
cators continued. By 1972, Sheldon and Land ( 1972) concluded that
the social indicators debate had
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2
. . . simmered down to more or less general recognition of three different types of indicators designed for different uses: (1) Problem-oriented or direct policy-oriented indicators which are intended for direct use in policy and program decisions. (2) Descriptive indicators intended primarily to describe the state of society and the changes taking place within it. (3) Analytic indicators that serve as components of explicit conceptual and causal models of the social systems or some particu lar segment thereof, (p. 139)
In addition, they offered as a point of departure the
following list of social indicators which located education among the sub
systems contributing to the social and economic welfare of society (p. 146).
A. Socioeconomic Welfare
1. Population (composition, growth, and distribution)
2. Labor force and employment
3. I ncome
4. Knowledge and technology
5. Education
6. Health
7. Leisure
8. Public safety and legal justice
9. Housing
10. Transportation
11. Physical environment
12. Social mobility and stratification
B. Social Participation and Alienation
1. Family
2. Religion
3. Politics
4. Voluntary associations
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3
5. Alienation
C. Use of Time
D. Consumption Behavior
E. Aspiration, Satisfaction, Acceptance, Morale, etc.
Educational Indicators
Given (1) the ferment in the social sciences leading to field work
on social indicators, (2) the fact that education is an important sub
system in the total societal system, and (3) the summons to accountabil
ity within the field of education itself, it is not surprising that a call
for educational indicators was sounded. Relatively early Ferriss (1969),
under the auspices of the Russell Sage Foundation, published Indicators
of Trends in American Education, a book which brought together time-
series data indicating the changing characteristics of education—
organizationally, and as a characteristic of the population (Sheldon,
1975, p. 7). Since then there have appeared other reports on
the educational attainments of young people, e.g., in the National
Assessment of Educational Progress and in The Condition of Education
(1975) published by the National Center for Educational Statistics, U.S.
Department of Health. Education, and Welfare (Sheldon, 1975, p. 8).
Activity in the development of indicators has not been limited
to work emanating from Washington, D.C., i.e., the U.S. Department of
Health, Education, and Welfare. States have constantly monitored input
measures and more recently have become involved in the systematic
measurements of input /output relationships and inventories of various
sorts. Cost benefit approaches have not been neglected either. In many
states there is legislation. Accountability activities in a range of states
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4
and territories are annotated by Hawthorne (1974). Publications cited
include: (1) Pennsylvania, Department of Public Instruction—publica
tions on Educational Quality Assessment (1968 onwards); (2) Florida,
Department of Education— Florida Statewide Assessment Programs (1971
onwards); (3) New York, State Education Department— Performance
Indicators in Education (1968 onwards); and (4) Oregon—'by Clemmer,
Fairbanks, Hall, Impara,and Nelson (1974) — Indicators and Statewide
Assessment.
Work on a cross-section of indicators was also underway over
seas. Studies on cost effectiveness (Coombs 6 Hallak, 1972), returns
to education (Psacharpopoulos & Hinchliffe, 1973), and various
publications through the International Institute for Education Planning
in Paris have all set a framework for international and cross-
cultural studies. Nor were the developing countries omitted. Kiros,
Mushkin,and Billings (1975), in a joint effort between Georgetown
University and the University of Addis. Ababa, produced the study
Educational Outcome Measurements in Developing Countries.
Educational Indicators— Further Needs
There are overwhelming data sources in education, yet there
remain deficiencies in crucial areas in many countries. Among the major
needs listed are:
1. A new tabulation format for old items (B. Duncan, 1968).
2. More evidence on many vital aspects of the system such as
the amount of learning that takes place (Ferriss, 1969) .
3. Segmenting the data to provide indicators that are not only
descriptive but also problem or policy-oriented, perhaps even analytical.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5
These indicators might be concerned not only with the objective,
quantitative outcomes of education but also with the subjective, qualita
tive outcomes— the "well being" of education. They might be concerned
with inputs as well as outputs, costs, returns on investment, and social
benefits. The educational operation could perhaps be sectored; and
final and intermediate products identified. Learning-domain expectations
might be established and achievements measured. In addition, the
"knowledge by-product" might be taken into account as well as its
return effect on the educational system. Indicators in the social environ
ment perceived as affecting education might also be included.
The question is not whether or not educational indicators should
be identified, but rather just what are the critical indicators and what
aspects of education are to be monitored. As Cooler ( 1975, p. 26)
questions: How does one arrive at a useful and valid mix of kinds of
educational indicators? What balance of qualitative and quantitative
indicators must be achieved? Further, how can indicators be "popular
ized" while retaining an acceptable depth of analysis? W. Cohen (1975,
pp. 60-62), provocatively suggests that there might be an educational
indicator as useful as the health indicator— Infant Mortality Rate—
that would come out monthly, that may be disaggregated, that may
be used in setting goals, and that may be used in identifying
deficiencies as well as progress.
These are questions which stimulate thought in the search
for educational indicators. They provide a reference for the assess
ment of the usefulness of proposed indicators.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6
Statement of the Problem
As Kiros et al. (1975, p. 9) recognized, ministers of education
and finance,as well as other ministry officials in developing countries,
feel acutely the need for quantitative bases for evaluating how and
w 'ere they might best direct their educational systems to further educa
tional goals and to stimulate national development. Together they face
the expensive and urgent task of human resource development against
the background of limited financial resources and other constraints—
human and societal.
Educators and administrators concerned with education and with
accountability in education feel the need for at least two types of
indicators:
1. Indicators of systems outputs which can be analyzed to see
how well the system is doing within itself and against external criteria.
2. Indicators of inputs of proven positive relationships with
achievement. These could be revalidated from time to time. Crucial
inputs may be reassessed. New inputs or inputs specific to time and
place may be identified.
Jamaican officials realize that Jamaica is no exception to these
problems. Both types of indicators are needed, particularly at the level
of primary education where the first extensive external assessment of
performance takes place at the Eleven Plus (11+) examination.
Purposes of Study
The primary purpose of the study was to identify and validate a
set of educational indicators for the primary-school system (grades 1-5)
in Jamaica. Subsidiary purposes were as follows:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7
1. To secure from a selection of Jamaican primary teachers
ratings of a number of educational indicators submitted to them and of
others which they might suggest. (The list would include input, process,
and output indicators as well as systems accessibility and environmental
indicators. )
2. To examine the predictive validity of certain inputs of the
primary school system by observing their co-varying relationship with
certain measures of pupil performance. (Those which apparently contri
bute most to achievement would be identified as significant input
indicators. )
3. To relate the two sets of input indicators which receive the
highest rankings in both investigations.
Significance of the Study
Indicators provide information on a system's performance. When
these are monitored over a period of time, they may be used to assess
progress towards goals, or they may reveal trends (including mainte
nance of or fluctuations around a norm) and tendencies toward growth or
decline. In a system as important as the educational system, it is highly
desirable to have such a profile of indicators for monitoring performance.
The stability of the measures from which the indicators are derived
would have to be established. The indicators themselves may be changed
or modified as planning considerations demand. The indicators identified
in this study should provide a useful initial profile for the study of the
Jamaican primary-school system.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8
H ypotheses
The study employed the survey technique to isolate a set of
highly rated educational indicators perceived to be essential in
effective educational management. It also employed a correlational/
predictive model to investigate the relationship between thirteen vari
ables and three sets of achievement scores in order to posit the
predictive validity of certain inputs.
The following research statement and general hypotheses were
therefore put forward:
1. There is a significantly differentiated list of indicators, as
rated by educators and administrators, such that at least ten ranked
indicators are secured.
2. There is a significant canonical correlation between a linear
combination of predictor variables and a linear combination of criterion
v a ria b le s .
3. There is a linear combination of predictor variables which
significantly separates the following groups of 11+ candidates:
(a) The top 33 percent from the bottom 33 percent of those
sampled in urban schools.
(b) The top 25 percent from the bottom 25 percent of those
sampled in rural schools.
(c) Among the ten best performing urban schools, the pupils
of the school with the highest mean score from the pupils
of the school with the lowest mean score.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9
List of Indicators
The list of indicators utilized in the study included the following:
1. Systems Utilization and Accessibility
(a) Ratio of enrollment to school-age population
(b) Degree of overcrowding by school
(c) Percentage of overcrowded schools
2. Systems Inputs
(a) Teacher/pupil ratio —trained teachers
(b) Teacher/pupil ratio—overall
(c) Number of library books per pupil
(d) Number of hours of educational television viewed
3. System s Process
(a) Percentage of teacher turnover
(b) Percentage of pupils attending 90 percent of school sessions
(c) Efficient principal
4. Environmental
(a) Index of malnutrition
(b) Percentage of unstable families
5. Systems Outputs
(a) Grade-by-grade average achievement in English
(b) Grade-by-grade average achievement in mathematics
(c) Percentage of students achieving below 50 percent in English and in mathematics
(d) Percentage of eligible age group obtaining passing mark in the Common Entrance Examinations
The rationale for the selection of these indicators for study is
presented in chapter 3.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10
Definition of Terms
Eleven Plus (11+) Examination
The Eleven Plus (11+) Examination is a national examination for
entry into forty-six "academic" high schools. The examination is adminis
tered to children between the ages of ten years nine months and twelve
years nine months, and comprises tests in mental ability, mathe
matics, and language arts. In January 1980, 35, 697 pupils from primary
and all-age schools and 2, *109 from independent schools—a grand total of
38, 106 entrants— sat for the examination. Awards totalled 8, 853 comprising
twenty-nine scholarships, where all expenses are paid, and 8,824 free
places, where qualifying students receive free tuition and assistance with
other expenses based on the criterion of need. Of these free places
7,578 went to Primary and All-Age Schools (Jamaica, Ministry of
Education, 1980— Educational Statistics 1979-80).
In d icato rs
Indicators are measures of performance which function as
informers of a system's condition. These are usually presented as a set
of statistics whose variations signal change in the larger world from which
they are derived. In economics a "leading indicator" leads or precedes
major economic change, a "lagging indicator" reports on a major economic
change after the fact, a "co-incident indicator" coincides with economic
change (Warner 6 Dublin, 1965, p. 29). Demographic indicators, though
not publicized as regularly as economic indicators, are nevertheless
respected and monitored for the insights provided. The birth and death
rates and even more importantly the infant mortality and life expectancy
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 11
rates are statistics that are observed and commented on in countries
all over the world.
Since social indicators and, more specifically, educational
indicators are the focus of this study it is useful to present some
definitional perspectives. Definitions of social indicators present a
dichotomy of functions. From the document Toward a Social Report,
published by the U.S. Department of Health, Education, and Welfare
(1969), the following statement is taken:
A social indicator . . . may be defined to be a statistic of direct normative interest which facilitates concise, comprehensive, and balanced judgment about the condition of major aspects of a society. It is in all cases a direct measure of welfare and is subject to the interpretation that if it changes in the "right" direction while other things remain equal, things have gotten better, or people are "better off." Thus statistics on the number of doctors or policemen could not be social indicators whereas figures on health or crime rates could b e . (p . 97)
Carlisle (1972) defined a social indicator as the operational
definition or part of the operational definition of any one of the
concepts central to the generation of an information system descriptive
of the social system (p. 25).
Two indicator functions are identifiable from the foregoing:
1. A statistic of normative interest.
2. A descriptor providing evidence on progress toward a goal,
or objective, or a measure of societal welfare.
In so far as utilization is concerned, it appears that the definition
of social indicators given by Sawhill (1969) is much to the point. They
are described as quantitative measures of [social] conditions designed
to guide choices at several levels of decision making (p. 473). In regard
to educational indicators. Cooler's ( 1975) definition seems comprehensive:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12
A definition of educational indicators might be deduced from definitions of social indicators, as applied to education. More simply stated, educational indicators are statistics that enable interested publics to know the status of education at a particular moment in time with respect to some selected variables, to make comparisons in that status over time, and to project future status. Indicators are time-series statistics that permit a study of trends and changes in education (pp. 11-12).
Educational indicators are not, however, synonymous with
evaluation. As Mushkin and Billings (1975, p. 66) note: "Educational
indicators are not always interchangeable with criteria for evaluation
which often require: much greater specificity."
Primary^School System
The Primary-School System comprises grades 1-6 found in four
types of schools. The number of each type of school in 1980 is shown
in parentheses. The grand total was 792 schools.
1. Primary schools, grades 1-6, ages 6^-12 (229 total)
2. Primary schools with infant departments, K-6, ages 4-15 (54 to ta l)
3. All-age schools, grades 1-9, ages 6*-15 (465 total)
4. All-age schools with infant departments, K-9, ages 4-15 (44 total).
Six special schools for the handicapped, including one with both
normal and physically handicapped pupils, are included in the numbers
g iv e n .
Secondary-School System
The Secondary-School System comprises five types of schools (138
total). The number of each type of school in 1980 is shown in parentheses.
1. New Secondary (78 total)
2. Government-aided "academic" High Schools (46 total)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 13
3. Technical High Schools (7 total)
4. Comprehensive High Schools (5 total)
5. Vocational and Agricultural Schools (2 total) .
System s
"Systems are bounded regions in space-time, involving energy
interchange among their parts which are associated in functional relation
ships and with their environments" (Miller, 1955, p. 514). They have a
number of interacting parts classified as subsystems which are inter
related in function and structure but are focused to a primary goal.
Systems are therefore characterized by boundaries, subsystems,
environment relationships (whether open or closed), and tendencies to
entropy (degeneration), steady state, or negentropy. Feedback is
vital for the maintenance of the system's steady state (Immergart 5
Pilecki, 1973, chapter 2).
Triennial Examinations— Primary and All-Age Schools, Jamaica
The triennial examinations for Primary and All-Age schools are
so arranged that each school undergoes assessment once every three
years. The evaluation involves written examinations and review of other
aspects of school operations by a panel of examiners. The written
examinations are administered at the discretion of the panel and all grades,
or a selected number of them, may be involved. Special attention is
paid, however, to grades 3, 5, 7, and 9. The written tests are not
standardized but are designed to be of comparable difficulty at each level.
Instructional management and teacher/pupil interaction are also reviewed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. w
Delimitations
The variables surrounding and involved in a systems operation
are multitudinous. The study confined its investigation mainly to
the school system and selected variables therein. The researcher's
knowledge of the Jamaican educational system and experience with its
operations influenced the initial selection of variables. Since the
opinion of other educators was also sought, the final selection was not
based on a sole opinion.
While it would seem that a study of educational indicators
should be more comprehensive in scope and should include education
(1) at secondary and tertiary levels, (2) in both formal and informal
systems, and (3) in both public and private institutions at primary level,
available resources for this study were too limited to permit this. The
study was, therefore, delimited to the public primary system comprising
grades 1-6. No attempt was made to study the private primary-school
system of preparatory schools or, for the most part, the extra-school
variables which might have a causal relationship with achievement in the
Common Entrance Exam ination.
Assumptions
In undertaking this study, it was assumed that
1. The systems model represents— if even grossly—the
functional relationships of the primary education system.
2. The data utilized in the study are close approximations of
reality and reflect significant causal or co-varying phenomena.
3. The known but as yet unmeasured variable, pupil motiva
tion, is normally distributed in the population of Jamaican primary-school
pup ils.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15
4. There is no sine qua non variable external to the school
variab les .
Organization of the Study
Chapter 1 discusses the activities in the area of social indicators
which have preceded and contributed to calls for educational indicators.
It identifies the problems to be addressed, and presents the aims to be
pursued in this study. Important terms are defined. Delimitations as
well as major assumptions are outlined.
Chapter 2 presents a review of the related literature, including
literature on indicators drawn from the fields of economics and health.
Chapter 3 deals with the methodology for deriving the pre
dictive/correlational relationships. The data used, the methods of
obtaining these, and the statistical analysis utilized are also outlined.
The method of securing and analyzing teacher opinion is also described.
Chapter 4 presents the results of the correlational studies
and the indicators identified in the survey of teachers and
principals.
Chapter 5 posits the family of indicators, draws conclusions,
and makes recommendations.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C H A PTER II
REVIEW OF RELATED LITERATURE
Literature and research findings relevant to this study are
reviewed under the following headings:
1. Indicators—other than educational
2. Educational indicators
3. Literature related to the variables of the study
4. Literature on Jamaican education relevant to the study
Indicators— Other than Educational
B ackg ro u n d
Cross (1966a) points out that, in the long history of Western
civilization, progress is repeatedly associated with efforts to obtain new
kinds of information. The censuses of Jewish and Roman times are
examples of such demographic information.- During the Middle Ages the
Latin phrase ratio status— the predecessor of "statistics"—was used to
refer to the factual study of politics and government. The word
"statistic," coined in Germany, first appeared in English in 1770.
The U.S. Constitution, framed six years later, showed the
framers' consciousness of the importance of this factual information of
politics and government. Article 1, Section 2, provides for a decennial
census. In 1790, by order of Congress, the free, white male population
16
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 17
was broken down into two categories, "16 years and over" and "16 years
and under." The nation was getting ready for business. By 1880
proposals for further breakdowns of the population were made, but these
were not implemented until twenty years later.
Article II, Section 3, of the Constitution provides that the
President "shall from time to time give the Congress information on the
state of the Union." The State of the Union message; the Budget
message—started under the Budget and Accounting Act of 1921; the
Economic Report— reference the Employment Act of 1946; the Manpower
Report—under the Manpower Development and Training Act of 1962; as
well as the Statistical Abstract together form the backbone of the
informational system on the "State of the Union" (Cross, 1966a, pp .xi,xiv).
Economic Indicators
The Economic Report, produced by the Council of Economic
Advisers, has received special attention, both developmentally and in
terms of publicity. It places order on a vast output of information.
The monthly Economic Indicators prepared by the Council is a care
fully organized set of some forty-five tables with charts. It has become
the means by which people "in-the-know" keep their fingers on the
pulse of the economy (Gross, 1966a, p. xii). The forty-five areas of the
Report are subsumed under the following topics;
1. Total Output, Income, and Spending
2. Employment, Unemployment, and Wages
3. Production and Business Activity
4. Prices
5. Money, Credit, and Security Markets
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 18
6. Federal Finance
7. International Statistics
Siderman (1966, p. 93) shows that up to 1964 the economic indicators in
State of the Union messages focused on:
1. Total activity and income statistics
2. Markets and production
3. Income (distribution)
4. Productive resources
5. "Sound" currency
6. Employment rates
7. Surpluses (Treasury)
8. Debts
Press, radio, and television utilize information from the "leading"
economic indicators regularly, and the Consumer Price Index is of
interest to the majority of citizens.
The indicators themselves are derived from a number of
sources—government departments (including the Treasury), public and
private corporations, and firms. The data from each economic area are
summarized and presented in analytic and interpretive tables, together
with graphs showing trends over a given period of time. The
analytic and interpretive tables are specific to and intended to be
appropriate to each area. Two sample pages from the Economic Indi
cator, published in 1984,are provided in appendix A. They are
(1) The Table of Contents and (2) Changes in Consumer Prices
(table and chart).
These economic indices reflect the development of measurement
technology in areas such as the development of index numbers, adjust-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 19
ments for changing monetary values, and plotting of real as against
apparent growth rates. It should be noted that the indicators are
subject to error. Oskar Morgenstern (1963, pp. 242-252] has called
attention to inaccuracies in the economic time series that are in
greatest use in making judgments about the economy. The reasons
for these errors are not discussed in this paper.
Demographic Indicators
The census, decennial or otherwise, is a familiar feature of the
developed and developing societies. Population growth is of such
vital and universal concern that the United Nations has made it one
of its areas of interest.
A substantial body of literature exists on the subject of demo
graphy, and although there are no monthly indicators, the public is
regularly made aware of vital statistics. Crude birth and death rates,
infant-mortality rates, incidence of deaths-by-disease categories, and
I if e-expectancy rates are all familiar and publicized statistics.
The statistical procedures used in deriving some of these
indicators are fairly simple; for example, the fertility rate showing
births per 1,000 women in the age group 15-45 or the crude death
rate showing deaths per 1,000 of population. Far more sophisticated
and often controversial statistical methods are employed in the deve
lopment of life-expectancy tables and population-projection rates.
The interpretation of all of these indicators must be made against
the background of additional information. A simple example is a deceptively
low death rate due to a predominantly young population.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission 20
Social Indicators
The development and use of social indicators have not shown as
long and fruitful a trend as is the case of the demographic and econo
mic indicators. Lazarsfeld (1961) affirms that the idea that social
topics could be quantitatively analyzed arose in the seventeenth century.
Biderman (1966, p. 76) reported that over a century ago Herbert
Spencer advocated the use of statistics for societal analysis.
A movement toward the development of social indicators did not
eventuate in the United States until the 1960s. The development of the
space program and its likely impact on the lives of people provided
impetus for this movement.
The Organization for Economic Cooperation and Development
(OECD) produced a document. Indicators of Performance of Educational
Systems (1973), which observed that there are two opposite views
concerning the definition of social indicators. On the one hand is the
definition of measures of output or results, put forward in the U.S.
Department of Health, Education, and Welfare Document 1969— Toward a
Social Report. On the other hand, there is the in-depth social report
ing view: i.e., the assessment of the condition of society vis-a-vis
its aspirations, goals, or problems (p. 23). This dichotomy in
definition is pointed out in chapter 1.
A great deal has been written in an effort to (1) provide
definitions (Carlisle, 1972; Land, 1971; U.S. Department of Health,
Education, and Welfare, 1969), (2) set conceptual foundations (Bauer,
1966; Biderman, 1966; Cross, 1966a; Moore & Sheldon, 1965), and (3)
present reflective and sometimes critical evaluations (Bell, 1969;
O. Duncan, 1969; Land, 1971; Olson, 1969; Sheldon 6 Freeman, 1970).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 21
Methodological work continued, as evidenced by the study by Andrews
and Withey (1976) which sought to measure Americans' perceptions of
well-being in order to see how these could be used to yield a series
of "social indicators."
In 1973 the U.S. Office of Management and Budget prepared
Social Indicators 1973, a book of statistics on eight major social areas—
health, public safety, education, employment, income, housing, leisure
and recreation, and population. It looked forward to subsequent edi
tions. Two others appeared in 1976 and 1980. The 1980 edition
expands the list of social areas to include (1) transportation, (2) social
security and welfare, and (3) social participation. These indicators
were based on two main criteria: measurement of individual and
family well-being (rather than institutional well-being) and measure
ment of outputs rather than inputs.
Concerning the question of indicators, Bauer (1966, p. 26)
observed that the statistical series collected by the U.S. Government
(or by anyone, for that matter) reflect those areas of concern that
have occupied the minds of the American people—though with some lag
in time. These interests, he claims, are reflected unevenly since the
probability of a given statistical series being developed is also
affected by:
1. The articulateness and power of the group whose interest
is involved
2. The susceptibility of the phenomenon to being measured
3. The extent to which the phenomenon is socially visible
4. The preferences and skills of agency personnel who gather
the statistics.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 22
Relatedly, Biderman (1966, p. 97) noted that a consensus of
the entire society, or of any given social group within it, concerning
the significance of a certain phenomenon, does not mean that this will
be reflected by available indices. Other factors determining the
availability of indicators are (1) measurement technology, (2) social
observability, and (3) data-agency perspective. Fear of political
misuse should not be overlooked. Indicators may be used either as
vindicators or indictors (Biderman, 1966, p. 78).
Educational Indicators
In the preface to the report on the proceedings of the 1975
Educational Testing Service (ETS) invitational conference on educa
tional indicators, David Krathwohl declared that "educational indicators
is a topic whose time has come" (p. x i). He pointed out that, in
America, massive federal legislation often has great impact upon the
schools, but key statistics are seldom available to help legislators,
policy makers, and laymen understand in uncomplicated terms how
education is changing. The need for such measures has been rein
forced by a mandate from Congress: the 1974 Educational Amendments
to the Elementary and Secondary Education Act require the Assistant
Secretary for Education to submit an annual report to Congress on
the condition of education in the United States. A response was there
after made to this order in the publication of The Condition of Educa
tion (1975; Golladay, 1976 ed.) published by the U.S. Department of Health,
Education, and Welfare. The 1976 document is divided into two sections:
1. "An Overview of Education" dealing with students and
schools, outcomes of schooling, and concerns of elementary and
secondary education.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 23
2. "A Closer Look at Selected Topics" such as post-secondary
education participation, relationships between education and work,
education personnel, and finance. International comparisons of educa
tion in the United States and in other countries are also included.
Although not labeled educational indicators, it is substantially that, as
it presents a wealth of tables and graphs.
The ETS Conference (1975) also dealt with the broad
context and problems of defining social indicators, including educational
indicators. In defining educational indicators at the conference,
Dennis Gooler (1975) suggested a framework for a comprehensive system
of educational indicators. The system is categorized in table 1.
Considered overall, the proposed indicators could be categor
ized according to potential use—i.e., (1) those of practical interest
and (2) those of academic interest. Indicators of cognitive and
affective outcomes, indicators as to aspirations, access, and concrete
ratios are of practical interest. Indicators of impact and of use of
resources such as cost to benefit/effectiveness ratios might be consi
dered of interest primarily to the professionals.
Other topics dealt with at the 1975 conference included
"Measurement and Efficiency in Education" (Olson), "Quality of Life
as an Educational Outcome" (Withey), and "Measures of Educational
Outcomes in Developing Countries" (Mushkin S Billings).
Gooler (1975) raised three questions especially pertinent to
this investigation:
1. How can one identify cause-effect or at least correlational
relationships among indicators?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 24
2. How can one arrive at a useful and valid mix of kinds of
indicators?
3. What balance of qualitative and quantitative indicators
must be achieved (p. 26) ?
TA B LE 1
CATEGORIES OF EDUCATIONAL INDICATORS
Access
How many and what kinds of people participate in educational activities Retention rates in educational activities Catalog of existing/available educational activities or services
Aspirations
Description of needs and desires of various kinds of people Individual self-assessment of personal capabilities Description of institutional goals
Achievement
What people know, do, and feel What people have earned (degrees, diplomas, certificates) What is taught
Impact
Consequences of having schooling Impact of education on social /economic/cultural systems Consequences of not having schooling
Resources
Capital, personnel, and materials expenditures Quality of human resources Cost to benefit/effectiveness ratios Quality of educational climate Time
Source: Dennis D. Gooler, "The Development and Use of Educational Indicators." In Educational Indicators: Monitoring the State of Education. Princeton, N.J.: Educational Testing Service, 1975, p. 15 (ERIC Document Reproduction Service No. Ed 150 147).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 25
Studies Aimed at Identifying Indicators: Input/Output Educational Studies
In the United States, the landmark input/output study is the
Coleman Report (1966) which surveyed 645,000 pupils in more than
3,000 schools in grades 3, 6, 9, 12 in all regions of the United States.
The survey showed minor differences in school characteristics but
great differences in achievement across the different racial and ethnic
groups. The Report argued that when social background and attitudes
of students are held constant, per pupil expenditures, pupil/teacher
ratios, school facilities, and curricula show very little relation to
achievement (Jensen, 1974, p. 176).
Twenty-three studies of input/output performance before,
including, and since the Coleman Report (seven of which analyzed part of
the Coleman Report) are summarized in table 2. These are based on data
gathered from (1) selected grade levels in primary and secondary
U.S. schools, (2) kindergarten and elementary schools in Great
Britain (one study), and (3) U.S. college freshmen (one study).
Bargen and Walberg (1974), who supplied the summary, note that
nearly all studies show that per student expenditures for teachers, and
teacher experience or education are significantly associated with student
achievement as measured in most cases by student verbal or reading
achievement. These results were striking since verbal skills are
importantly "conditioned by . . . home environment and by heredity"
(p. 243). Average class size showed mixed results and should be
interpreted cautiously.
It is observed from the twenty-three studies listed that the
investigations centered around the following inputs:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE 2
MAJOR STUDIES OF TEACHER QUALITY AND CLASS SIZE
Investigator Crades Sample Control Variables Significant Correlates with Student Achievement
Armor (1972) 6 U.S. schools SES*, background Teacher salary, verbal ability
Averch and Klesllng 9 U.S. schools and SES* Teacher salary, class size ( 1970) individuals
Benson ( 1965) 5 California schools SES (district) Teacher salary
Bowles ( 1969) 12 U.S. Negro Indi Background Teacher verbal ability, education; viduals class size
Bowles and Levin 12 U .S. individuals Background Teacher verbal ability, experience ( 1966)
Burkhead, Fox, and 12 Chicago and Atlanta Median family Income Teacher salary, experience; class Holland (1967) schools (census), lower-grade size nonsignificant achievement
Cohn (1968) 12 Iowa districts Prior achievement Teacher salary; class size non significant
Coleman et al. 3,6, 9, 12 U.S. individuals Background Teacher verbal ability; teacher ( 1966) experience, education and class size nonsignificant, but these findings are disputed
Goodman ( 1 959) 7, 11 New York Stale SES Teacher expei lence, observer rating individuals of ability to relate subject matter to student ability and interest. Instruc tion expenditures per student
Guthrie, Klelndorfer, 6 Michigan SES Teacher experience, attitude toward Levin, and Stout (1971) individuals teaching, school, and other teachers, verbal ability
Hanushek ( 1970) 3 California SES*, prior achieve Teacher experience and education individuals ment nonsignificant
verbal ability nonsignificant tion major in college ability, academic rather than educa salary - negative; class size - mixed observer rating class of size teacher nonsignificant quality; significant; only significant teacher experience when all included control In model variables significant; class size High school teacher salary; class size Teacher experience, tenure, verbal in in school; class site attendance at universitycollege; teacher rather verbal than ability non Teacher experience and education non Expenditures per student student; class size Teacher salary; expenditures per Teacher experience, accreditation; education, lack of teacher turnover Expenditures per student Teacher experience, education-mixed; Teacher experience, satisfaction, and Teacher experience, attitude, and
SES SES SES SES*, background SES (by census tracts) Teacher experience, verbal abilitySES, prior achievement SES SES, student attitudes,grade aspirations, par SES, backgroundSES Teacher experience, education; SES, SES, background SES, SES, characteristics of home and community ents' attitudes
individuals individuals U .S. schools Individuals U .S. Individuals U.S. northern U .S. schools individuals Eastern city British districts individuals Boston districts districts districts Virginia 9, 9, 12 6 Eastern city 8 New York Statu 6 New York State 6,9, 12 6 Northern urban 6 10, 10, 12 Kinder- freshmen High garden, elementary School
Source: M. Bergen and H. J. Walberg, "School Performance." In Evaluating Educational Performance. Edited by Herbert Walberg. (Berkeley, Calif.: McCutchan Publishing Corporation, 1974), pp. 240-242. number of books In the home, appliances, and other factors thought to be related to student achievement. *SES *SES refers to status socioeconomic measured by parental or average community education and/or income. Background refers to (Central Advisory Rlblch 1968) ( Hanushek 1968) ( Klesllng (1970) Levin 1970) ( 6 Plowden — 1967) ( Raymond 1968) ( College Kalzman (1971) Klesllng (1969) Smith 1972) ( Thomas Thomas 1962) ( Council for Education) Mlchaelson Mlchaelson 1970) ( Melville ( 1956) Mollenkopf and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28
1. T each ers
their education
their experience
their accreditation
their verbal ability
their observed and rated quality
their attitudes to teaching, to school, to other teachers
their salary
their tenure
their satisfaction
their lack of turn over
their ability to relate subject matter to student ability and interest
2. Class Size
3. Expenditures Per Pupil
The results of these studies, in terms of the times that the
aforementioned inputs were found significant/not significant or had
mixed results, are summarized in table 3.
The following points taken from the Mushkin and Billings
(1975) presentation at the ETS Conference are relevant at this point:
1. The findings of the James Coleman study—that the chief
determinants of educational achievements of children are the educational
level and socio-economic status of their families— might not be applicable
in every type of case. Combinations of school inputs seem responsible
for differences in some cases. For example, Schiefelbein's (1974)
Chilean study report showed textbooks to be significant; Blaschke's (1975)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 29
TABLE 3
MAJOR STUDIES OF TEACHER QUALITY AND CLASS SIZE— FREQUENCY COUNT OF INPUT VARIABLES' EFFECTS
V a ria b le Times Times Found Times Found M ixed Studied Significant not R esu lts Sig. or Neg.
Teacher variables
Education 9 6 2 1 Experience 14 9 4 1 Accreditation 1 1 0 0 Verbal ability 9 7 2 0 Observed quality 1 1 0 0 Attitudes to teaching, to other teachers 1 1 0 0 S alary 8 7 1 Neg. 0 T enure 1 1 0 0 Satisfaction 1 1 0 0 Lack of turn-over 1 1 0 0 A b ility to relate subject matter to student ability & in te re s t 1 1 0 0 Class size 11 6 4 1 Expenditures 4 4 0 0
Michigan study showed teacher behavior to be significant; the study by
the International Association for the Evaluation of Educational Achieve
ment (IEA) Sweden (see Comber 5 Keeves, 1973; Platt, 1973; Purves,
1973; Thorndike, 1973) showed that there are differences in the extent
to which achievement in different subjects is influenced by family back
ground, family having less to do with achievement in science and mathe
matics than with achievement in vocabulary skills or literature.
2. Achievement measures have been found to vary with
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30
measures of self-concept and perception of locus of control (pp. 68-69).
Moats (1970) utilized an input-process-output model to
validate ten educational indicators (input and output indicators) taken
from the Pennsylvania Bureau of Educational Quality Assessment and
asked two questions: "Are the indicators statistically related to
quality education as defined and measured by some reliable method?"
and "How many and what are the factors which make up the domain
represented by the indicators?" After canonical analysis, the highest
weighted indicator variables emerged as (1) net total expenditure,
(2) percentage of graduating class receiving post high-school educa
tion, and (3) dollars spent on curriculum materials. Factor analysis
produced four factors accounting for 71 percent of the variance.
These were: (1) staff expenditure, (2) orientation to broad and
continuing education, (3) staff working conditions, and (4) pupil
dropout rates.
Kuhns (1972) investigated the predictive relationships among
certain specified manipulative inputs and selected outputs as measured
by the Pennsylvania Department of Education, Bureau of Educational
Quality Assessment. Multiple-linear regression was used and three
input-output models were compared. The three models utilized
manipulative and non-manipulative inputs selectively. A number of
interesting findings resulted; among them:
1. The manipulative inputs generally were adequate
predictors of school outputs.
2. A major determinant of student performance on school-
output measures was non-school (non-manipulative) inputs such as
family socioeconomic status.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 31
Outcome Indicators
Kiros, Mushkin, and Billings (1975) undertook "an initial
effort" to bring together information about educational outcomes,
including types of measures, the role and use of measurements, and
national policy consequences of measurements in developing countries.
Table 4 provides their potential list of outcome indicators and possible
approaches for measuring these. The important point was made that
"output" was an analytical construct— that through some rationale a
developmental output or set of outputs had to be settled upon. Review
of table 4 shows their emphases to be on affective and qualitative
long-term outputs. These represent important outcomes for study in
developing countries but are mainly outside the scope of this study.
Criteria for Selection of In d ic a to r^
Clemmer, Fairbanks, Hall, Impara, and Nelson (1974, p. 3),
in their work on indicators for statewide assessment in Oregon, made
several worthwhile points on the selection of indicators. To be of
greatest use in assessment, the indicators should be:
1. Related to agreed-upon goals
2. Derived from reliable and valid data
3. Derived from data that will continue to be collected so
that comparisons over time may be made
4. Derived from data for which the measurement techniques
have stability over time.
Their classification of indicators was as follows:
1. Input indicators—conditions or variables over which the
school has some c o n tro l.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 32
TABLE 4
SELECTED OUTCOME INDICATORS AND POSSIBLE APPROACHES FOR MEASUREMENT OF THESE OUTCOME INDICATORS
Outcome Indicators Approaches for Measurement
"Personal Development" Indicators
1. Achievem ent cognitive tests self-reports unobtrusive (non-reactive) measures
2. Attitude towards work, professionalism self-report (interview, question naire employer reports
3. In itia tiv e observation (structured) self-report (interview) employer reports
4. Leadership self-report (questionnaire) unobtrusive (non-reactive) measures (electio n to o ffic e , prom otion) employer reports non-cognitive tests
5. Motivation self-report (interview, question naire
6. Rationality self-reports (non-cognitive tests, in te rv ie w )
7. Risk-taking, "seculariza self-reports (non-cognitive tests, tion," self-reliance interviews, questionnaires)
8. Tolerance self-report (non-cognitive tests) observation reports of associates
"Social Outcome" Indicators
1. Social m obility . self-reports (interviews, questionnaires) . unobtrusive measures (back ground, income)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 33
TABLE 4—Continued
Outcome Indicators Approaches for Measurement
2. Social integration . self-reports (questionnaire on inter-tribal marriage, associa tes, language spoken at home) . unobtrusive measures (migra tion, urbanization data)
3. Social participation . self-reports (questionnaire on membership in organizations, civic and community participa tion) . unobtrusive measures (member ship, contributions)
4. Health-level improvement . self-report (questionnaire on health practices, attitudes) . unobtrusive measures (infant mortality, incidence of certain diseases)
"Political Outcome" Indicators
1. N atio n al in teg ratio n . self-reports (questionnaire on attitudes, political participa tio n ) . unobtrusive measures (migra tion, urbanization)
2. Democratization . self-reports (questionnaire on attitudes towards justice, merit and reward, voting practice) . Unobtrusive measures (voting p a tte rn )
Source: F. Kiros, S. Mushkin, and B. Billings, Educational Outcome Measurement in Developing Countries. Washington, D.C.: Georgetown University, T975, pp. 93, 94 (ERIC Document Reproduction S ervice N o . Ed 120 25 3 ).
2. Context indicators—conditions or variables over which
the school has little or no control; for example, ability of student or
socio-economic status of students.
3. Performance indicators—measurable or observable
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 34
behavior or variables used to determine program effectiveness or
efficiency (data may concern [a] student performance scores or
[b] a program variable such as instructional process or availability
of learning experience).
4. Societal indicators—measurable aspects of a social con
dition affected to some degree by education; for example, (a) arrests
for delinquency, (b) employment rates for recent graduates (p. 3).
Clemmer and his associates (1974) have noted that the
criteria for the acceptance of an indicator may change depending on
desired use. They also produced a summary classification of outputs
and examples (see figure 1). Their secondary indicators may be
classified as longer-term outputs and, in this regard, would be
similar to those put forward by Kiros, Mushkin,and Billings (1975).
Economic Considerations for Educational Indicators
Noah (1971, pp. 199-207) points out that, in the common
situation where there are limited resources available to satisfy the
wants or needs of an individual or organization, there arises the
problem of a choice between alternative means of pursuing given ends.
In the business world, the businessman is forced by competition to
behave in an economizing and innovative way, though entrepreneurs
seek more than just maximum profits. A high rate of innovation is
consistent with the achievement of many of his objectives.
In education as in industry, the efficiency problem can be
raised. Knowing that we want to produce a given output, what is the
most economical way of doing the job?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 35
Phase 1 . Primary Effects
P roduct In vestm en t Consumption
Q u a n tity Q uality 1 ncome Employmentj
Num ber o f A ttitu d e s Value Added, School drop s tu d e n ts . A ttrib u te s Earnings, outs. High School A ptitu d es A dded Unem ploy completion, Achievement e a rn in g s , m ent R ate, etc . (e.g., self e tc . e tc . esteem, creativity, I.Q . S A T scores)
Phase 2. Secondary Effects
In vestm en t Consumption j F eedback Feed b ack |
Economic Consumer information growth (e.g., Consumer efficiency years o f Medical care schooling, U 1>C of leisure time lifetim e Moral and citizenship earnings dif values, etc. ferentials)
Phase 3. Tertiary E ffects
Intergenerational Im pact
Educational Motivation of Children
Fig. i. Educational Outputs and Examples
Source: R. Clemmer, D. Fairbanks, M. Hall, J- Impara, C. Nelson, Indicators and Statewide Assessment, Salem, Oregon: Oregon State Department of Education, 1974, p. 14 (ERIC Document Reproduction Service No. Ed 091 869) .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 36
Most decisions in education, as in other areas of adminis
tration, are made in conditions of partial uncertainty and ignorance.
The major need is to organize information around the decision require
ments, to limit the area of uncertainty, and to define its general
characteristics (Noah, 1971, p. 207). As a minimum the following
actions are required by boards of education:
1. To identify clearly the objectives of the educational
enterprise being financed (not in the usual shibboleths of "education
for life" or "educating each child to the extent of his capabilities11)*
2. To explore the full costs of the decisions taken,
recognizing the commitments to future costs.
3. To identify and explore alternative ways of achieving
stated objectives.
4. To estimate the full costs of each of these alternatives.
5. To associate measures of output (benefit, yield) with
each of these alternatives so that some meaningful ratios of expected
output per unit of input may be calculated.
6. To deliberately and, as a matter of constant concern,
regularly develop new alternative ways of achieving given objectives which
may be more satisfactory than conventional ones. (p. 207)
A board of education which does this is undertaking program
budgeting and forging a powerful tool for improvement of its decision
making. And one could add that it is laying the foundation for
performance budgeting.
Kiros, Mushkin, and Billings (1975) commented at some length
on the economic outcomes of education, including macro- and micro-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 37
economic measures. In the introduction to their work it is noted that
education in the developing world is playing a major multipurpose
role— economic, social, and political— with diverse weights assigned to
these purposes in different nations. Later it is pointed out that
macroeconomic and microeconomic estimations of the contribution of
education at the various levels may be made, despite difficulties in
identifying and quantifying the variables involved. At macroeconomic
levels, estimations of the contribution of education to the total growth
in real national income may be made. At microeconomic levels, there
may be estimations of benefits through cost-benefit analyses, utilizing
internal rates of return, and net present discounted values for these
analyses. Costs, for example, would include the following:
1. Direct costs:
(a) Salaries of teachers and auxiliary staff
(b) Supplies provided by and for the school
(c) Rent for land usage
(d) Interest (debt service).
2. Indirect costs:
(a) Room and board if the student has to live away from home
(b) Additional clothing required for school
(c) Transportation to and from school
(d) Cost of personal supplies.
3. Opportunity costs:
(a) The value of the tax exemption enjoyed by non profit educational institutions
(b) The opportunity cost of the student's time (employ ment foregone) while in school, (pp. 121, 122)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 38
The benefits would include:
1. Private benefits:
(a) Increased earnings for the individual
(b) Consumption benefits for the individual
(c) Increased capacity of individuals to adjust to new circumstances, jobs, job opportunities.
2. Social benefits:
(a) Provision of manpower for economic growth
(b) Increased productivity for society
(c) Benefits to society of invention and innovation that are generated by increased levels of education
(d) Better citizenship (and hence lower crime costs)
(e) Intergenerational effects
(f) External effects on other individuals, family, employees, and employer
(g) Discovery and cultivation of talent by the educa tional system, (p. 124)
It was also pointed out that studies of social rates of return
to education performed for many developing countries showed that there
was a decline in the rate of return as the level of education increased.
Returning to the question of output measurement, Kiros et
al. ( 1975) point out that, when output data in educational programs and
projects are available, gaps in knowledge, skill, and attitudes in the
population can be identified. Deficiencies, especially those deemed
functional for the society, become the basis for issue definition and
analysis. Such analysis can inform policy officials about the size and
nature of the problem and set the stage for generation of policy and
project options. Determination of outputs of education, in a measurable
way, underlies improved management of resources.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 39
Alexander and Simmons (1975) attempted to identify schooling
inputs which might be recommended to policy makers as having a
significant impact on schooling. They proposed an educational produc
tion function (EPF) as an analytical tool for this purpose: despite
acknowledged limitations for economic analysis, it was felt that the
results could be used for analyses of the cost effectiveness of selected
educational inputs. The production function expresses the maximum
product from an input combination at the existing state of technical
knowledge. Its nature and underlying assumptions derive from the
theory of the firm which is extensively examined in microeconomic
literature. In order to maximize output subject to a budget constraint, .
it is necessary that the marginal product of the last dollar spent be
the same for all inputs. The prescription is to estimate the marginal
products of all inputs, determine their relative prices, and equate
the ratios of marginal product to price over all inputs. Given this,
output efficiency in the educational sector would be maximized (p. 1).
The Education Production Function is generally written
as (Alexander S Simmons, 1975, p. 2) :
Ait = 9 (Fi(t)' Si(t)' P i( t)' °i( t)' 'it' U) The i subscript refers to the i^ student; the t subscript in parenthesis (t) refers to an input that is cumulative to time t, w h ere
Ajt = a vector of educational outcomes for the i^ student at tim e t
F.f = a vector of individual and family background charac teristics cumulative to time t
S.,..= a vector of school inputs relevant to the i^ student cu m u la tiv e to t
P., .= a vector of peer or fellow student characteristics cumulative to t
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 40
0 .,., = a vector of other external influences (the community, for example) relevant to the L. student cumula tiv e to t
I.. = a vector of initial or innate endowments of the i.. student at t
U = an error term
The use of the EPF would depend on two factors. These are:
(1) identification of those inputs which contribute to learning with
estimations of their relative contributions, and (2) financial inter
ventions to test the outcome of attempting to maximize certain inputs.
Alexander and Simmons (1975) provided information on the
relative significance of various school inputs researched in developing
countries (see table 5) . It is interesting to note that most variables
showed differing results, among the countries studied.
Educational Indicators: A Comprehensive Work
Johnstone (1981) dealt extensively with the subject of educational
indicators. He considered the concept and classification of educational
indicators as well as statistical formulas for developing them, indicators
were discussed in terms of the systems model, whether they were
referenced as "normative," " criterion, " or "self-referenced"; whether
they were defined as pragmatic, theoretical, or empirical; and whether
they should be classified as representative, disaggregated, or composite
(which he favored) .
He provided a schema for educational indicators based on the
systems concept (see figure 2) . In this illustration the systems environ
ment included the economic, social, and political systems. The education
systems itself included as input, " resources" and " preference for educa-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE 5
RESULTS OF EDUCATIONAL PRODUCTION FUNCTION STUDIES FOR POOR COUNTRIES
Variable, and its Relationship to Investigator(s) with Country Where the Study Was Done Student Performance
Expected Statistically Significant Non-Statisticaliy Significant, Variable Sign with Expected Sign or with Opposite Sign
Background Variables and Schooling Variables
SES has strong effect at primary and Beebout ( 1972) Malaysia Carnoy (S.F.) (1971) Puerto Rico lower secondary grades for all + Carnoy (R .F .) (1971) Puerto Rico Carnoy-Thias (197*1) Tunisia subjects Epstein (1971) St. Lucia
science* (1972_) '" ‘"-national S'bein-Farrell ( 1973) Chile Simmons-Askoy (1972) Tunisia SES has less effect in developing c .f. developed countries IEA reading (,972'> International S'bein-Farrell (?) ( 1973) Chile SES has diminishing effect through Beebout (English) ( 1972) Malaysia Beebout (1972) Malaysia out secondary cycle Carnoy (1971) Puerto Rico Carnoy-Thias (197*1) Tunisia
I eA science (1972- } '" ‘'•‘national Schooling variables have some effect ? Carnoy (1971) Puerto Rico at lower grades, but a stronger Carnoy-Thias (197*0 Tunisia effect in upper secondary grades IEA reading international IEA science Schooling variables have some effect on + IEA reading ( 1972-) International Simmons-Askoy ( 1972) Tunisia native language beyond primary grades
Individual Schooling Variables
Por-pupll expenditures on school + Beebout ( 1972) Malaysia facilities or teachers Levy (1971) International Thlas-Carnoy ( 1969) Kenya (grade 11) Average class size, or pupil: Beebout (1972) Malaysia teacher ratio Carnoy (1971) Puerto Rico I eA scfence ( ' 972-> International Levy (1971) international S'bein-Farrell ( 1973) Chile Thias-Carnoy ( 1969) Kenya (grade 11) Teacher certification and academic Carnoy (1971) Puerto Rico Levy (1971) International qualification at primary and lower S'bein-Farrell ( 1973) Chile secondary grades Thias-Carnoy (1969) Kenya (grade 11)
(-) Carnoy (1971) PuertoCarnoy-Thias Rico (1979) Tunisia Carnoy-Thias 1979) ( Tunisia Carnoy-Thias 1979) ( Tunisia Carnoy (1971) Puerto Rico Carnoy (1971) Puerto Rico Beebout 1972) ( Malaysia Beebout 1972) ( Malaysia Carnoy-Thias (1979) Tunisia Beebout (1972) Malaysia Thias-Carnoy (1969) Kenya (grade 7) S'bein-Farrell (1973) Chile
' re?dln9 ( 1972-) International (+) scferlce <,972'> '" ‘^national a IEA IEA science 1972-) ( International Beebout 972) ( 1 Malaysia Carnoy (1971) Puerto Rico !1 IEA science 1972-) ( International IEA science 1972-) ( International IEA science IEA IEA reading (1972-) International Carnoy (1971) Puerto Rico S'bein-Farrell 1973) ( Thias-Carnoy Chile 1969) ( Kenya (grade 7) Beebout 1972) ( Malaysia !pa IEA science 972-) ( J International Beebout 1972) ( Malaysia Carnoy-Thias (197 Source: L. Alexander and J. Simmons, The Determinants of School Achievement in 1975, pp. 57-59. (ERIC Document Reproduction Service No. Ed 28 1 991). Developing Countries: The Educational Production Function. Washington, D .C .: World Bank, secondary grades qualification at upper secondary grades Teacher certification and academic Teacher experience at secondary upper grades Homework and free reading Teacher experience at primary and Teacher contract (tenure) at upper Performance and attitudes of class Teacher sex - males at primary and Boarding at secondary grades Double sessions lower secondary grades Teacher motivation room room peer-group secondary grades Size Size of school enrollment at upper lower secondary grades; females at Textbook availability grades at primary at home upper secondary grades Grade repetition Availability and use of library Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. OJ system Political of society and Skills Utilization Future framework of Resources system Social system Economic with and Skills Output Education subdivision Resources Satisfaction XXXXXXXXXXXXXX X Xxxxxxxxxxxxxxxxx and Process subdivision System of Inputs S tru ctu re Distribution Education System- for for Input Education subdivision Resources Preferences Education ;xxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxx Fig. 2. A framework for the development of indicators of education systems. Source: J. Johnstone, Indicators of Education Systems. Paris: UNESCO, International Institute for of Past system society framework Social system Political system Economic Educational Planning, 1981, p. 26 (ERIC Document Reproduction Service No. Ed 212 091). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 44 tion" ; as process, " system structure and distribution of inputs" ; and as outputs, "[utilization of] resources and skills" and "satisfaction with education." In addition, Johnstone (1981) listed a number of indicators already developed (many under the auspices of the United Nations' Education, Scientific, and Cultural Organization— UNESCO), formulas for deriving these, and his selection and categorization according to the systems concept. The following summaries are considered of special interest and informational value: 1. A summary of education—economy studies together with the education variables used (see table 6). These variables include (a) per centage of Cross National Product and percentage of government budget spent on education, (b) numbers of eligible population in education, and (c) literacy rates and levels of education reached. 2. A summary of eight core " input" indicators (see table 7). These include (a) enrollment ratios, (b) rates of change in the ratios, (c) financial (expenditure) indicators, (d) access indicators, i.e., per centage of eligible group in education including accessibility to females, (e) minimum attainment (grade 4), and (f) maximum second level attain ment. 3. A summary of eighteen "process" indicators (see table 8) . These include (a) availability of student places (ratios of enrollment), (b) availability of classes and schools at consecutive levels, (c) rates of change for (a) and (b) above, (d) female involvement (students and teachers), (e) orientation of second- and third-level studies, (f) expected number of years and expected completed grades for an entrant. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 45 TABLE 6 SUMMARY OF EDUCATION VARIABLES USED IN SELECTED COMPARATIVE EDUCATION STUDIES Education—economy Education— socio-political studies studies S tu d y Education variable S tu d y Education Variable Edding (1 962) % of CNP on Jahoda(1961) Level of education education reached S venn il son Enrollment ratio Fox and Miller Combined first-and et al. (1 962) (1965) second-level enrollment ratios Bowman an d % of population Almond and Years of schooling Anderson in primary edu Verba (1963) (1963) cation % of population in post-primary education Harbison and Combination of Myers (1964) second- and third-level enrollment ratios M cC lelland Secondary school C u trig h t Combination of (1966) enrollment per ( 1963) literacy rates and 10,000 population higher education enrollments per 10, 000 population Meyer (1971) Ratio of enroll Rummel Expenditure on edu ments at each (1 972) cation as a % of level government budget first-level pupil- teacher ratio first-level unad justed enrollment ratio second- and third- level students per 10,000 population Source : J. Johnstone, Indicators of Education Systems. Paris: UNESCO, International Institute for Educational Planning, 1981, p. 11 (ERIC Document Reproduction Service No, Ed 212 091). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ■ c CTl Division Division Method of Gin) Glnl coefficient coefficient Combination Variables 1. 1. Index numbers Unweighted 1. 1. Index numbers Unweighted 2. Standard scores average 1. Index numbers Unweighted 1. Index numbers Unweighted 2. Standard scores average 2. Standard scores average Conversion of 2. Standard scores average None None None None INDICATORS TA B LE 7 Variables Used in Calculation Total amount spent on education as a proportion of the variables nominated for input Indicator 1 and and number of teachers per 10,000 population variables listed for input indicator Percentage 3 of enrollments with in a nominated certain groups and characteristic percentage of total population specifically the total enrollment of girls in each region CNP CNP and the total publicproportion expenditure of the on total education government as a expenditureVariables measuring the rate of change in both of the Adjusted enrollment ratios for first- and second-levels Proportion of the total population of an age theoretically in in those groups same Proportion of the total population of an age theoretically who are enrolled In that grade or above (for two years) Variables measuring the rate of change in each of the enrolled in gradeor above SUMMARY OF CORE EDUCATIONAL INPUT Source: J. Johnstone, Indicators of Education Systems. Paris: UNESCO, International Indicator Name In In the provision of education educational attainment Indicator involvement Equality of access of girls to education indicator changing human resource level educational attainment Change In the minimum Change in the maximum second- . 1. 1. Involvement of human resources 3. 3. Financial involvement 2. 2. Development emphasis on 5. 5. Equality of access a. a. Financial expansion 8 Institute for Educational Planning, 1981, p. 150. (ERIC Document Reproduction Service No. Ed 212 091). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. values Combination Method of Unw eighted Unweighted 2. Ratio of C average average average Index numbers Index numbers Unweighted Index numbers Unweighted Standard scores average Standard scores Standard scores . 1. 2. 1. Index numbers 1 2. 1. Index numbers Unweighted 2. Standard scores 2. 2. Standard scores average No C index No 1. C index TABLE 8 Ratio of enrollments(each ratio at third-level separately for to males second-level and females) Ratio of enrollmentsRatio of enrollments at special at education second-level to first to first level level Ratio of enrollments at pre-school to first-level second- and third-levels Ratos of change for all variables listed for indicators 1-3 1. Ratio of numbersRatio of of classes numbers In of specialclasses educationat second-level to first-level to first-level Percentages of second-level studentsPercentages enrolled fieldsof In third-level general students enrolled In nominated second- and third-levels Percentages of enrollment who are female at the first-. Ratio of numbers of classesSimilar at variables third-level to those to above second-level except numbers of schools Ratio of numbers of classes at pre-school to first-level education and teacher training courses Percentages of teaching force who are female at the first-, are used Instead of numbers of classes SUMMARY OF CORE EDUCATIONAL PROCESS INDICATORS classes (4), Indicator Name Variables in Calculation Conversion student places at ( ( 5) and schools 6) ( level studies for students consecutive levels education schools at consecutive the provision of levels Development emphasis levels level studies expansion of system Female Involvement in Orientation of third- on i]uaritilalive classes at consecutive 1. Relative availability of 3. Relative availability of 7. 2. Relative availability of 8. Orientation of second- 9. 6. i(- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10. Expected number of Promotion and repetition rates for all grades In the first- No Special formula years of education level and/or second-level received by an entrant 11. Expected number of Promotion and repetition rates for all grades In the first- No Special formula grades of education level and/or second-level completed by an entrant 12. Availability of Percentage of teachers in various qualification categories No C index qualified teachers 13. Distribution of Percentages of qualified teachers In each area of a country No Clnl coefficient qualified teachers as well as percentages of total teaching force in each area 1 15. Regional educational 1. Promotion and repetition rates for each grade in each region No 1. Special formula flow differential 2. First and final grade enrollments and total enrollments for ♦ Stand ard the level analyzed Deviation 2. Ratio of Clnl Coefficients 16. Index of structural Variety of characteristics of institutions No Cuttman scale differentiation 17. Dependence on a Percentages of all expenditure on a system provided by the No C index single fund source national government of a country and by state or provincial governments 18. Education cost Unit cost for first-, second- and third-level over a given time Index numbers Laspeyres indlca tor period and enrollments at each level in a base year formula Source: J. Johnstone, Indicators of Education Systems (Paris: International institute for Educational Planning, 1981, ERIC Document Reproduction Service, No. ED 212 091). 49 (g) availability and distribution of qualified teachers, (h) male/female educational flow differential, (i) regional flow differential, (j) index to categories of institutions, and (k) funding and cost indicators. 4. A summary of five "output" indicators (see table 9). These include (a) achievement levels, (b) an academic production index, (c) specialization of graduates, (d) proportion of entrants graduating, and (e) satisfaction with educational performance. 5. A summary of five " educational system— societal congruence indicators" (see table 10). These are (a) educational profile of the un employed, (b) availability of the mass media, and (c-e) rates of return on the three levels of education. A listing of the formulas referred to in tabies 7-10, as well as others of interest are included in appendix B. Literature Related to the Variables Under Study In addition to the literature already cited, literature and research findings relevant to the variables included in this study are presented. Eight predictor variables are reviewed below: Sex, Age, Achievement Bernard (1972), in his book Psychology of Learning and Teaching, states that boys as a group show greater variability than girls in the range of intelligence. More boys than girls are defective in intelligence but also more boys than girls are superior. On the average, males demonstrate slight superiority over females in general reasoning, arithmetic reasoning, ability to detect similarities, and certain aspects of general information. They tend to excel over girls in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50 TABLE 9 SUMMARY OF CORE EDUCATIONAL OUTPUT INDICATORS Variables Used in Method of Indicator Name Conversion Calculation Combination 1. Achievement Scores on either individual No A ddition level items or on sub-tests 2. Academic Numbers of third-level 1. Index Weighted or production graduates obtaining numbers Unw eighted index various types of degrees 2. Standard average or scores Numbers of second-level graduates completing various types of education 3. Specializa Proportion of third-level No 1. C index tion of graduates in nominated 2. Ratio of graduates fields of study C indices 4. Proportion Promotion, repetition and No Special of entrants graduation rates for all formula who finally grades in the f ir s t - and/ graduate or second-level 5. Satisfaction Ratings on an ordinal scale No W eighted w ith the to indicate degrees of average or performance satisfaction or responses Laspeyres of education on a nominal scale to formula systems specific features, problems, etc. of the education system Source: J. Johnstone, Indicators of Education Systems. Paris: UNESCO, International Institute for Educational Planning, 1981, p. 175. (ERIC Reproduction Service No. Ed 212 091). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 51 TABLE 10 CORE INDICATORS SYSTEM— SOCIETAL CONGRUENCE M ethod o f Indicator Name Variables Used in Calculation Conversion Combination 1. Educational Percentage unemployment rates No G ini p ro file of for selected educational coefficient th e u n attainment groups and per em ployed centage of 15-64 age group in each of these groups 2. Availability Rates per 1,000 population for 1. S ta n d a rd 1. Addition of the mass the availability of radios, scores 2. Multipli media television, daily newspapers 2. In d ex catio n and non-daily newspapers num bers 3. Rates of re tu rn fo r first-level Many estimates including V ario u s education income foregone, forecast earnings, etc. (See 4. Rates of Psacharopoulos (1975), re tu rn fo r Thias and Carnoy (1969), second-level for full details.) education 5. Rates of re tu rn for third- level education Source: J. Johnstone, Indicators of Education Systems. Paris: UNESCO, International Institute for Educational Planning, 1981, p. 181. (ERIC Document Reproduction Service No. Ed 212 091). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 52 speed, coordination of large body movements, perception, and me chanical aptitude. Girls tend to have slight superiority in memory, language usage, manual dexterity, numerical computation, and perceptual speed. Bloom (1964, cited in Bernard, 1972, p. 155) found that at least during the ages of seven to sixteen, there is little justification for separate analysis of boys and girls on intelligence test data. It should be noted that in the Common Entrance Examination the cut off score for boys is lower than that for girls—perhaps supporting the observation of greater variability in intelligence or achievement scores of boys as a g ro u p . Numbers of Years in School Good, Biddle, and Brophy (1975, p. 64), in their review of elementary-school studies, indicate that, based on two general reviews by Rosenshine and Furst (1973) and Dunkin and Biddle (1974), several studies have demonstrated that attendance at elementary school makes a difference and/or that some elementary schools are more successful than others. At the most fundamental level several studies have compared children who have stayed in school with children who have left school early. The studies all show that school has considerable impact not only on infusing knowledge but also in developing children's cognitive abilities and problem-solving skills. One such study was conducted in Prince Edward County, Virginia. The county closed its public schools from 1959-1961 rather than comply with a court order to integrate the schools. Some Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 53 children in the county did not attend school during this period. Comparisons of children who stayed in the county but did not attend school with peers who went to private schools in the county or who left the county to attend school elsewhere showed that children who did not attend school failed to progress and in many cases seriously regressed in their intellectual abilities (Green et al., 1966). A second study yielding similar results was conducted by Schmidt (1966). This research was done on a South African community where education is not compulsory. Children who started school early and stayed in school longer were compared with children who began later or did not stay in school very long. Among other things, his comparisons revealed a strong relationship between years of schooling and intelligence test scores even with age held constant and socio economic status (SES) controlled statistically. Good, Biddle, and Brophy (1975) go on to comment that all of these studies no doubt involved a certain amount of selection bias, because even where SES was controlled it is likely that children who attended school came from families who placed a greater value on education than families who did not. Thus, differences seen in these studies reflect not only the effect of school but probably also differ ences in family values. Nevertheless, these and similar data leave little doubt that schooling has important effects upon cognitive development. Children with more schooling not only learn more facts, they develop their thinking and problem-solving abilities more fully. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 54 Teacher/Schools Differential E ffects Barak Rosenshine (1971, pp. 66-95), in How Teachers Make a Difference, makes a general note on the lack of research on how teachers make a difference. While the studies cited may show certain relationships, further research would be necessary to show what curriculum inputs and what aspects of teacher presentation make a difference in pupil outcomes. On the question of some schools being more effective than others. Good, Biddle, and Brophy (1975, pp. 65, 66) cite an especially interesting study done by Brookover, Gigliotti, Henderson, and Schneider (1973). These investigators took into account the socio economic and racial composition of school populations by matching schools within categories before making comparisons. Then the investi gators identified pairs of schools within each category that were similar on every variable measured except student achievement. These data demonstrated that schools can differ significantly from one another even when student populations are carefully matched. Students, teachers, and principals of twenty-four schools were asked to fill out questionnaires. The responses were then analyzed and related to measures of school success in providing student learning. These results showed that the most important single negative factor was the student's sense of futility. Students in schools that had the worst records showed the strongest sense of futility. These students were cynical and apathetic, feeling that very little of value went on in their schools and that they had little to say about it. Other important differences between successful and unsuccessful schools Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 55 included teacher perceptions of future evaluations, teacher expecta tions of the students, and the teachers' reported willingness to push students to achieve. In short, when teachers believed that students could learn and worked to see that they did, students learned; when teachers did not think the students could learn and did not work very hard to see that they did, the students did not learn much. These data are consistent with those from studies reviewed by Good, Biddle, and Brophy (1975); a relationship between teacher expectations and student achievement is shown. Whether the unit of analysis is the individual teacher or the entire school, teacher expectations and student learning tend to go together. Good et al. (1975) also point out that several studies indicate that curriculum materials provided to teachers tend to set the pace for the class by fixing the expectations of both teachers and students concerning what is appropriate for students to learn. A number of studies was cited as well as the review of thirty-two studies of innovative curricula covering elementary school through college (Walker & Schaffarzink, 1974). The conclusion is that much of what students learn is a function of teacher and student expecta tions and that these are determined in large part by the curriculum used (p. 67). It was also affirmed that some teaching methods are more effective than others even when the curriculum is identical. Studies of affective behavior of teachers are also provided. Several studies have indicated the importance of certain aspects of teacher personality—especially the affective or emotional aspects. Certain Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 56 behaviors may produce different learning outcomes with students from different racial or class backgrounds. Good, Biddle,and Brophy (1975) conclude their review by stating that, although a variety of elementary-school research has been considered, the review has been illustrative rather than exhaustive. Two main points were brought out: 1. Teachers differ significantly from one another both statistically and practically in their relative impact on student-learning and student-affective variables such as self-concept and attitudes towards learning and school. 2. Much is already known and more is being discovered every day about what kinds of teacher behaviors are appropriate or inappropriate for particular kinds of students in particular situations. Although a teacher is partially an artist with a unique individual approach, teaching is (or should be) only partially an art. As knowledge about effective teaching accumulates, teaching should increasingly become an applied science .... The skilled teacher will be an individual who has mastered a large body of principles and skills, and who is capable of diagnosing a situation correctly and deciding which of the many options available to him are appropriate to the situation. (p. 79) Educational Television The vast majority of research studies on the effectiveness of educational television had established, by the late 1960s, that there were no significant differences in measured performance between students who were instructed via television and those who were taught directly. A few studies showed differences in favor of television and a few found differences in the opposite direction (Reid 5 McLennan, 1967). Gordon (1965) concluded that television in an educational setting may be used for: (1) enrichment; (2) cooperative or team teaching. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 57 and (3) total teaching. Of 393 comparisons between instructional television and live instruction, there was no significant difference in 65 percent of the cases studied. There was an apparent tendency for televised instruction to provide more effective learning in the primary grades. There was also variation in the effectiveness of television teaching according to subject matter. For example, at the time of the study, language skills and health and safety education showed a fair amount of effectiveness. Mathematics, science, and social studies showed outstanding success for the groups tested, while history, humanities, and literature had the smallest measure of sucess. Tele vision students held their own in comparative tests of critical thinking, problem-solving, and sundry other incidental virtues of education (pp. 66, 84, 85, 87). In addition, learning from televised classes, like all learning, seems to be a correlate of the motivation of students— a class with low motivation to learn will probably learn as little from a live instructor as from instructional television (Cordon, 1965, p . 8 6 ). Carpenter (1963), cited by Cordon ( 1965), listed factors based on research at that time, which seemed to affect positively, student learning in instructional television courses of any kind. These factors were: 1. The quality of sources and resources used. 2. The way in which materials are related and organized. 3. The surroundings on the television screen in which materials are selected and organized. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 58 4. The characteristics of the students in terms of aspira tion, motivation, abilities, previous training, and state of physical and mental health. 5. The nature of students' responses to the materials on television. 6. The rewards, penalties, or reinforcements which occur with or during televised instructions. 7. The evaluations and comparisons of achievement rates and performance levels to which students are subjected. (Cordon, 1965, pp. 87, 88). School Libraries Caver (1963) points out that the most authoritative pronouncements on school libraries were issued by the American Association of School Librarians. An arresting paragraph in this docu ment reads: Whatever form the soul-searching regarding the education of youth may take, sooner or later it has to reckon with the adequacy of the library resources in the schools. In the educa tion of all youth from the slowest learner in kindergarden to the most intelligent senior in high school an abundance of printed and audio visual materials is essential. (American Library Asso ciation—ALA— 1960, Standards for School Library Programs, p. 3) It was further stated that educators and librarians agree that even under the best conditions public library service is not accessible enough for all school children. The advent of video and computers has introduced teachers to even more flexible resources for teaching by audiovisual technology. These resources are not, however, being utilized in Jamaican primary schools. Socioeconomic Status: Classifi cation Index A convenient socioeconomic measure is the McGuire White (1955) Index I of socioeconomic status which provides categorizations Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 59 based on three sources of data: (1) occupation of family head, (2) source of family income, and (3) education of family head. A weighting method for aggregating the variables is described and a rating provided. These are included at appendix C. Literature on Jamaican Education Relevant to the Study The Common Entrance Examination Vernon (1961), in his report to the Minister of Education on Selection for Secondary Education in Jamaica,made a number of state ments and observations relevant to this study: 1. The majority of pupils obtaining free or grant places in Kingston schools then had most of their education in primary schools but also attended infant, private, or preparatory schools (p. 37). 2. Boys tended to start a little behind (girls) but appeared to catch up after three years in secondary school (p. 37). 3. Age of entry seemed to have no clear effect. Probably those aged 11$ did rather better in secondary school than those under this age (p. 37). 4. Girls did better than boys on all three Moray House tests, but especially in English. Compared with the United Kingdom 11+ children, Jamaican candidates were relatively more advanced in mechanical but poorer in problem arithmetic, and somewhat weaker in a test of non-verbal reasoning (p. 38). 5. In discussing intelligence and the intelligence tests, Vernon (1961, p. 43) pointed to a distinction made by Hebb (1949) between Intelligence A—defined as an innate quality of the central Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60 nervous system and brain—and Intelligence B— which is the result of Intelligence A interacting with environment in the growing child. Such an intelligence is similar to the conception of health— fundamentally limited by genetic endowment but also dependent on environmental conditions from conception onwards. Vernon (1961) considered that the intelligence under reference is a middle-class European and American invention—a kind of intelligence particularly well-adapted to scientific analysis, control, and exploitation of the physical world, for large-scale and long-term planning and the carrying out of materialistic objectives (p. 45). 6. The intelligence tests utilized in the battery attempted to measure skills which are somewhat less affected by schooling than are such measures as English and arithmetic. Home and school environ ment would nevertheless affect the measures on these tests (p. 46) . 7. Studies of maturation have shown that western education is apt to introduce children to reading, numbers, and grammatical and mathematical operations before they are ready for them and that much quicker progress is made when they are postponed for a year or two; also that such children do completely catch up with those who made an earlier start (p. 49). 8. In 1959 nearly twice as many girls as boys entered for the examination. Because of a 10-point lower limit in total marks arbitrarily set for boys, 853 boys as against 902 girls secured places (p . 5 7 ). 9. In England the average intelligence and English quotients for boys were a little lower than for girls, though they often exceeded girls in arithmetic. Their test spread was also wider Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 61 with relatively more very high and very low scores (p. 57). Manley (1963), in his study Mental Ability in Jamaica, analyzed the records of a random sample of 1,730 children from the 1959 entries for the Common Entrance Examination. These children were given the Moray House Verbal Reasoning Test 43, Moray House Arithmetic Test 19, and Moray House English Test 19. Manley discussed the sociological and educational implications of the results and examined some of the possible reasons for these. The means of the test scores were all between 85-90 with standard deviation of about 10. This compared with the means of the United Kingdom test scores in the region of 100 with standard deviation of 15. The Jamaican spread was a skewed distribution. Table 11 displays the Jamaican results (Manley, 1963, p. 5). TABLE 11 MEAN QUOTIENTS IN MORAY HOUSE BATTERY (1959 - TOTAL SAMPLE JAMAICA) N o. o f Intelligence A rith m e tic E nglish Examinees Quotients Q uotients Quotients Total Boys 657 87.1 89. 9 8 5.6 262.6 G irls 1, 073 87.5 89.1 86. 3 262.9 B oth sexes 1, 730 87. 3 89.4 86.1 262.8 S ta n d a rd d eviatio n 10.8 10.3 10. 5 25.4 The results were classified by parishes and then oy all rural areas as against Kingston and suburbs. Rural areas, however, did include rural towns. The results in respect to candidates from the primary schools only (candidates from the high schools and private schools excluded) are summarized in table 12 (Manley, 1963, p. 8). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 62 T A B L E 12 MEAN QUOTIENTS— PARTIAL SAMPLE—MORAY HOUSE BATTERY 1959 - JAMAICA (PRIMARY SCHOOL ENTRANTS ONLY) No. o f Intelligence A rith m e tic E nglish Parish Examinees Q uotients Q uotients Quotients Total All ru ra l areas 1178 84.2 87.1 82. 7 253.7 Kingston (and suburbs) 228 91.3 92.0 8 9 .6 272.9 Total p rim ary 1406 85.3 87.8 8 4 .0 257.1 Although children from rural parishes obtained a larger percentage of free places than any other single group, they obtained far fewer than their proportionate numbei— a disproportionate number going to high- school pupils, nearly all of whom were taking the examination for the second time. Again, girls from all types of schools showed a slight margin of superiority in intelligence and English (pp. 9, 11). Manley (1963) made special note of the fact that boys appeared to be dropping out of the system. Whereas there were twice as many girl candidates as boys in the Common Entrance Exam of 1959, there were also fewer boy candidates in 1961 for the Technical High Schools Entrance Examinations which are set at age thirteen. He concluded that if these boys had not dropped out of the educational system, they had failed to approach the secondary stage (p . 12). Based on crude categorizations of parents' occupations, Manley (1963) found that children drawn from certain social classes Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 63 showed a decided superiority (p. 18). Tables 13 and 14 illustrate this finding (pp. 15. 16). T A B L E 13 SOCIAL CLASS AND TEST PERFORMANCE—MORAY HOUSE TESTS (JAMAICA) 1959 Intelligence Arithmetic English Group Quotients Quotients Quotients Total 1. Professional and managerial 98.6 96.4 98.4 293.4 2. Teachers 9 2 .3 94.4 90.4 277.1 3. Clerical - 91.5 92.1 89.3 272.9 4. Skilled & semi-skilled 8 7 .5 90.3 84.4 262.2 5. Farm ers 8 2 .9 85.6 83.7 252.6 6. Unskilled 82.1 85. 7 82.4 250.2 T A B L E 14 THE SIGNIFICANCE OF DIFFERENCES BETWEEN THE MEANS OBTAINED BY DIFFERENT SOCIAL GROUPS Intelligence Arithmetic English G roup Q uotients Quotients Quotients 1. Professional and Managerial and Teachers 6 .3 * 2. 0 8 .0 * 2. Teachers and Clerical 0.8 4.1 * 1.1 3. Clerical and Skilled and Semi-Skilled 4.0* 1 .8 * * 4 .9 4. Skilled and Semi-Skilled and Farmers 4 .6 * 4 .7 * 0 .7 5. Farmers and Unskilled Workers 0.8 0.1 1 .3 * p < .01 ** p < . 05 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 64 An "elite" group from wealthy homes who attended prepara tory schools showed even greater superiority in performance. Their overall mean score was 19.7 points above the mean of primary school children classified under the professional/managerial group. A sizeable proportion of children from the professional managerial group appeared to attend rural primary schools. It was pointed out that this was not as unusual as might be expected since primary schools might have been the only ones available in those areas and all social groups either used them or sent their children away to boarding schools—an expensive alternative (p. 20). Correlates of Achievement in a Common Entrance Level Examination The landmark Jamaican study of factors influencing per formance of primary-school pupils in an examination of similar level to the common entrance level was that of Reid (1964). Specially prepared tests in English and arithmetic which produced normalized standard scores with a mean of 100 and a standard deviation of 15 were the criterion measures used. The averaged scores of each school were treated as the unit. A test of mental ability devised by Vernon for use with Jamaican children was also used. Each school, therefore, had a criterion score made up of the averaged scores in English and arithmetic combined, as well as an averaged score of mental ability, together with a score or rating on each independent variable. The independent variables were: 1. School Size—the average enrollment of children for the entire school. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 65 2. Class Size— the number of children enrolled in the class from which the sample was drawn, i.e., the highest in the Junior department. 3. Attendance— the average number of sessions, i.e ., half days for which the children attended in the whole of the previous school year expressed as a percentage of the total number of possible sessions. 4. Length of Schooling— the average number of years pupils attended school up to the time of the sampling, inclusive of infant or "basic" schooling. 5. Teacher Education or Professional Qualification— coded according to the levels of education and grade responsibility. 6. Buildings rated on a five-point scale following assessment by staff of the Building Section of the Ministry of Education. 7. Accommodation— the amount of teaching floor space available per child for the entire school. 8. Equipment, Library, Radio, Piano—points given according to the number of books, for example. 9. Family Pattern— the proportion of children sampled in each school living with two (2) parents, one (1) parent, or guardian, expressed as a percentage of the school sample. 10. Family Size—number of living children, siblings as well as cousins and other kin residing with the family. 11. Parents'Reading Habits—rated according to whether either parent was a member of a lending library, read the newspaper daily or weekly, and according to the number of books in the home. 12. Parental Interest— number of visits to school, member- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 66 ship of PTA, attention to term test results, help with home work, number of visits to school, concern for pupil getting a high-school p lace. 13. Economic Level of the Community—parents classified in occupational categories: Unskilled laborers Small farm ers Semi-skilled workers Skilled workers C lerical and small p ro p rie to rs Managerial and professional 14. Progressiveness of Community—whether the community had certain community facilities such as a public library, electricity, e tc . Reid's (1964) analyses began with the Pearson product- moment correlation of each independent variable with the criterion achievement score. This was followed by factor analyses which, he pointed out, contained an element of arbitrariness since, as Burt (1940) held, the analyst must first have formulated a theory and then arbitrarily chosen variables to be included in the research; factor analysis, therefore, cannot of itself permit a determination of causal influences. Reid's (1964) factor analyses yielded four factors, three of which he considered significant: 1. A general intellectual factor, which was contributed to by the intelligence of the child, early schooling, well-trained staff, a progressive communal environment, and interested and educated parents with not very low incomes. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 67 2. A physical and social environmental factor, represented by good buildings, generous space, small classes, and, mainly, the small two-parent family. 3. A school-environmental factor, to which size of enroll ment, teacher status, early start of schooling, and equipment contributed. Reid (1964) initially selected the following variables as the most promising to include in his final analyses by means of multiple corre lation, i.e ., correlation of one criterion variable with a battery of o th e rs : 1. Parental interest 2. Reading habits 3. Occupational level 4. Progressive community 5. School size 6. Years of schooling 7. Teacher status 8. Maternal family headship 9. Equipment 10. Buildings 11. Class size Intelligence was omitted since it would have been difficult to argue whether it should be treated as a cause of attainment or as an effect of the contributory variables. Items 1 and 2 were combined into a new variable termed "Parental Education." The ten finally selected variables and their regression weights are listed in table 15 (Reid, 1964, p. 1 3 5 ). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 68 T A B L E 15 REGRESSION WEIGHTS OF PROMISING VARIABLES V a ria b le Regression Weights Parental education .5395 School size .1868 Teacher status .1485 Years of schooling .1271 Occupational level .0710 Progressive community .0153 E quipm ent - .0004 Maternal headship - .0979 Buildings - .0055 Small class - .1488 Reid (1964) concluded that the most essential contributing factor to school achievement was good home background, e.g ., the education of the parents themselves and their interest in their children's education. When this was taken into account, socioeconomic level had very little further contribution to make to predictions, for it was already covered by parental educational level. Years of school and qualifications of staff did make a small independent contribution, and the larger school with good equipment provided a definite advan tage. Other variables—progressive community (favorable) and maternal headship of family (unfavorable)—were considered largely covered by the previously listed variables so that any additional contribution appeared insignificant (pp. 135, 136). Reid's (1964) study looked at the gross effects of primary schooling. Measurement and computational limitations apparently did Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 69 not allow for profiling of the factors influencing individual pupil achievement. The factor analyses depended on his (1964) inter pretation of the environment, which influenced his decisions as to what factors to enter in the equation. His analysis was on gross school rather than individual measures and the results focused on non school rather than school effects. The present study, while utilizing six of the variables of this earlier study, focuses on school effects and examines these, not on gross measurements of school means but on individual pupil measures, a procedure made possible by advanced computing machines. Roach (1981), in a study of 418 boys and girls aged 10-12 chosen from five urban elementary schools in Jamaica, used mathema tics achievement as the dependent variable and sex, family size, birth order, occupational level, father's presence, preference for conceptual style, field dependence, reading achievement, and mental ability as the independent variables. Step-wise regression analyses yielded mental ability, reading achievement, and family size in that order as the best predictors of achievement in mathematics. In a wide-ranging review, Christopher Bagley (1979) in his article "A Comparative Perspective on the Education of Black Children in Britain "showed that 1. The educational standards reached by the Jamaican popu lation lagged behind those of other islands. 2. One reason for underachievement of Jamaican children in English schools seemed to be the relatively low level of educational achievement of their parents (this stood in contrast to the educational achievement of parents [and their children] from other Caribbean areas). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 70 3. Jamaican class-room absenteeism was high among boys and teachers interacted more favorably with and gave more positive feedback to pupils who were of higher socioeconomic back ground, who were fair rather than black, and who were girls rather than boys (Girling, 1974), 4. Well-motivated children whose parents could afford to make the sacrifice benefitted from a parallel system of education— private classes. 5. Authoritarianism and its linked style of teaching involving rote learning seemed inimical to creativity and breadth of enquiry. A number of studies from various cultures shows that the method, on the whole, is not successful in training children in lasting skills (Simmonds, 1976; Wilson, 1973). 6. Curriculum was not relevant to the needs and culture of the mass of the population. The curriculum is a Europeanized one which has ignored the linguistic culture of the masses. Dwyer (1971) showed that the concept formations transmitted by Jamaican Creole are severe impediments in taking tests such as the Moray House instru m ent. At this point it seems useful to cite the results of a study done by Ragbir (1977) in respect to motivation of high-school students. Although not focused on primary-level pupils, there is some relevance to this review. Studying academic-achievement motivation in three types of male Jamaican high-school students, Ragbir (1977) compared the three types based on certain combinations of high and low levels of the variables: intelligence, academic achievement, and social class. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 71 Certain motivational influences were identified which distinguished between middle-class high-ability high achievers, and middle-class high-ability low achievers. It was found that a group of working- class high-ability high achievers showed a middle-class type of academic-achievement motivation on par with their middle-class high- ability counterparts. Further, this group of working-class high achievers showed the middle-class type of academic-achievement motivation to a greater degree than middle-class high-ability low achievers. He concluded that,since it has been widely found that the working-class student performs at a lower level than his middle-class counterpart of equal intelligence, links between motivation and achievement could be established. Smaller Studies It is relevant and worthwhile at this point to cite the results of certain smaller studies done by Bachelor of Education students at the University of the West Indies. On Attendance/Absenteeism. Clunie (1983), in a study of 110 grade-6 pupils in Primary and All-age schools , found that peak periods for absenteeism were Fridays followed by Thursdays and the end of each school term, particularly the Christmas term. Parental attitudes, parents' low socioeconomic status, teachers' attitudes towards students and work, streaming (mainstreaming), the program of activities, and physical conditions of the school were factors contributing to absenteeism. Erskine (1979), in a study of 186 grade-8 students from the All-age schools in rural and urban Jamaica found, among other things. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 72 that over the period of All-age school exposure, from ages 6^-15 (approximately eight years) many working-class children attended school for less than 4-5 years. Students of higher socioeconomic status had significantly less absenteeism than students from lower socioeconomic status. Absenteeism was greater in the mornings than in the afternoons, and, during the week, Friday was the peak day, followed by Thursday. Low nutritional levels and lack of parental interest were also contribu ting factors. On Teachers. Monteith ( 1978), utilizing a sample of twenty schools and 119 teachers, found that distribution of trained teachers and teacher retention were more favorably weighted towards urban than rural and deep rural areas. Retention rates were higher among older rather than younger teachers, and a significant relationship existed between teacher retention and certain community and school factors. Fenton (1979), on a sample of 100 teachers from primary and preparatory schools in Jamaica, compared with 50 primary-school teachers in Montserrat found, among other things, that 1. The strongest single influence in the schools was felt to be the principal. 2. Jamaican primary-school teachers perceived themselves as performing under very difficult and constraining circumstances as opposed to preparatory-school teachers who did not thus perceive themselves. On St. Thomas, a Parish in the Present Study. Kean (1974) studied a sample of seven schools on factors influencing school out comes. He found them underproducing in terms of percentage of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 73 pupils in the relative age cohort presented at the Common Entrance Examination. Towards Indicators in the Jamaican Educational System Johanson (1969) proposed an improved information system for the Ministry of Education in Jamaica. His systems approach implied considering the problem as a whole. The project produced three principal results: 1. A criterion model for analysis. 2. A model of a total information system for education, prescribing units of information grouped into three general areas— pupils, personnel, and property. 3. A first step toward achieving the total system for Jamaica, the design of a detailed system of pupil accounting. This would entail (a) registering all children at age six, (b) numerating pupils currently enrolled, (c) registering all independent schools, (d) establishing de facto school zones, (e) revising procedures for enrollment and admission, and (f) keeping track of all movements by pupils between schools. Johanson's (1969) study focused on the organizational (and to some extent— the management) aspects of the education system. Quantitative Information Other relevant literature on Jamaica related mainly to quantitative information and utilized in this study include: 1. The Annual Report of the Ministry of Education. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 74 2. Education Statistics (1979/80). 3. The Directory of Educational Institutions, 1976-77. All are published by the Ministry of Education, Jamaica. The Annual Report 1975-76 announced that in addition to the administration of the usual examination, a national test in reading was administered in 256 of the 775 primary schools. The booklet. Education Statistics, published annually since 1975/76 (except for 1978/79), contains the fo llo w in g : 1. "Summary of Educational Statistics," showing data on numbers of educational institutions, by type, together with enrollments (male/female), and numbers of teachers categorized as trained and pretrained. 2. "Plant Statistics," showing numbers of institutions by type and by parish, capacity, enrollment, and the statistic percentage excess enrollment. 3. "Teacher Statistics," showing teachers in various categories (i.e. by grade), by type of educational institution, by sex, and by parish. Teacher-pupil ratios in the various levels of institutions are also provided. 4. "Student Statistics," showing data on enrollments in the public institutions at the various levels,by age, by grade, by sex, and by parish. Data on numbers of repeaters, newcomers, and those promoted are given. Data on enrollments in private primary and secon dary institutions, in the teacher training institutions, and in the University of West Indies are also provided. 5. "Examination Statistics," showing results of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 75 (a) the Common Entrance Examination to secondary level institutions ("High" Schools and "Technical" high schools); (b) final examinations of the Teacher Training institutions; (c) degrees awarded at the University of the West Indies; Data at (a) and (c) above are categorized by sex. 6. "Financial Statistics," showing (a) size of the education budgets—capital and recurrent—for the year of interest as compared with pro visions for other Ministries and the Departments not attached to Ministries; (b) the functional classification of the Government's b u d g et; (c) a program allocation of the Ministry of Education's budget for the year of interest and for the preceding two years; (d) a three-year overview of the changes in the levels of allocations to the various programs within the Ministry of E d u catio n . Educational Television: Jamaica It is noteworthy that this media form was not in the schools in 1964, the year of Reid's study. Introduced in 1965 the success of Educational Television (ETV), in terms of utilization, has been less than outstanding. Ministry of Education 1978-80 reports concluded that, although the programs were well received, there were major problems Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 76 contributing to the under-utilization of ETV in the schools. These centred around: 1. Too few sets in schools (ninety-four schools were, at the time, requesting sets). 2. Poor maintenance of sets. These factors contributed to scheduling problems. Table 16 shows the results of a relevant survey. T A B L E 16 ETV UTILIZATION SURVEY— 1978/79 RESULTS No. of No. N o. of N o. o f Schools School Schools O w ning Sets D e fe c tiv e Without T y p e Responding Sets Owned Sets Electricity Other P rim ary 74 41 58 37 6 2 A ll-A g e 139 45 52 35 23 3 Secondary 14 13 30 20 0 0 and High T o tal 227 99 140 92 29 5 Source: Jamaica, Ministry of Education, "Education Broad casting Service ETV Reports, 1978-80." (Mimeographed). School L ib ra rie s : Jamaica The "Schools Library Service Annual Report (1979-80)" lists several relevant points. First is that the majority of older primary schools were still without suitable library accommodation. Next, new schools with library accommodation were given generous bookstocks. Third, some older schools were provided library rooms through the efforts of the Parent/Teachers' Associations. One such library room was opened at the St. Richard's Primary School in 1979. Fourth, keen Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 77 interest was shown by some principals and teachers in getting their libraries to function despite drawbacks. Data on book stocks supplied to primary schools of all categories taken together are shown in table 17. T A B L E 17 SCHOOLS WITH BOOKSTOCKS, SUMMER 1979 Number of Books Received U p to 500 500 to O ver Total Number of 1000 1000 Schools Served , by Parish Parish Number of Schools Serviced K ingston 23 11 9 43 S t. A n d rew 47 18 12 77 S t. Thomas 25 10 2 37 Total 95 39 23 157 Source: Jamaica, "Schools Library Service Report for Period April 1979 to March 1980," Kingston, Jamaica. (Mimeographed). Sum m ary This literature review provides an historical and descriptive overview of economic, demographic, and social indicators. The dis cussion on educational indicators located these as one segment of the social indicators which could be partitioned according to the elements of the systems model. They could be classified as: 1. Context indicators 2. Input indicators 3. Process indicators 4. Output indicators. In terms of the management of the formal education system. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 78 some of these indicators, particularly context indicators such as socio economic status, are not directly "manipulative." The "manipulative" indicators are the "inputs" and "processes" of the school system. Focus was placed on the question of whether there could be a single significant composite indicator, and whether indicator variables should be utilized in economic analyses. This chapter has reviewed literature on the influence of school as well as home effects on the educational product. It has established (1) that schooling is a necessary input; (2) that there are differentiated effects of schooling depending on teacher training, ability, experience, remuneration, and motivation, and their interaction with pupils; (3) that curricula, media, equipment, and adequacy of accommodation are important manipulative variables also; (4) that pupil motivation, whether engendered by the home or the school, is of critical importance; (5) that where motivation is not fostered by the home, it may be encouraged by the school; and (6) that there are physical and developmental variables associated with sex and age to be borne in mind. As measured by research the impact of these variables may differ in significance from country to country and from time to time. In so far as the Jamaican educational universe is concerned, it has been shown that the conditions generated by poverty and the cumulative effects of poor schooling militate against high achievement in the primary-school population, since the important home effects are thus adversely affected. The school effects in primary education could also be regarded as deficient in certain areas. Intelligence of the Jamaican child must therefore be understood as innate ability fundamentally determined by genetic endowment but moderated by environmental Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 79 factors from conception onwards (Vernon, 1961). In so far as a statistical information system for Jamaica is concerned, there has been a proposal for an information system—a model of a total information system and the design of a system of pupil accounting. It is also shown that the Ministry of Education compiles considerable data and a certain number of statistics on the educational system. The present study searches for a limited number of indicators, statistics of normative interest, descriptors proven by research, where necessary, to be capable of providing early evidence of progress towards or regression from primary-education objectives. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER III METHODOLOCY Overview This was, first, a correlational/predictive study which inves tigated the extent to which variations in certain primary-schools, inputs and pupil personal variables corresponded with variations in the achievement scores of a sample of Jamaican primary-school pupils in the 1980 Common Entrance Examination. Second, it was an exploratory field study aimed at securing the perceptions of a sample of primary-school teachers regarding sig nificant indicators of educational performance in the primary-school system in Jamaica. Third, drawing on the findings of these two studies and on the relevant literature, the study proposes a set of indicators for monitoring the primary-school system in Jamaica. The Correlational/Predictive Study The correlational study aimed at identifying from among selected primary-school inputs and pupil personal variables those which covaried with or seemed predictive of achievement in the Age Eleven Plus (11+) Common Entrance Examination to high schools in three parishes in eastern Jamaica. It studied the behavior of these variables in schools and children in urban area as opposed to schools 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 81 and children in the rural areas of the parishes, utilizing for greater differentiation a stratified sample of the ten best and of the ten least performing schools in both areas. As a p re lim in a ry , u tiliz in g the resu lts o f the Common Entrance Examination of January 1980, a correlational analysis was done to deter mine the extent to which scores on the English and mathematics tests covaried with the scores on the test of mental ability. The means and standard deviations were secured and compared with the findings of Manley's 1963 study done on 1959 data. The means of the ten best per forming urban schools were also compared. Next, utilizing the results of the three batteries of the Common Entrance Examination of January 1980 as the dependent variables, and a selection of thirteen school input and pupil personal variables as inde pendent variables, identification was made of those independent variables correlating most highly with the three sets of test scores. Investigation also was made by discriminant analysis of the vari ables which separated the top 33 from the bottom 33 percent of the urban sample and the top 25 from the bottom 25 percent of the rural sample based on individual total scores. These were the percentages which by observa tion of the histograms and initial analyses best differentiated the groups. Selection of Sample Schools The Common Entrance Examination is taken each year by pupils from public primary and private preparatory schools who are not less than ten years nine months of age or more than thirteen years of age on the first day of September that year. In January 1980, a total of 38, 106 pupils islandwide sat the examination with 35,697 ( 12,718 boys and 22,979 girls) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 82 sitting from the public schools and 2,409 from private independent schools. The sample of schools was taken from three eastern parishes of Jamaica—Kingston, St. Andrew, and St. Thomas. Using a classification of " rural” and ” urban” schools developed in the Ministry of Education, Jamaica, in 1976,it was observed that, among the three parishes, there was an acceptable ratio of urban to rural schools. Data in table 18 bear out this fact, showing eighty urban and seventy rural schools (including two special schools) in the three parishes. TABLE 18 NUMBER OF SCHOOLS IN THE PARISHES OF KINGSTON, ST. ANDREW, AND ST. THOMAS CLASSIFIED AS URBAN (CITY/TOWN) OR RURAL/REMOTE BY MINISTRY OF EDUCATION (1976) Number of Rural/ Parish Numcbe/- ° f U rban Total Schools Remote Schools Kingston 27 1 28 St. Andrew 48 31 79 St. Thomas 5 38 43 Total 80 70 150 As school factors contributing to pupil achievement were a prime focus of study, it was decided to stratify the schools sampled based on "measure of success” in the Common Entrance Examination. Success was not measured in terms of numbers of candidates awarded high-school places, since it was established that a few schools screened candidates while others did not. Rather, success was measured by the percentages of awardees to numbers of students in the grades from which Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 83 candidates would be drawn, i.e ., grades four to six in the 1979/80 school year- Utilizing this measure, success was observed not only on performance in the 1980 Common Entrance Examination but also in the Common Entrance Examination of 1979. Appendices D and E show the performance of all schools in the three parishes entering pupils in the Common Entrance Examinations of January 1979 and 1980. Based on this measure, therefore, an initial selection was made of a total of forty schools—ten urban and ten rural schools with the highest percentage of awards to number of students in the relevant grades and ten each of the urban and rural schools with fewest or no awards. Comparison of the schools in appendix F shows that the best performing schools in urban areas differed considerably from the best performing schools in rural areas in 1980 (15 percent-34 percent awards to students in grades 4-6 in urban areas as against 5 percent-10 percent in rural areas). The selection of schools with the fewest awards was made from urban area schools by random selection from schools obtaininq 5 per cent awards and below. In the case of rural areas, a random selection was made from schools obtaining zero awards. A description of the schools follows. Description of Sample Schools The ten best performing urban schools. The ten best schools together with their grades 1-6 enrollments for 1979/80 are shown in table 19. The group includes seven very large schools—Dunrobin, Mico, New Providence (the largest, with 2,383 pupils in two shifts). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 84 TABLE 19 TEN BEST PERFORMING URBAN SCHOOLS (Grades 1-6) Schools Enrollments 1979/80 Parish of Kingston Alpha II - Jessie Rippol Primary 598 St. Aloysius Primary 1,406 St. George's Girls All-Age & Infant 759 Parish of St. Andrew Dunrobin Primary 1,778 Hope Valley Experimental Primary & infant 522 Mico Practising All-Age 917 New Providence Primary 2, 383 Pembroke Hall Primary 1, 495 St. Richard's Primary 1,081 Tarrant Primary 1, 507 Pembroke Hall, St. Aloysius, St. Richard's, and Tarrant. There are three medium-sized schools—Alpha II - Jessie Rippol, Hope Valley Experimental, and St. George's Girls'. These schools may be classified as magnet schools to a greater or lesser degree, being renowned for their successes in the Common Entrance Examination. Some, like St. George's, have a long history in this regard. Alpha II (orJessie Rippol, a school named in honor of a nun), St. Aloysius, and St. Richard's are Roman Catholic schools known for their discipline. Mico Practising, an All-age school taking pupils up to age 15, is Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 85 attached to the Mico Teachers' College. Hope Valley Experimental is classified as a Special school as it maintains special relationship with the U n iversity of West Indies' Polio Rehabilitation C entre, admitting physically handicapped as well as normal children. Mico, St. Aloysius and St. George's are downtown schools. Dunrobin and St. Richard's are located in uptown middle class areas. All these schools draw children from all socioeconomic levels in a varying mix of numbers. Alpha II was omitted from analyses involving canonical correlation and discri minant analysis due to difficulties in securing pupil data from the school. The ten best performing rural schools. The ten best rural schools together with their grades 1-6 enrollments for 1979/80 are listed in table 20. There are four relatively large schools. Lawrence Tavern, the largest school, is followed by Duckenfield, Lyssons, and Trinity- ville. The other six schools are relatively small. Most of these schools are located in very rural areas. At Red Hills and Lawrence Tavern, there is some degree or urbanization. Lyssons is located close to the town of Morant Bay. Ten urban schools—fewest awards. The ten urban schools with fewest awards together with their enrollments are listed in table 21. Two schools— Dupont and Seaward—are very large. Next in descending order of enrollment size are St. Peter Claver, Tavares Gardens, Allman Town, Mountain View, and Balcombe Drive. The last three— Mannings Hill, Gordon Town, and Boys' Town are small schools. The political violence in Kingston during the 1979/80 period affected all schools and Boys' Town very significantly. This school had registered enrollments of 280 in both 1976/77 and 1977/78. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 86 TABLE 20 TEN BEST PERFORMING RURAL SCHOOLS (Grades 1-6) Schools Enrollments 1979/80 Parish of St. Andrew Dallas All-Age 375 Lawrence Tavern Primary 1,080 Red Hills All-Age 404 Parish of St. Thomas Dalvey Primary 247 Duckenfield Primary 742 Golden Grove Primary & Infant 330 Johns Town All-Age 142 Lyssons All-Age 677 Middleton Primary 310 Trinityville Primary 661 All these schools are located in low-income-earning areas of the city, with Mountain View being perhaps on the upper end of the scale. Gordon Town and Mannings Hill are located on the outskirts of the city. Ten rural schools--fewest awards. The rural schools with no awards selected for the study together with their enrollments are shown in table 22. All of these are relatively small schools. Three or four are very remote. After repeated unsuccessful attempts to secure pupil data from Mt. Felix it was decided to omit this school from the analyses involving canonical correlation and discriminant analysis. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 87 TABLE 21 URBAN SCHOOLS WITH FEWEST AWARDS (Grades 1-6) Schools Enrollments 1979/80 Parish of Kinqston Allman Town Primary 800 Boys Town All-Age 79 Parish of St. Andrew Balcombe Drive All-Age 514 Dupont Primary & Infant 1,184 Gordon Town All-Age 256 Mannings Hill All-Age 261 Mountain View Primary 761 St. Peter Claver Primary 942 Seaward All-Age 1, 570 Tavares Gardens All-Age 940 Summary. The sample of the fo rty schools chosen included schools from each of the four types in the primary education system. There were sixteen primary schools, three primary schools with Infant Departments, twenty All-Age schools, and one All-Age school with an Infant Department. In addition, three schools operating double shifts were included. These represented the response of the education system to enrollment pressures. Table 23 shows the distribution of these types of schools among the performance groupings of the sample as well as the overall distribution for the three parishes. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 88 TABLE 22 RURAL SCHOOLS WITH NO AWARDS (Grades 1-6) Schools Enrollments 1979/80 Parish of St. Andrew Bowden Hill All-Age 111 Grove Primary 168 Halls Delight All-Age 234 Mt. James All-Age 297 Westphalia All-Age 104 Parish of St. Thomas Bethesda All-Age 240 Hayfield All-Age 70 Mt. Felix All-Age 89 T hornto n All-Age 159 Whitehorses A ll-A ge 316 The next section discusses the determination of samole size and the number of subjects for each of the two studies. The method of securing the subjects is described as well as the statistical parameters for the sample and the tests. Sampling of Pupils Table 24 shows that of the 150 urban and rural schools in the population, 146 put forward 14,072 candidates in the Common Entrance Examination of 1 980, gaining 3,177 awards. Overall, urban areas secured 23 percent awards over entries while rural schools secured 15 percent Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 89 TABLE 23 DISTRIBUTION OF THE FOUR TYPES OF SCHOOLS IN THE STRATIFIED SAMPLE Performance Primary Primary All-Age All-Age Grouping /Awards with Infant with Infant U rb an Most 7 1 1 1 Urban Fewest 3 1 6 - Rural Most 5 1 4 - Rural Fewest 1 - 9 - Sample Total 16 3 20 1 Total School Types in the Three Parishes 1979/80 59* 8* 69 14 * Includes special schools awards over entries. Four schools put forward no candidates. For certain analyses both urban and rural areas were treated as populations. Utilizing the tables for determining sample size from a given population developed by Krejcie and Morgan ( 1970), a sample of 302 would have been adequate for a signifiance level of a = .05 level in respect of the rural areas. A sample of 319 cards with achievement scores was drawn and complete data were secured in 2 85 cases. The urban population of entrants was 12,673 from which— using the sample table mentioned above—a sample of 373 would have been adequate. A sample of 729 cards with achievement scores was actually drawn and complete data secured on 517 cases. Appendices C and H present details by stratum and by school of the sample drawn. An overall total Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 90 TABLE 24 1980 COMMON ENTRANCE EXAMINATION (Entries and Awards from Urban and Rural Schools) No. of No. of No. of Category Schools Entries Awards Urban schools 79 12,673 2, 968 Rural Schools 67 1,399 209 Total 146 14,072 3,177 of 802 cases was achieved for the multivariate analyses, i.e ., canonical co relation, and discriminant analysis. For the total of nineteen variables originally projected, 190-285 cases (with a ratio of ten-fifteen subjects for every variable) would have been necessary for the stability of the correlation matrices. (See Kendall, 1975, p. 11 for a recommended ratio of ten to one— 10:1.) For the sixteen variables finally utilized, a total of between 160-240 cases was the actual requirement. Obtaining 285 complete cases for the rural areas and 517 for the urban areas—a grand total of 802— ensured the soundness of the multivariate analyses. The minimum effect size set for the Pearson r product-moment correlation test was ± .30. This is considered to be of medium effect. The test for the null hypotheses was set at ci= .05. With these two parameters and a sample of 1,048 subjects the power of the test was .99. For the Analysis of Variance test, the level for the null hypothesis was set at a = .05. The power achieved on the test which involved 466 subjects was .99. Power is defined as the probability of rejecting the null hypothesis when it is in fact false (Welkowitz, Ewen, 6 Cohen, 1962, chapter 4). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 91 The sample was drawn by random selection of score cards of entrants. A table of random numbers as well as manual random selection was used, as cards were filed in two ways. Award cards were filed under forty-six high schools and four comprehensive schools. Non award cards were batched under the schools from which the pupils entered. Data Gathering— Correlational Study The following data were secured from the Common Entrance score cards of 1980 : 1. Mental A bility test scores 2. Mathematics test scores 3. English test scores 4. Sex 5. Age 6. Parent's occupation 7. Num ber of years in school (in most cases) 8. W hether there was previous school experience in P rep arato ry or Basic schools. A sample card is included in appendix I. Where data item number 7, number of years in school, was not secured from the cards, it was necessary to seek it directly from the schools. The schools were approached by letter and followed by visits for the following data items: 9. Number of sessions attended for the 1 978/79 school year, the year preceding the term in which the candidates sat the examination. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 92 10. Grade from which the individual pupil sat the Common Entrance in January 1980. 11. Qualifications of the teacher of the particular grade in 1980. 12. Years of experience as teacher. 13. Whether use was made of Schools'Library Service and whether the school had a lib ra ry . 14. Whether use was made of television in the Common Entrance classes. 15. Qualifications of the principal. 16. Years of experience as principal. The following information was obtained from the Ministry of Education : 17. Pupil/teacher ratios for the school years 1977/78 - 1979/80. 18. Percentage of trained teachers at the school 1977/78 - 1979/80. 19. Classroom accomodation. Obtaining the Quantitative Indices: Criterion Variables 1-4 The test scores were all numbers which together formed the total scores. The total score was used in differentiating the groups for the discriminant analyses. Obtaining the Quantitative Indices: Predictor Variables 1-16 Variables 1 and 2. Sex was coded one (1) for male and two (2) for female. Age presented no difficulty. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 93 Variable 3. Parent's occupation presented difficulty. For example, occupations quoted on the cards such as " housewife" and 11 farmer," being general rather than specific, meant that a range of incomes and education could be involved. After several fruitless attempts to secure Jamaican guidelines to categorizations, the re searcher decided to classify occupations in broad bands related to the 1955 McCuire-White Social Class index (see appendix C) and the occupational categories of the Annual Jamaica Labour Force Survey. These bands were coded: 1. 10— Lowest paid categories—primary education only, assumed. 2. 20--A broad band covering assumed secondary-level education and lower-middle and middle-middle incomes. 3. 30— Professionals and upper-middle, lower upper-class incomes. The list of occupations together with their coding is given in appendix J. Nonspecific occupations were placed in category 20. Variables 4 and 5. Number of years in school presented no difficulty, neither did number of years in Preparatory, Infant, or Basic School. Variables 6 and 7. Number of sessions as well as grade from which the Common Entrance was sat presented no difficulty. Variable 8. Teacher's qualification was coded as follows: Code 1 = Pre-trained Code 2 = Teacher in the In-Service Teacher Education Thrust (ISTET) program Code 3 = Trained Teacher or Intern Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 94 Code 4 = Trained teacher with Vocational Training Code 5 = Trained teacher with Management Training Code 6 = Teacher with the Certificate in Education Code 7 = Teacher with a First Degree Variable 9. Number of years in teaching presented no coding difficulty. Variables 10 and 11. Use of Library and Television was each coded one (1) for No, two (2) for Yes. Variable 12. Pupil-teacher ratios for grades 1-6 were secured for three years and averaged; the average was used. Variable 13. Percentage of trained teachers was secured, averaged over three years,and the average was used. Variables 14 and 15. Principal1 s qualification was coded as for teacher's qualification: principal's years of service as principal presented no coding difficulty. Variable 16. Degree of overcrowding was coded on the following basis: data on classroom size collected by the Ministry of Education in 1978 were used to derive the desirable enrollment level based on 10 square feet per pupil (cf. 9 square feet per pupil in Reid's (1964) study). The actual enrollment was divided by this number and the resulting quotient was the number used. Modifications were made in a few cases where these were obviously needed. Qualitative Data Finally, the collection of data from the schools included a request for the principal's comments on the factors which negatively or positively influenced performance in the Common Entrance Examination. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 95 Categorization of the Predictor Variables The predictor variables may be divided into pupil variables and school variables. The pupil variables were: 1. Sex 2. Age 3. Parent1 s occupation 4. Preparatory school experience 5. Number of years at the particular school 5 . Number of sessions attended 7. Grade from which the Common Entrance was sat. The school variables were: 1. The grade teacher's qualification 2. The grade teacher1 s experience 3. Use of lib rary 4. Use of television 5. Teacher/pupil ratio for the school, 1977-80 6. Percentage trained teachers, 1977-80 7. Principal1 s qualification 8. Number of years as principal 9. Degree of overcrowding Rationale for Selection of the Variables Chapter 2 presented literature on the variables utilized in the study. It is useful to restate that the variables examined were school factors directly manipulable by those responsible for the provision of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 96 education. No attem pt was made to p ro file, measure, and correlate variables pertaining to the students' homes and communities. In respect to school variables, it was not possible, though desirable, to measure certain other variables identified in the literature as having a bearing on pupil achievement. These included pupil motivation and supply of text books. Environmental Factors to Be Considered Mention should be made of the very unstable environment existing in urban Kingston and St. Andrew, and to a lesser extent throughout Jamaica in 1979-80. Political violence, with regular outbreaks of gunfire in low income areas, led to reduced enrollments in many schools and must have had its effects on pupil performance. Strength, Relevance, and Degree of Differentiation of Certain Variables of the Study Data collection proved to be very difficult. Numerous visits had to be made to certain schools. Some records had been misplaced or were lost, due in one case to a break-in of the principal's office. In the larger schools, teachers were not passing in registers to the principal's office at the end of the school year as required. Smaller and rural schools presented less difficulty in this matter. Principals pointed out that records were easier to maintain when the Ministry of Education had required them to prepare annual Class Lists. A sample list is shown in appendix K. This class list summarized a good deal of the required data which otherwise could only be obtained by reference to the Attendance Registers and other registers in the school. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 97 Finally, as mentioned earlier, complete pupil and school data were possible on only thirty-eight schools. The hypotheses requiring complete data were therefore tested on thirty-eight schools and a total 802 pupils. Number of years in preparatory school also presented uncertain ties. It appeared that these data were not consistently entered on the c a rd s . Information on Library and Television was not sufficiently differentiated for the schools. Measures were not secured for the number of books per pupil or numbers of books, other than readers, which pupils were required to read. Data on hours of television exposure were not secured, so these data became dichotomous—whether or not TV was used with the Common Entrance class. It was decided, therefore, to omit these from the analyses. Comment is, however, offered on the variables in view of their perceived importance to and potential for the learning process. The data would be useful in the final conclusions and recommendations. Mention should be made of the extra coaching which CEE pupils customarily receive within and outside their schools. School principals were questioned about this. After school and holiday classes were held in most schools for a fee. Two schools had separate day-time classes. The Exploratory Study The population for this area of the study comprised originally approximately 700-750 teachers drawn from the forty schools sampled. Principals felt that those teachers (approximately 300) working in the Common Entrance grades were the ones who would be most qualified to Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 98 respond. Opinionnaires were circulated to 300 teachers and question naires were received from 177 teachers. This response gave the sample validity at a significance level of a = .0 5 . The letter (sample in appendix L) sought the teachers' ratings of a total of sixteen indicator items broken down into thirty-one responses specifying (1) school level, (2) parish level, (3) islandwide level. These were to be rated on a continuum of 1-5, 1 beinq most important. Table 25 classifies and lists these sixteen indicator items. In addition, teachers were asked to add any indicators which they considered to be important and rate these indicators also on a continuum of 1-5, 1 being most important. More than one attempt had to be made to secure return of the questionnaires. It became apparent that returns were best received when the researcher circulated the questionnaire and waited for the completed responses. Stamped self- addressed envelopes did not help very much, though small tokens such as book marks did help. Hypotheses and Statistical Analyses Used Fourteen hypotheses were originally proposed but thirteen were finally tested. Three hypotheses—hypotheses 1(a)-1(c)—were tested utilizing the Pearson product-moment correlation test; foui hypotheses 2(a)-2(d)—by the analyses of variance test; two, by canonical correla tion and four, by the discriminant analysis tests. The fourteenth test involving discriminant analysis became unnecessary, when no significant difference was found among the school means of the total test battery of the 1980 CEE for the ten best performing urban schools. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 99 TABLE 25 SIXTEEN INDICATORS PRESENTED FOR RATING TO THE SAMPLE OF TEACHERS Indicators Subsets Systems Accessibility 1. Ratio of enrollment to school-age population 2. Degree of overcrowding by school 3. Percentage of overcrowded schools Systems Inputs 4. Teacher/pupil ratio by trained teachers 5. Overall teacher/pupil ratio 6. No. of library books per pupil 7. No. of hours of educational tele vision Systems Process 8. Percentage of teacher turnover 9. Percentage of pupils attending 90 percent of school sessions 10. Efficient principal Environmental 11. Index of malnutrition 12. Percentage of unstable families Systems Outputs 13. Crade-by-grade average achieve ment in English 14. Crade-by-grade average achieve ment in mathematics 15. Percentage of students achieving below 50 percent in English and in mathematics 16. Percentage of students passing Common Entrance Examination Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100 The hypotheses stated in the null form are as follows: Hypotheses 1(a)-1(c) Hypothesis 1(a) reads: There is not a significant correlation between the scores of the intelligence and the English tests of the 1980 CEE, fo r th e total sample o f 1,048 p u p ils . Hypothesis 1(b) reads: There is not a significant correlation between the scores of the intelligence and the mathematics tests of the 1980 CEE,for the total sample of 1,048 pupils. Hypothesis 1(c) reads: There is not a significant correlation between the scores of the English and mathematics tests of the 1980 CEE, for the total sample of 1,048 pupils. The significance levels for the above three hypotheses are Y = ± . 30 and a = .05. Hypotheses 2(a)-2(d) Hypothesis 2(a) reads: There is not a significant difference among the school means of the total scores of the 1980 CEE test battery, for the ten best performing urban schools. The statistical test is analysis of variance and the significance level a = .05. Hypothesis 2(b) reads: There is not a significant difference among the school means of the mathematics test scores of the 1980 CEE, for the ten best performing urban schools. The statistical test is analysis of variance and the significance level a= .05. Hypothesis 2(c) reads: There is not a significant difference among the school means of the English test scores of the 1980 CEE, for the ten best performing urban schools. The statistical test is analysis of variance and the significance level a = .05. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 101 Hypothesis 2(d) reads: There is not a significant difference among the school means of the intelligence test scores of the 1980 CEE, for the ten best performing urban schools. The statistical test is analysis of variance and the significance level is cl = .05. Hypothesis 3 There is not a statistically significant canonical correlation between the linear combination of thirteen predictor variables—that is, thirteen pupils' personal and school variables—and the linear combina tion of the criterion variables of intelligence, arithmetic, and English scores of the Common Entrance Examination 1980. The thirteen predictor variables were: 1. Pupil1 s sex 2. Pupil1 s age 3. Parent1 s occupation expressed as an index of socioeconomic status 4. Number of years at the particular school 5. Number of sessions attended in 1978/79 6. Crade from which the 1980 Common Entrance Examination was sat 7. Qualifications of the grade teacher 8. Length of service of the grade teacher 9. Overall teacher/pupil ratio of the school, averaged over the three-year period— 1977/78 to 1979/80 10. The average percentage of trained teachers at the school — 1977/78 to 1979/80 11. Principal's qualifications 12. Principal's years of service as principal 13. Degree of overcrowding Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 102 Hypothesis 4 There is not a statistically significant canonical correlation between the linear combination of twelve predictor variables— that is twelve students' personal and school variables—and the linear combination of the criterion variables of intelligence, arithmetic, and English scores of the Common Entrance Examinations of 1980. The predictor variables are those listed in Hypothesis 3 with the exclusion of parent's occupation which is expressed as an index of socio economic status. The significance level is a = .05. Hypothesis 5 There is no linear combination of input variables which significantly separates the top 33 percent from the bottom 33 percent of pupils sampled in the urban schools. The statistical test is discrimi nant analysis and the significance level is a = .05. Twelve input variables are utilized. These are: 1 . Pupil's age » 2. Pupil's sex 3. Number of years at the particular school 4. Number of sessions attended in 1978/79 5. Crade from which the 1980 Common Entrance Examination was sat 6. Qualification of the grade teacher 7. Length of service of the grade teacher 8. Overall teacher/pupil ratio of the school averaged over the three-year period 1977/78 - 1979/80 9. The average percentage of trained teachers at the school 1977/78 - 1979/80 10. Principal's qualifications Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 103 11. Years of experience as principal 12. Degree of overcrowding Hypothesis 6 There is no linear combination of input variables which significantly separates the top 33 percent from the bottom 33 percent of the pupils sampled in urban schools. Thirteen variables are used, i.e ., the variables in Hypothesis 5, with socioeconomic status added. The statistical test is discriminant analysis and the significance level is a = .0 5 . Hypothesis 7 There is no linear combination of input variables which significantly separates the top 25 percent from the bottom 25 percent of pupils sampled from the rural schools. The twelve input variables are those listed for Hypothesis 5. The statistical test is discriminant analysis and the significance level is ct = .05. Hypothesis 8 There is no linear combination of input variables which significantly separates the top 25 percent from the bottom 25 percent of pupils sampled from the rural schools. Thirteen input variables are used, that is, those variables listed in Hypothesis 5, with socio economic status added. The statistical test is discriminant analysis and the significance level is a = .05. Analytical Methodology The statistical methods used were the most appropriate for examining the relationships of the variables under study. These were: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 104 1. Pearson Product-Moment Correlation Coefficient 2. Analysis of Variance 3. Canonical Correlation 4. Discriminant Analysis 5. Categorical Scaling Pearson Product-Moment Coefficient The Pearson product-moment correlation technique produces a single descriptor statistic which quantitatively describes the degree to which change in the values of one variable parallels a degree of change in the other, indicating also whether the companion change is in a similar (positive) or opposite (negative) direction, or whether there is no significant co-variance. This method was used to determine the covarying relationships among the scores of the three batteries of the Common Entrance Test. Analysis of Variance The analysis-of-variance technique provides a statistical answer to the question of whether there is a significant difference in the measures on the dependent variable when two or more groups are involved. In the case of two groups, the t-statistic is usually used. The analysis-of-variance technique is commonly employed when three or more groups are involved and has its greatest usefulness when two or more independent variables are studied (Runyon & Haber, 1977, p. 288). Utilizing the F-ratio, this technique determines whether the variance between the groups is more significant than the variance within groups. It answers the question of whether or not there is an indication of an Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 105 overall effect of the experimental treatments before specific hypotheses are investigated. In this study, analysis of variance was used to determine whether there were significant differences in school means of the ten best performing urban schools on the three batteries of the test and the total test scores of the 1980 Common Entrance Examination. Canonical Correlation Canonical correlation is described as the generalization of multiple-regression analysis to any number of dependent variables. Canonical correlation utilizes K independent or predictor variables and M dependent or criterion variables. Through the least squares analysis two linear composites are formed, one for the independent (predictor) variables X and one for the dependent (criterion) variables Y. The correlation between these two composites is the canonical correlation Rc. The square of the canonical correlation R^ is an estimate of the variance shared by the two composites. The canonical coefficient is the maximum corre lation possible between the two sets of variables. (Kerlinqer & P ed h azu r, 1973, p p . 3*11-342) In illustrating its application, the above cited authors state that more than one source of common variance can be identified and analyzed. The first and largest source of variance is extracted first, then the next largest source of variance left in the data is identified. The respective coefficients are the indices of the relationship (pp. 344-345). The appropriateness of the canonical correlational technique to this analysis may best be appreciated by reference to the critique of the Coleman R eport by Elazar j . Pedhazur (1982) . He points out that multiple-regression analysis, the methodology used in the Coleman study, is more validly applied to experimental rather than non- experimental research. In experimental research the independent Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 106 variables are manipulated and the dependent variables observed; there fore causal assumptions may more validly be made. In non-experimental research such as the Coleman study, correlation, even though substan tial, does not establish direct relationships because it is difficult: 1. to untangle the effects of each variable 2. to know whether variables are serving as proxies for the true variables (p. 175). The incremental partitioning of variance of multiple regression rests on a crucial question: In what order are the variables entered into the equation? Pedhazur (1982) suggests that because one inde pendent variable precedes another independent variable in time, this fact is not sufficient justification for regarding it as causal and thus entering it first. In commenting on the statistical methodology applicable to non- experimental (or experimental) multi-dimensional research, Pedhazur ( 1 982) suggests that studying it piecemeal does not hold promise of understanding it. Multi-variable phenomena require analytic methods commensurate with them (p. 686). The development of the computer allows for the utilization of procedures such as canonical analysis, as developed by Hotelling in 1 936 (Pedhazur, 1982, p. 720). This technique tests the correlations of two sets of variables simultaneously rather than sequentially. Canonical correlation was therefore consid ered most appropriate to this study. Discriminant Analysis Having established through canonical correlation that there was a significant combination of predictor variables covarying with the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 107 criterion variables, discriminant analysis was used to isolate those vari ables which separated the top segment of pupils from the bottom segment in urban and in rural areas. The discriminant analysis technique, developed by R. A. Fisher in 1936 (Pedhazur, 1982), enables the weighting of individual scores so as to lead to maximum discrimination between the groups and identification of the dimensions along which the groups differ (see Pedhazur, 1982, pp. 692-717). Analysis of Teacher Responses by the Categorical Scalings Model The categorical scaling model was appropriate for this situation where stimuli were presented to teachers for ranking on a continuum of importance—that is, 1-5, with 1 being most important. Torgerson (1955, p. 207) identifies three modes for eliciting judgments: 1. Replication over trials within a single subject. 2. Replication over individuals—each stimulus being judged once by each subject. 3. Mixed models involving replication over both individuals and trials, each stimulus being judged several times by each subject. The second method was the one utilized in this study. Torgerson ( 1955) also notes that a great number of analytical procedures dealing specifically with categorical judgments have been devised (p. 214). Categorical scaling commences with a matrix showing the frequency with which each stimulus is rated in each category. From this matrix cumulative frequencies and proportions are derived. These are converted into normal deviates (z scores) . A matrix of intercorrelations among Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 108 these deviates is subjected to matrix algebra procedures, the first eigenvector giving positions of the stimuli on an arbitrary interval scale. In the analysis, the scale was modified by addition of a constant so that the lowest scale value was zero. Data Analysis All data collected were recorded on the xerox sigma 6 computer at Andrews University, Berrien Springs, Michigan and carefully rechecked. Analyses utilized the packaged BMDP statistical programs. The Family of Indicators The survey identified ratings of the indicators presented to the teachers of the schools which were sampled. Other indicators were suggested by them as well. The opinions of principals on factors contributing to success in the Common Entrance Examination were also secured. The correlational studies as well as the discriminant analyses were performed to identify input variables of greatest importance. The significant variables identified were gauged against the findings of the literature review of chapter 2. The family of indicators was formulated against this theoretical and empirical background. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER IV FINDINGS OF THE STUDY This chapter presents the results of analyses of data gathered in a correlational study of pupil performance in the Common Entrance Examination (CEE) of 1980. The aim of this investigation, the first and major segment of a two-part study, was to examine the extent to which certain school and personal variables were predictive of or co-varied with test scores of the Common Entrance Examination of January 1980. The second segment involved the collection and analysis of data on an opinionnaire circulated to a sample of teachers in the schools selected for the first part of the study. Results— Correlational/Predictive S tu d y Thirteen hypotheses were tested on the data gathered for the correlational/predictive segment of this study. Hypotheses 1(a)-1(c) tested the degree of correlation among the scores for all entrants on the three batteries of the 1980 Common Entrance Examination. The Pearson product-moment correlational technique was utilized for these hypotheses. Hypotheses 2(a)-2(d) established by analysis of variance whether there was a significant difference among the school means of the ten best performing urban schools. Hypotheses 3 and 4 established by canonical analysis whether there was a linear combination of input variables which 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 110 significantly correlated with the scores on the three batteries of the 1980 CEE for all entrants. Hypotheses 5-8 utilized discriminant analysis to isolate those factors which most significantly separated the top 33 percent from the bottom 33 percent of urban entrants, and the top 25 percent from the bottom 25 percent of rural entrants. Individual total CEE scores were used in distinguishing the groups. The respective percentages were selected on the basis of breaks appearing in the histograms of the two sets of scores (urban and rural) and preliminary analyses of these. Hypotheses 1(a)-1(c) Hypothesis 1(a) reads: There is not a significant correlation between the scores of the intelligence and the English tests of the 1980 CEE, for the total sample of 1,048 pupils. Hypothesis 1(b) reads: There is not a significant correlation between the scores of the intelligence and the mathematics tests, of the 1980 CEE for the total sample of 1,048 pupils. Hypothesis 1(c) reads: There is not a significant correlation between the scores of the English and the mathematics tests of the 1980 CEE, for the total sample of 1, 048 pupils. The significance tests for the above three hypotheses were Y = ± . 30 and a= .05. Table 26 displays the obtained correlations all above '(= +.30 and significant at p < .001 level. The three hypotheses were therefore rejected. The highest correlation was between the English and intelligence scores (.85). There was a slightly lower correlation between the intelligence and mathematics scores (.82) and the lowest correlation was between English and mathematics (.78). Table 27 shows the means and standard deviations. When compared with Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 111 TABLE 26 CORRELATIONAL ANALYSIS FOR FORTY SCHOOLS-- OBTAINED CORRELATIONS Intelligence Mathematics English Intelligence 1.00 .82 .85 Mathematics .82 1.00 .78 English .85 .78 1.00 p < .001 T A B L E 27 CORRELATIONAL ANALYSIS FOR FORTY SCHOOLS— MEANS AND STANDARD DEVIATIONS Variable Cases Mean Standard Deviation Intelligence 1048 102.89 18.06 Mathematics 1048 86.63 21.00 English 1048 87.23 18.97 the analyses done by Manley (1963) on 1959 data, it is observed that the arithmetic quotient shows little change (less by 1.2 points rounded). The English and more so the intelligence scores have risen (greater by 3.2 and 17.6 points rounded, respectively). See table 28. Hypotheses 2(a)-2(d) Hypothesis 2(a) reads: There is not a significant difference among the school means of the total scores of the 1980 CEE test battery, for the ten Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 112 TABLE 28 COMPARISON OF QUOTIENTS IN MANLEY'S STUDY— 1959 DATA; PRESENT STUDY — 1980 DATA No. of Mean Data Year Cases Mathematics Intelligence English (arithmetic) 1 959 1406* 85. 3 87.8 ( Prim ary Schools only) ro ro o 00 00 00 1 980 1048 102.89 86.63 CO * On a total sample of 1730 which included Preparatory Schools, all means were said to be between 85 and 90 with standard deviations about 10 (Manley, 1963, p. 6). best performing urban schools. The significance level was a= .05. Table 29 shows the means on the total test scores for the ten best performing urban schools. Table 30 shows the analysis of variance for the test of the hypothesis. A significant F probability was not achieved (p = .58). The h yp o th esis was th e re fo re re ta in e d . Hypothesis 2(b) reads: There is not a significant difference among the school means of the mathematics test scores of the 1980 CEE, for the ten best performing urban schools. The significance level was a = .05. Table 31 shows the means on the mathematics scores for the ten best performing urban schools. Table 32 shows the analysis of variance for the test of the hypothesis. A significant F probability at the level set was not achieved (p =.07). The hypothesis was therefore retained. Hypothesis 2(c) reads: There is not a significant difference among the school means of the English test scores of the 1980 CEE, for the ten best performing urban schools. Table 33 shows the means on the English Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 113 TABLE 29 MEANS OF TEN BEST PERFORMING URBAN SCHOOLS— TOTAL BATTERY OF 1980 CEE SCORES Croup Name Count Mean S tandard No. Total Deviation 01 St. Richard1 s 44 300.71 49. 84 02 Dun robin 53 300.04 54. 16 03 Pembroke Hall 47 300.85 53.01 04 St. George1 s Girls 44 293.71 4 8.82 05 Alpha II 42 298.10 49. 71 06 T a rra n t 45 295.07 4 4.28 07 St. Aloysius 44 308.21 47. 75 08 Hope Valley 35 291.71 45.71 09 New Providence 67 299.05 50.49 10 Mico 45 281.87 6 1 .3 2 Total 466 297.18 50.82 T A B LE 30 ANALYSIS OF VARIANCE ON SCHOOL MEANS OF TOTAL BATTERY OF CEE SCORES—TEN BEST PERFORMING URBAN SCHOOLS Source d f SS MS F P Between Croups 9 19546.09 2171.79 .8 4 .58 W ithin Croups 456 1181422.00 2590.84 Total 465 1200968. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 114 TABLE 31 MEANS OF TEN BEST PERFORMING URBAN SCHOOLS— MATHEMATICS TEST OF 1980 CEE SCORES Group Name Count Mean Standard No. Deviation 01 St. Richard's 44 9 2 .2 3 21.28 02 Dunrobin 53 91.81 19. 90 03 Pembroke Hall 47 9 8 .1 7 21.66 04 St. George's Girls 44 8 9 .4 5 19.79 05 Alpha II 42 101.71 19.42 06 T arran t 45 95. 29 18.23 07 St. Aloysius 44 9 0 .3 4 19.88 08 Hope Valley 35 9 2 .7 4 19.86 09 New Providence 67 95.04 21.33 10 Mico 45 88.44 22.61 Total 466 9 3 .5 7 20.65 T A B L E 32 ANALYSIS OF VARIANCE ON SCHOOL MEANS OF MATHEMATICS TEST — TEN BEST PERFORMING URBAN SCHOOLS S ource d f SSMS F P Between Groups 9 671 2 .7 745.86 1.776 .0705 Within Groups 456 191493.6 419.94 T otal 465 198206.3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 115 scores for the ten best performing urban schools. Table 34 shows the analysis of variance for the test of the hypothesis. A significant F probability was not achieved (p = .35). The hypothesis was therefore re ta in e d . Hypothesis 2(d) reads: There is not a significant difference among the school means of the intelligence test scores of the 1980 CEE, for the ten best performing urban schools. Table 35 shows the means on the intelligence scores and table 36 the analysis of variance for the test of the hypothesis. This was significant with an F probability of p = .0003. The hypothesis was therefore rejected. Because of the significant F ratio the Newman Keuis test (Winer, 1962, pp. 77-89 ) was used to compare the means for each pair of schools. Table 37 presents this comparison. The schools are ordered according to size of the mean. Any entry in the table shows the difference between the mean of the school at the column head and the mean of the school at the row head. For example, 3.54 = 105. 52 - 101. 98. An asterisk indicates a significant difference at the ct= .05 level. It was proposed to isolate by discriminant analysis those factors which separated the students in the school with the highest mean on total scores from the school with the lowest mean. As no significant difference among the school means on the total CEE test scores was found this analysis was abandoned. Hypothesis 3 The third hypothesis centered around whether out of thirteen predictor variables there was a combination of variables covarying significantly with achievement scores on the intelligence, mathematics. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 11G TABLE 33 MEANS OF TEN BEST PERFORMING URBAN SCHOOLS— ENGLISH TEST OF 1980 CEE SCORES C ro u p Name C ount Mean S tan d ard No. Deviation 01 St. Richard's 44 97.11 18.51 02 Dun rob in 53 96.77 20. 55 03 Pembroke Hall 47 95. 51 18.15 04 St. George's Girls 44 94. 59 18.56 05 Alpha II 42 91.33 18.22 06 T a rra n t 45 93.62 15.78 07 St. Aloysius 44 99.36 18.41 08 Hope Valley 35 91.86 16.33 09 New Providence 67 93.28 17.77 10 Mico 45 89.84 21 .49 Total 466 94.38 18.52 T A B L E 34 ANALYSIS OF VARIANCE ON SCHOOL MEANS OF ENGLISH TEST —TEN BEST PERFORMING URBAN SCHOOLS Source d f SS MS F P Between Groups 9 3432.1 381.34 1.11 .35 W ithin Groups 456 156027.7 342.16 Total 465 159459.8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 117 TABLE 35 MEANS OF TEN BEST PERFORMING URBAN SCHOOLS— INTELLIGENCE TEST OF 1980 CEE SCORES Croup Name Count Mean Standard No. Deviation 01 St. Richard1 s 44 111.36 13.85 02 Dun robin 53 111.45 18.30 03 Pembroke Hall 47 107.19 16.18 04 St. George1 s Cirls 44 109.89 15.22 05 Alpha II 42 105.52 15.72 06 T a rra n t 45 106.16 12. 96 07 St. Aloysius 44 118.25 12.43 08 Hope Valley 35 107.11 13.67 09 New Providence 67 110.70 14.94 10 Mico 45 101.98 21.45 Total 466 109.09 16.14 T A B L E 36 ANALYSIS OF VARIANCE ON SCHOOL MEANS OF INTELLIGENCE TEST —TEN BEST PERFORMING URBAN SCHOOLS Source d f SS MS F P Between Groups 9 7920.4 880.05 3.55 .0003 Within Groups 456 113142.1 248.12 Total 465 121062.5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 118 - 6.80* 7.55* 8.36* 12.09* 16.27* 12.73* - .09 6.89* .75 9.97* 5.93* 1.56 5.29* 9.39* 11.19* 9.26* 11.06* - 5.89* 5.20* 1.97 9.25* - .66 5. 18* 5. 8.72* 9.38* 3.59 3.51 9.17* 9. 9. 59* 7.91* 3.73" 2.78 2.70 .08 - 1.03 . 95 . 1.59 1.67 9.37* 5.13* 5.21* Hope Hope Pembroke St.Ceorge's New St. St. T A B L E 37 - .69 3. 59*3. 9.18* SCHOOLS— INTELLIGENCE TEST -- - Mico Mico Alpha II Tarrant Valley Hall Clrls Providence Richard's Dunrobln Aloysius 105. 52 105. 101.98 106.16 111.15 118.25 111.36 110.70 107.19 B9 109. - .81 107.11 - Means Means 101.98 105.52 16 106. 1 107.1 107.19 109.89 110.70 36 . 1 11 111.<15 118.25 Newman Keuls a posteriori test MATRIX OF DIFFERENCES OF MEANS (ROUNDED)—TEN BEST PERFORMING URBAN * = means which are significantly different based on the Means p < .05 Dunrobln St. Aloysius St. Richard's New New Providence Tarrant Hope Hope Valley Pembroke Hall St. George's Clrls Alpha II Mico Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 119 and English tests of the CEE 1980. The predictor variables were: 1. Sex 2 . Age 3. Parent's occupation expressed as an index of socioeconomic status *». Number of years at the particular school 5. Number of sessions attended 1978/79 6. Crade from which the Common Entrance was sat in 1980 7. The grade teacher's qualifications 8. The grade teacher's years of experience 9. Overall average teacher-pupil ratio for the school for the three years 1977/78 - 1979/80 10. Three-year average of the percentage of trained teachers at the school 1977/78 - 1979/80 11. The principal's qualifications 12. The principal's years of experience as principal 1 3. The factor of overcrowding The null hypothesis reads: There is not a statistically significant canonical correlation between the linear combination of thirteen predictor variables—that is, thirteen pupil personal and school variables—and the linear combination of the criterion variables of the intelligence, mathe matics, and English scores of the 1980 Common Entrance Examination. The null hypothesis was rejected at a = .05 level. The results of the analysis are shown in table 38 and discussed below. First Canonical Function— Hypothesis 3. Since ±.30 was to be taken as the acceptable level of correlation, the linear combination on the first canonical function is primarily discussed. The first function showed a significant canonical correlation coefficient of .575 at a probability level Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 120 TABLE 38 CANONICAL CORRELATION AND STANDARDIZED COEFFICIENTS FOR EACH CANONICAL FUNCTION OF INTELLIGENCE, MATHEMATICS, AND ENCLISH TEST SCORES WITH THIRTEEN PERSONAL AND SCHOOL INPUT VARIABLES (N = 802) Function 1 2 3 Canonical Correlation .575 .217 .157 C h i-S q u are 375.16 58.04 19.75 Degrees of Freedom 39 24 11 Probability < .000005 < .00012 < .04892 Test Scores (Set 1) Rank Rank Rank Intelligence .391 2 -1.701 2 -1 .2 8 4 2 Mathematics .122 3 1.708 1 - .6 8 9 3 English .541 1 .139 3 1 .8 7 9 1 Personal + School Input Variables (Set II) Sex .103 11 - .1 7 7 8 .539 2 Age .027 13 -.2 4 1 6 -.2 2 6 6 Parent's Socioeconomic Status (SES) . 346 3 .373 3 .243 4 Years at School .046 12 .497 2 .125 9 Sessions Attended 78/79 . 352 2 -.1 0 9 10 -.7 3 3 1 Crade CEE Sat 1980 . 334 5 .068 13 .1 1 5 10 Teacher's Qualifications . 221 7 . 234 7 . 228 5 Teach. Years Experience .248 6 .090 11 -.1 0 7 11 Avg. Teach./Pupil Ratio, 77-80 .115 9 .265 5 .1 5 5 7 Avg. Percentage Trained Teach., 77-80 .371 1 -.1 5 2 9 .154 8 Principal1 s Qualifications .1 38 8 -.3 2 7 4 .074 12 Principal's Years Experience - .3 4 5 4 .605 1 - .3 4 3 3 Overcrowding - .1 1 2 10 .078 12 -.0 1 8 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 121 of p < .000005 and explained 33 percent of the variance between the two sets of variables. The important criterion variables are English (.541) and intelligence (.391) correlating with the predictor variables of (1) percentage of trained teachers (.371), (2) number of school sessions attended (.352), (3) higher socioeconomic status (.346), (4) principal's administrative experience (-.345), (5) grade level in 1979/80 (.334), (6) number of years the grade teacher taught (.248), and (7) teacher qualification (.221). This suggests that the pupils who performed highest on the intelligence and English tests were: 1. from higher socioeconomic status (.346) 2. had higher rates of attendance (.352) 3. were at higher grade levels— in the 1979/80 school year (.334), and attended schools with: 1. higher proportion of trained teachers (.371) 2. principal with less administrative experience (-. 345) 3. more experienced grade teacher (.248) 4. more qualified grade teacher (.221). Second Canonical Function— Hypothesis 3. The second function had a canonical correlation coefficient of .217. Mathematics (1.708) and intelligence (-1.701) from the first variable set ranked first and second. The variables which were significantly linearly related to mathematics and intelligence were (1) the principal's number of years service as principal (.605), (2) number of years the pupil spent at the school (.497), ( 3) higher socioeconomic status (. 373), (4) principal's qualifications (-.327), (5) averaged teacher/pupil ratio (.265), (6) age (-.241), and (7) teacher's qualifications (.234). This suggests in a weaker correlation that pupils who did well on the arithmethic but not as well on the intelligence tests were Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 122 younger pupils, from higher socioeconomic levels, who had a longer exposure to school, under more experienced but less qualified principals who had better trained teachers, but who were not in schools with necessarily low pupil/ teacher ratios. The explained variance is so small, however (5 percent), that much emphasis should not be placed on these findings. Third Canonical Function. The third canonical function is significant with probability < .05. However, as only 2 percent of the variance was explained, further discussion is not merited. Hypothesis 4 Since certain limitations were perceived in the categorizations on the socioeconomic variable, the researcher decided to examine the relationships between the criterion and predictor variables with socio economic status excluded from the predictor variables. The null hypothesis reads: There is not a statistically significant canonical correlation between the linear combination of twelve predictor variables—that is twelve students' personal and school variables— and the linear combination of the criterion variables of intelligence, mathematics, and English scores of the 1980 Common Entrance Examinations. The null hypothesis was rejected at a = .05 level. The results of the analysis are shown in table 39. First Canonical Function—Hypothesis 4. The results show two, rather than three, significant canonical functions (see table 39). The first function shows a significant canonical correlation coefficient of . 543, signficant at a probability p < .000005 with 29 percent of the variance explained. The significant criterion variables are intelligence (. 522) and English (.485) with predictor variables in the following order: (1) averaged percentage of trained teachers (.417), (2) number of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 123 TABLE 39 CANONICAL CORRELATION AND STANDARDIZED COEFFICIENTS FOR EACH CANONICAL FUNCTION OF INTELLIGENCE, MATHEMATICS AND ENCLISH TEST SCORES WITH TWELVE PERSONAL AND SCHOOL INPUT VARIABLES (N = 802) Function 1 2 3 Canonical Correlation . 543 .202 .1 50 Chi Square 327.49 51.08 18.03 Degrees of Freedom 36 22 .10 Probability <.000005 <.00042 < .05438 Test Scores (Set 1) Rank Rank N.S. Intelligence .522 1 -1 .4 1 7 2 -1 .5 5 5 Mathematics .039 3 1.808 1 -.366 E nglish .485 2 -.1 7 6 3 1.891 Personal + School In p u t V ariab les (S e t ID Sex .127 9 -.226 7 .542 Age .025 11 -.2 3 1 6 -.3 0 4 Y e a rs at School .009 12 .482 2 .234 Sessions Attended 78/79 .408 2 .073 11.5 - .7 0 4 G rad e CEE Sat 79/80 .383 4 .144 8 . 217 Teacher's Qualifications .216 6 .236 5 .314 Teach. Years Experience .257 5 .143 9 -.0 5 7 Avg. Teach./Pupil Ratio, 79- 80 .157 8 .325 3 .291 Avg. Percentage Trained Teachers, 77-80 .417 1 -.100 10 .195 Principal's Qualifications .209 7 -.2 7 0 4 .072 Principal's Years Experience - .4 0 4 3 .603 1 -.2 7 2 Overcrowding -.1 1 6 10 .073 11.5 -.0 0 8 Not Significant Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 124 sessions attended in 1978/79 (.408), (3) principal's experience in administration (-.404), (4) higher grade in 1979/80 (.383), (5) number of years experience of the grade teacher (. 257), and (6) grade teacher's qualifications (.216). These findings correspond with the findings of hypothesis 3 which included socioeconomic status among the predictor variables. Second Canonical Function— Hypothesis 4. The second function shows a canonical correlation coefficient of .202 and is significant at P< -0004 level. The highest correlations on the first set of variables are mathematics (1.808) and intelligence (-1.417). The highest weightings on the second set are (1) principal's years of experience as principal (.603), (2) number of years pupil spent at the school (.482), (3) average teacher/ pupil ratio (.325), (4) principal's qualifications (-.270), (5) teacher's qualifications (.236), and (6) age (-.231). The explained variance was only 4 percent; hence, further emphasis will not be placed on these findings. The third function was not significant at a= .05 level and is therefore not discussed. Hypothesis 5 Stated in the null form, hypothesis 5 reads: There is not a linear combination of input variables which significantly separates the top 33 percent from the bottom 33 percent of pupils sampled in urban schools based on the individual totals of the 1980 CEE scores. The twelve input variables of hypothesis 4 were utilized in this analysis. Table 40 displays the results of the analysis. The approximate Chi Square was 126.5466, the degrees of freedom 12, and probability level p < .00005. The hypothesis was rejected. The group means on the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 125 TABLE 40 DISCRIMINANT FUNCT IONS— UPPER 33 PERCENT COMPARED WITH LOWER 33 PERCENT OF URBAN STUDENTS— TWELVE VARIABLES Discriminant Variable No. Name Weights Rank 1 Sex 1.0161 11 2 Age .5757 12 3 Years at school 2.0535 10 4 Sessions attended 1978/79 4.2258 9 5 Grade CEE sat 1979/80 7.5358 1 6 Teacher's qualifications 5.6805 4 7 Teacher’s experience 5.5157 5 8 Teacher/pupil ratios 5.7581 3 9 Percentage of trained teachers 4.6942 6 10 Principal's qualifications 4.6018 7 11 Principal's years of experience -6. 5189 2 12 Overcrowding -4.3689 8 Croups Means (D.F.)* Chi Square (approximate) : 126.5466 Upper 11 .295 Degrees of freedom : 12 Lower 10*272 P = < .00005 * D .F . = Discriminant Function Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 126 discriminant function were 11.295 for the upper 33 percent and 10.272 for the lower 33 percent of entrants. The standardized weights enable comparison to be made between the relative importance of the variables in separating th' groups. Examination of the table shows that the two rr. jst outstanding factors separating the groups were (1) the higher grade levels attained (7.5358) and (2) principals with fewer years of administrative experience (-6.5189) . The other factors in descending order of importance were (3) higher teacher/pupil ratios (5.7581), (4) better qualified teachers (5.6805), (5) grade teachers with greater experience (5.5157), (6) percentage of trained teachers (4.6942), (7) principal's qualifications (4.6018), (8) overcrowding, a negative factor (-4.3689), and (9) attendance 1978/79 (4.2258). That is, the upper 33 percent of students in urban areas differed from the lower 33 percent in that they took the examinations at higher grade levels, attended school more regular!-/, had more qualified principals who were younger rather than older in administrative experience. In their schools the teacher/pupil ratios were higher but teachers were more qualified and experienced. The school historically had a higher percentage of trained teachers and were less c ro w d e d . Hypothesis 6 Stated in the null form, hypothesis 6 reads: There is no linear combination of input variables which significantly separates the top 33 percent from the bottom 33 percent of pupils sampled in urban schools based on the individual totals of the 1980 CEE scores. Hypothesis 6 examined the differentiation when socioeconomic Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 127 status was added to the twelve predictor variables. Table 41 shows the variables together with the standard discriminant functions and ranking. The Chi Square is 155.2898, with 13 degrees of freedom and probability p < .00005. The hypothesis was rejected. The group means were 11.868 for the upper 33 percent and 10.564 for the lower 33 percent of urban entrants. Examination of the results shows that a cluster of variables similar to those identified in the earlier analysis differentiated the groups. Higher socioeconomic status (7.7222), higher grade level attained when the pupils sat the Common Entrance Examination in 1980 (6.3814), and principals with less experience in administration (-6.3587) were the chief factors separating the two groups. Next in descending order of importance were (1) teacher/pupil ratio (5.5071), (2) grade teachers' experience (5.5052), (3) grade teachers' qualifications (5.4806), (4) overcrowded schools (-5.0513), (5) principal's qualifications (4.5139) and (6) sessions attended 1978/79 (3.8607). That is,with the socioeconomic variable introduced, the upper 33 percent of students in urban areas differed from the lower 33 percent in that they came from higher socioeconomic levels, took the examination at higher grade levels, and attended school more regularly. Their principals were qualified but were younger rather than older in administrative experience. The teacher/ pupil ratio tended to be higher but the grade teachers were qualified and experienced and there was less overcrowding. Hypothesis 7 Hypothesis 7 tested the degree of differentiation between the upper and lower 25 percent of pupils sampled in the rural areas. Hypothesis 7 states: There is no linear combination of input variables Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 128 TABLE 41 DISCRIMINANT FUNCTIONS— UPPER 33 PERCENT COMPARED WITH LOWER 33 PERCENT OF URBAN STUDENTS THIRTEEN VARIABLES S tan d ard Discriminant Rank Variable No. Name W eights 1 Sex 1.3205 12 2 Age 1.0102 13 3 Parent's socioeconomic statu s 7.7222 1 4 Y ears in school 2.7320 11 5 Sessions attended 78-79 3.8607 9 6 Crade CEE sat 1979/80 6.3814 2 7 Teacher's qualifications 5.4806 6 8 Teacher's experience 5.5052 5 9 Teacher/pupil ratio 5. 5071 4 10 Percentage o f train ed teachers 2.9147 10 11 Principal's qualifications 4. 5139 8 12 Principal's years of experience -6.3587 3 13 Overcrowding -5 .0 5 1 3 7 Groups Means (D .F.) * Chi Square: 1 55.2898 Upper 11.868 Degrees of freedom: 13 Lower 10.564 p = < .00005 *D.F. = Discriminant Function Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 129 which significantly separates the top 25 percent from the bottom 25 percent of pupils sampled from rural areas based on the individual totals of the 1980 CEE scores. The twelve input variables are those given for hypothesis 4. Table 42 shows the variables, together with standard discriminant functions and rankings. The approximate Chi Square is 31.8618 with 12 degrees of freedom and probability p< .00005. The hypothesis was rejected. The group means were 9.108 for the upper 25 percent and 7.921 for the lower 25 percent of urban entrants. The percentage selected (25 percent) gave better differentiation between the groups than the 33 percent for urban groups. The most important variable was (1) percentage of trained teachers (8.2524), followed by (2) principals with less administrative experience (-5.1115), (3) higher grade levels attained when the pupils sat the Common Entrance Examination in 1979/80 (4.7123), (4) number of sessions attended in 1978/79 (4.4307), (5) overall lower teacher/pupil ratios (-3.7651), and (6) principal's qualifications (3.0646). Overcrowding did not emerge as a significant factor separating the groups. This implies that the upper 25 percent of students in rural areas differed from the lower 25 percent in that they took the examinations at higher grade levels, attended school regularly, their principals were generally qualified but were younger rather than older in administrative experience. Their schools had higher percentages of trained teachers and lower teacher/ pupil ratios. Hypothesis 8 Hypothesis 8 in the null form reads: There is no linear combination of input variables which significantly separates the top 25 percent from the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 130 TABLE 42 DISCRIMINANT FUNCTIONS— UPPER 25 PERCENT COMPARED WITH LOWER 25 PERCENT OF RURAL STUDENTS TWELVE VARIABLES S tan d ard Discriminant Variable No. Name R ank W eights 1 Sex . 5106 11 2 Age -.3 4 6 6 12 3 Y ears in school -.9 4 7 2 10 4 Sessions attended 1978/79 4.4307 4 5 C rad e CEE sat 1979/80 4. 7123 3 6 Teacher's qualifications 1.4171 8 7 Teacher's experience 2.0023 7 3 Teacher/pupil ratio -3. 7651 5 9 Percentage of trained teachers 8.2524 1 10 Principal's qualifications 3.0646 6 11 Principal's years experience -5.1115 2 12 Overcrowding .9773 9 Groups Means (D.F.)* Chi Square (approximate): 81.8618 Upper 25% 9.108 Degrees of freedom: 12 Lower 25% 7.921 p = < .00005 *D .F. = Discriminant Function Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 131 bottom 25 percent of pupils sampled in rural schools based on the indivi dual totals of the 1980 CEE scores. The thirteen input variables used were those of hypothesis 3, i.e., with socioeconomic status added. The results are presented in table 43. The approximate Chi Square was 87.2743 with 13 degrees of free dom and probability p< .00005. The hypothesis was rejected. The group means on the discriminant function were 9.893 for the upper 25 percent and 8.639 for the lower 25 percent of entrants. Examination of the results shows that with the socioeconomic variable introduced in the rural areas the most differentiating factor was still (1) percentage of trained teachers (8.6974), followed by (2) principals with less administrative experience (-5.2244), (3) higher grades attained when pupils sat the Common Entrance Examination in 1979/80 (4.4863), (4) lower teacher/pupil ratios (-4.4706), and (5) number of sessions attended (4.2511). Socioeconomic status did not emerge as a strong factor contrary to expectation. This implies that the upper 25 percent of students in the rural areas differed from the lower 25 percent in that they attended school regularly, and longer, were in schools with higher percentages of trained teachers over the years, their principals yvere qualified but younger rather than older in administrative experience, and the teacher/pupil ratio was lower. The Exploratory Study The exploratory segment of the study obtained the opinions of 177 teachers on sixteen categories of indicators subcategorized under school, parish, and Island. This resulted in a total of 31 stimulus statements being presented to the teachers for rating on a continuum Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 132 TABLE 43 DISCRIM INANT FUNCTIONS— UPPER 25 PERCENT COMPARED WITH LOWER 25 PERCENT OF RURAL STUDENTS THIRTEEN VARIABLES Standard Discriminant Rank Variable No. Name Weights 1 Sex .3674 12 2 Age -.0916 13 3 Parent's socioeconomic status 2.3687 7 4 Years in school -.9828 11 5 Sessions attended 1978/79 4.2511 5 6 Grade CEE sat 1979/80 4.4863 3 7 Teacher's qualification 1.7821 9 8 Teacher* s years experience 2.2957 8 9 Teacher/pupil ratio - a . 4706 4 10 Percentage of trained teachers 8.6974 1 11 Principal's qualifications 2.9918 6 12 Principal's years of experience -5.2244 2 13 Overcrowding 1.1031 10 Croups Means (D.F.) * Chi Square (approximate): 87.2743 Upper 9.893 Degrees of freedom: 13 Lower 8.639 p = < .00005 *D.F. = Disciminant Function Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 133 of 1-5, 1 being most important. The data gathered were submitted to the Andrews University computer center. Analysis was made by the categor ical scaling method involving conversions to an arbitrary interval scale and rank ordering of the derived values. Results of the Opinionnaire Table <14 presents the rank ordered list. The ordering of the sixteen categories of variables when parish and Island levels are brought together is shown in table 45. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. .8430 .7310 .7680 .7850 .8070 .6200 . 7190 .6630 . . 3290 . 5040 .4480 .1600 .2060 .1210 . . 0000 Values O rd ered Indicators TABLE 44 RANK ORDERINC OF INDICATORS BASED ON TEACHER RATINGS Percentage of overcrowded schools (for the Island) Percentage overcrowdedTeacher/pupil schools (by Percentage ratio--trained parish) of unstable teachers families (by (by parish) school) Index of malnutrition (by parish) Ratio enrollment to school age population (by parish) Degree of overcrowding (by school) Overall teacher/pupilGrade-by-grade ratio (by average parish) achievement in mathematics (by school) Percentage of pupils attending of 90% school sessions by term Teacher/pupil ratio--trained teachers (by school) Grade-by-grade average achievement in English (by school) or year (by school) Efficient principal (by school) 9 2 5 Overall teacher/pupil ratio (by school) 7 1 8 3 46 Index of malnutrition (by district) 15 11 10 12 13 14 Rank O rd er Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16 Percentage of Pupils attending 90% of school sessions by term or year (for the Island) .8530 17 Percentage of pupils attending 901 of school sessions by term or yearly (by parish) .8590 18 Ratio enrollment to school-age population (for the Island) . 9010 19 Percentage of teacher turn-over annually (by school) . 9130 20 Percentage of students achieving below 50% in English and mathematics (by school) . 9283 21 Crade-by-grade average achievement in mathematics (for the 1 stand) 1.0076 22 Percentage of unstable families (by parish) 1.0201 23 Percentage of teacher turn-over annually (for the Island) 1.0865 214 Crade-by-grade average achievement in English (by parish) 1.0759 25 Percentage of eligible age group obtaining passing mark in Common Entrance Examinations (not necessarily awards) 1.0802 (by school) 26 Grade-by-grade average achievement in mathematics (by parish) 1.0997 27 Grade-by-grade average achievement in English (for the Island) 1.1502 28 Number of library books per pupil (by school) 1.1505 29 Percentage of teacher turn-over annually (by parish) 1.2535 30 Percentage of eligible age group obtaining passing mark in Common Entrance Examinations (not necessarily awards) 1.2761 (by parish) 31 Number of hours of educational T .V . viewed (by school) 1.8387 136 15 . . 22 23 . 29 . 20 . . . 21 . 12 . 19 . 25 . . 30 28 . . . 11 . 26 . . 31 ...... Name Name by Teachers Indicator Rank Accorded for the Is la n d for the Is la n d by school by school by school by school by school by parish by p a ris h by p a ris h by p a ris h by school by p a ris h for the I s l a n d in in mathematics Percentage students achieving Percentage unstable families Percentage teacher turn-over Percentaqe overcrowded schools Percentage eligible group obtaining pass mark In CEE (not necessarily Number of library books per pupil Crade-by-grade achievement Number of hours of E T V below below 50 percent In English and mathematics awards) annually 9. 15. 16. 10. 19. 11. 12. Order of Importance TABLE 1*5 9 1 7 3 8 9 17 29 27 by Teachers Rank Accorded ...... 18 ...... 2 . . . . 13 5 ...... 10 ...... 16 ...... 6 ORDERING OF SIXTEEN INDICATORS ...... Name Indicator ...... overcrowding for the I s l a n d by s c h o o l by p a r is h by p a r is h by s c h o o l by s c h o o l by s c h o o l by p a r i s h by p a r is h by s c h o o l by district by s c h o o l for the I s l a n d by p a r i s for h the I s l a n d by p a r i s h 90 90 percent of school sessions 13. teachers Index of malnutrition Teacher/pupll ratlos--tralned Degree of Ratio enrollment to school-age Percentage pupils attending In English Crade-by-grade achievement by term or year population Overall teacher/pupil ratio , 1 . 1 2. 7. 8. 5. 9 3. Efficient principal 6. Order of Importance Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 137 Additional Indicators Provided by Teachers Teachers provided a total of eighty-nine additional items. The details of this material are in appendix M. Table 46 provides the categories assigned the items and the tallies. TA B L E 46 ADDITIONAL INDICATORS PROVIDED BY TEACHERS In d icato rs T a lly 1. Teachers and staff 18 2. School working conditions 16 3. P aren ts and home 13 4. Pupil behaviors 12 5. C u rric u lu m 7 6. D iscip lin e 6 7. Diet and nutrition 5 8. Ministry of Education inputs 3 9. Pa rent/Teachers Association 2 10. O th e rs 7 Total 89 The above included items not listed anywhere in the study. These were: 1. Teacher/teacher, teacher/pupil interaction 2. Communication with parents 3. Teachers'salaries 4. Teacher efficiency, interest in pupils, and training 5. Public-relations training 6. Discipline 7. Textbooks related to Jamaica Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 138 8. A national syllabus 9. Health facilities for schools 10. School security 11. Ministry of Education supervision 12. School Boards of quality 13. School tra n s p o rta tio n 14. School/community relationships 15. Parent/Teachers Association, and 16. No p o litic s in school. Factors Contributing to CEE Success as Supplied by Principals Principals identified, mainly through interview, a number of factors contributing positively or negatively to Common Entrance Examination success (see appendix N). These generally paralleled factors supplied by the teachers, focusing on home and environmental conditions, organized and disciplined class work, good attendance, adequate books, pupil motivation, native intelligence, and, from one rural school, distance to be travelled in order to get to school. Summary of the Findings This was a two-part study, part one being a correlational/ predictive study and part two an opinionnaire survey. In part one ten of the thirteen hypotheses tested were rejected. Three were retained. Table 47 brings together the findings of both parts of this study by displaying the rankings of the variables within their respective analyses. This enables comparison of the rankings across analyses and highlights the following: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 139 Survey 9 7 5 3 8 4 6 13 11 Rural Thirteen Variables 3 7 1 1 2*9 2* 10 6 11 12 12 10 1 - 9 4 6 8 3 7* 8 13 11 10 Discriminant Analyses 1 2 5 5 3 4 5 9 7 - 2* 11 12 Urban Twelve Variables Urban Variables Thirteen Variables Twelve Rural 5 6 4 7 8 - 12 10 7 3 5 4 1 1 6 2 2 8 4* 3* 13 11 12 11 9 12 TABLE 47 1 0*1 10* 8* Thirteen Variables Twelve Variables Canonical Correlation RANKING RANKING OF FINDINGS FOR SIX STATISTICAL TESTS AND TEACHER SURVEY OPINIONNAIRE (3 year-average) 7. Teacher's qualifications 2. Age 3. SES--Parent's socioeconomic status 1. Sex 5. Sessions attended 978/79 1 9. Teacher/pupil ratio (3 year average) 9 8. Teacher's experience 6 4. Years in school 6. Grade CEE sat 1979/80 V ariab le 11. Principal's qualification 10. Percentage of trained teacher 12. Principal's years of experience 13. Overcrowding Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. h I Oh 7 9 8 2 3 4 6 11 12 14 10 13 15 16 teachers 50 50 percent--English and mathematics passmark in CEE school sessions 19. Enrollment to school age population 14. Teacher/pupil ratio--trained 18. English achievement 25. Percentage pupils achieving 26. Library 27. T . V . 20. Mathematics achievement21. Percentage overcrowded22. Unstable schools family 23. Teacher turn-over24. Percentage pupils achieving below 15. Efficient principal * = Negative Correlation 16. Malnutrition 17. Percentage attending 90 percent of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 141 1. Ratio or percentage of trained teacher occupied the highest rank on both canonical analyses and on the two rural discriminant analyses. This factor was also important to the teachers, being ranked second. It occupied a lower ranking on both urban discriminant analyses. 2. Socioeconomic status had a high ranking in urban areas when this was introduced into the analysis. It was ranked first in the discriminant analysis of the two urban groups. 3. Number of sessions attended was ranked highly on both canonical analyses (second on both) and in the discriminant analyses on the two rural groups. 4. Taking the examination when in higher grades in school had a clear advantage in all analyses. 5. The grade teacher's experience, followed closely by his/her better qualification, was also important. 6. Principal's years of administrative experience correlated negatively with achievement in the analyses. 7. Overcrowding was negatively related in four analyses. 8. Overcrowding was perceived by the teachers to be the most important factor affecting performance. This was followed by five factors, in descending rank: the trained teacher/pupil ratio, an efficient principal factor, a malnutrition assessment factor, an overall teacher/pupil ratio factor, and an attendance factor. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C H A P TE R V SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS Summary B ackground The main purpose of this study was to postulate a single or a set of indicators which would provide timely and significant statistical information for planning and monitoring primary education in Jamaica. This was to be accomplished through relating the research findings in the field to the findings of (1) a correlational study of Jamaican primary schools inputs and outputs, and (2) a survey of teacher opinions on significant factors contributing to educational achievement. Chapter 2 reviewed the literature to provide a contextual framework. Indicators were defined as measures of a system1 s performance, statistics of normative interest signalling change in variables of interest, registering progress towards or retrogression from a goal. The systems model was suggested by some authors as a basis for the classification of educational indicators. Accordingly there could be "context," "input," "process," and "output" indicators. Another categorization listed " Access," "Aspiration," "Achievement," "Impact," and "Resources" indicators. A third categorization proposed "Personal Development," "Social Outcomes," and "Political Outcomes" indicators. Indicators were looked at in terms of " primary" or present 1U2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. effects, " secondary" or long-term effects, and " tertiary" or inter- generational effects, indicators were discussed in the context of economic analysis. Reference was made to the "leading," "co-incident, and "lagging" economic indicators. Indicators were said to be "manipulative" and " non-manipulative." Finally, it was questioned whether there could be a single educational indicator as informational as the demographic indicator. Infant Mortality Rate. Actual work on the development of educational indicators in the U.S. and abroad was reviewed. The most extensive work was that of Johnstone (1981) who surveyed inter- and intra-national indicators put forward by various professionals and United Nations' Education, Scientific, and Cultural Organization (UNESCO). Johnstone (1981) himself recommended a profile of indicators categorized according to the systems model, together with statistical formulas. Literature on the variables used in the study was also reviewed. Summaries of research findings on school input/output relationships in the United States and overseas countries were provided Socioeconomic status was, in most cases, positively correlated with achievement. In some developing countries, however, it had less effect Bargen and Walberg (1974), in a summary of twenty-three studies on United States and British subjects, found that teacher variables were, in the main, significantly correlated with school achievement. A summary of overseas research provided by Alexander and Simmons (1975) showed variable effects for most of the individual schooling variables such as per pupil expenditures, class size, teacher certifica tion, tenure, experience, and sex. Text book availability, availability Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 144 and use of library, homework and free reading were also significantly related to achievement in the studies done on these variables. Literature on the primary-school system in Jamaica was reviewed. The commentary offered by Vernon (1961) on the Common Entrance Examination (CEE) selection for secondary education in Jamaica provided information on the sex differential in performance, the effect of maturation, what the tests were supposed to measure and how Jamaican children performed in comparison to English children. Manley's (1963) study on the results of the 1959 Common Entrance Examination—the entrance examination to the academic high schools in Jamaica—pointed out the differential effects of socioeconomic status and the sex differential in terms of examination performance and participation. The apparent lower participation of boys in the education system was underscored. Reid (1964), in a comprehensive study, examined the relation ship of a profile of home, community, and school inputs with primary- school attainment. His aim was to measure the interaction of these factors on the criterion of attainment. The most essential contributing factor to school achievement was good home background, the education of the parents themselves,and their interest in their children's education. Years of schooling and qualifications of staff did make a smaller inde pendent contribution. Other Jamaican studies reviewed showed that underachievement of Jamaican children in English schools was related to the parents' low educational achievement (Bagley, 1979). Motivation was, however, important (Ragbir, 1977). The pattern of absenteeism, the deficiencies Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 145 of educational television, the level of provision of library books,and the annual data and statistical information published by the Ministry of Education were also reviewed. Methodology— Correlational Predictive Analyses The methodology for the correlational predictive study involved a stratified sample of forty urban and rural schools. These were stratified on a dimension of performance in the 1980 Common Entrance Examination. Ten each of urban and rural schools showing high percentages of CEE awards to the relative age cohort in the relevant grades (grades 4-6) were selected. Ten each of urban and rural schools showing fewest or no awards were also selected. Data were secured in three ways: they were (1) taken directly from the CEE score cards, (2) secured from school principals, and (3) secured from the Ministry of Education. Difficulties in securing data from the schools resulted in two schools having to be omitted from six of the thirteen hypotheses. The first seven hypotheses were therefore tested on forty schools and 1,048 sub jects,and the last six, consisting of canonical correlation and discriminant analyses,were tested on thirty-eight schools and 802 sub jects. Of the completed cases 285 were from rural and 517 were from urban areas. Sixteen predictor variables were originally selected for analy sis. These were later reduced to thirteen through the elimination of the variables (1) "Library," (2) "Educational Television," and (3) pre paratory school experience. Sufficiently differentiated data were not secured on the first two; reliability of the last was questioned. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1H6 The indexing of socioeconomic status from parent's occupation listed on the CEE cards presented assessment difficulties. Accordingly the last six hypotheses were tested with this variable included in three analyses and excluded from three. The thirteen variables finally used were : 1. Pupil1 s sex 2. Pupil1 s age 3. Parent1 s socioeconomic status 4. Number of years in the particular school 5. Number of sessions attended in 1978/79, the year prior to the 1980 CEE 6. Grade from which the 1980 CEE was sat 7. Crade teacher1 s qualifications 8. Crade teacher's years of experience 9. Averaged pupil /teacher ratios for three school years— 1977/78 - 1979/80 10. Percentage of trained teachers averaged over the three years— 1 977/78 - 1979/80 11. Principal's qualifications 12. Number of years administrative experience of principal 13. Degree of overcrowding at the school. The data secured gave a fair coverage of the pupil's personal variables, including his attendance rate and cumulative experience in the particular school. School variables were also well represented, for the year of interest and historically for three years previously . The aims of the analyses were: 1. To examine the intercorrelations of the three test batteries of the Common Entrance Examination. This would establish whether Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 147 the tests showed low correlations among themselves. 2. To ascertain whether there was a significant difference in test performance among the ten best performing urban schools. 3. To ascertain whether by utilizing the test scores of all entrants there was a significant linear combination of school and personal inputs correlating with a linear combination of the scores of the three batteries. 4. To ascertain what factors significantly separated: (a) the first of the ten best performing schools from the last, should a signifi cant difference among the school means on the overall scores be found; (b) the top 33 percent from the bottom 33 percent of urban entrants; and (c) the top 25 percent from the bottom 25 percent of rural entrants. A Pearson product-moment correlation coefficient of ±.30, and probability p < .05 were the statistical parameters set. Survey Methodology The second part of the study, the sampling of teachers' opinions,was accomplished through random circulation of a questionnaire to teachers mainly in the CEE grades in the schools sampled. Ratings on a continuum of 1-5, 1 being most important, were requested on sixteen indicator items, disaggregated into thirty-one responses related to school, parish, and Island as appropriate. The responses from 177 teachers were tallied. Categorical scaling was then used to place the items along an arbitrary interval scale. These were then ranked. F indings Thirteen hypotheses were tested under eight headings. Hypotheses under seven headings were rejected and retained under one. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 148 Hypotheses 1(a)-1(c) tested the degree of correlation among the three batteries of the 1980 CEE. Significantly high correlations among the test scores of 1,048 entrants from forty schools were found. English and intelligence tests were the most highly correlated (.85), followed second by intelligence and mathematics (.82), and last by English and mathematics (.78). Hypotheses 2(a)-2(d) tested the significance of the difference between the school means of the ten best performing urban schools on the total CEE test, and on the three individual tests of the 1980 CEE battery. This hypothesis was retained in respect to (1) total scores, (2) scores on the English, and (3) scores on the mathematics tests, there being no significant differences in the school means on these tests. There was, however, significant differentiation on the intelligence test. The table of means was examined utilizing the Newman Keuls a-posteriori test. St. Aloysius, in particular, and four other schools were shown to have achieved significantly different means from other schools. The reasons for this finding merits further investigation which will not be undertaken in this study. As no significant difference was found among the means of the total CEE battery for these schools, no discri minant analysis was performed to isolate the factors separating them. Hypotheses 3 and 4 established by canonical analysis that there was a significant linear combination of predictor variables covarying with the criterion scores of the three test batteries of the 1980 CEE in respect to 802 entrants. Hypothesis 3 utilized thirteen input variables; hypothesis 4 utilized twelve variables, the variable socioeconomic status being excluded. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 149 Hypotheses 5 and 6 utilized discriminant analysis to identify those variables which separated the top 33 percent from the bottom 33 percent of urban entrants. Hypothesis 5 utilized twelve variables; hypothesis 6 utilized thirteen variables, socioeconomic status being the added variable. Individual total CEE scores separated the groups. Hypotheses 7 and 8 utilized discriminant analysis to isolate variables significantly separating the top 25 percent from the bottom 25 percent of rural entrants. Hypothesis 7 utilized twelve variables; hypothesis 8 utilized thirteen variables, with socioeconomic status being the added variable. The findings of the study are summarized in tables 48-50. The canonical analyses and the discriminant analyses were all significant and identified factors associated with school success. The tables list the variables in order of the weights they showed in the analyses. Essentially the same pattern of school inputs correlated with success with or without socioeconomic status included. Socioeconomic status had a high weighting in the urban discriminant analysis. The co varying variables were different in respect of urban and rural areas when top and bottom percentages of entrants were contrasted. The ranking of the sixteen major indicators based on ratings by the teachers is shown in table 51. Conclusions Both canonical analyses show the correlates of high performance in 1980 CEE to be: 1. Higher percentage of trained teachers 2. Greater number of sessions attended Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 150 TABLE 48 SUMMARY RANKING OF PERSONAL AND SCHOOL VARIABLES (Canonical Correlations) Thirteen Variables Twelve Variables Rank Item Standardized Item Standardized C o efflcen t Coefficient 1 Percentage of trained teachers (.3 7 1 ) (.417) 2 Sessions attended 78/79 (.352) (.408) 3 Parent's socioeconomic status (. 346) Not included 4 Principal's years of experience -( .345) -(.4 0 4 ) 5 C ra d e CEE sat 1 980 (.3 3 4 ) (.383) 6 Teacher's years of exp erien ce (. 248) (.257) 7 Teacher's qualifications ( .221) (.216) 8 Principal's qualifications (.1 3 8 ) (.2 0 9 ) 9 Teacher/pupil ratio (.1 1 5 ) (.157) 10 Overcrowding - ( .1 1 2 ) Sex (.1 2 7 ) 11 Sex (.103) Over -(.1 1 6 ) cro w d in g 12 Y ears in school (.0 4 6 ) Age (.025) 13 Age (.027) Y ears in (.0 0 9 ) school KEY: Broken line ( ------) separates most important variables Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 151 TA B LE 49 SUMMARY RANKINC OF PERSONAL AND SCHOOL VARIABLES—URBAN (Discriminant Analysis) Rank Thirteen Variables Twelve Variables Discriminant Discriminant 1 tern IUm Function Function 1 Parent1 s socioeconomic status (7.72) Not included 2 Crade CEE sat 79/80 (6.38) (7 .5 4 ) 3 Principal's years of experience -(6.36) -(6.52) 4 Teacher/pupil ratio (5 .5 1 ) (5 .7 6 ) 5 Teacher's experience (5.50) T eacher's (5 .6 8 ) qualifications 6 Teacher's qualifications (5 .4 8 ) T e a c h e r's (5 .5 2 ) exp erien ce 7 Overcrowding -( 5.05) P ercentage (4 .6 9 ) tra in e d teachers 8 Principal's qualifications (4. 51) (4 .6 0 ) 9 Sessions attended 78/79 (3 .8 6 ) O v e r - ( 4 .3 7 ) crow ding 10 Percentage of trained Sessions (4 .2 3 ) teachers (2 .9 1 ) atten d ed 11 Y ears in school (2.73) (2.05) 12 Sex (1 .3 2 ) (1 .0 2 ) 13 Age (1 .0 1 ) ( -58) KEY: Broken line ( -----) separates most important variab les Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 152 TABLE 50 SUMMARY RANKINC OF PERSONAL AND SCHOOL VARIABLES—RURAL (Discriminant Analysis) Thirteen Variables Twelve Variables Rank Item Discriminant Item Discriminant Function Function 1 Percentage of trained teachers (8 .7 0 ) (8 .2 5 ) 2 Principal's years of e x p e rie n c e - ( 5 .2 2 ) - ( 5 .1 1 ) 3 Crade CEE sat 79/80 (4.49) (4.71) 4 Teacher/pupil ratio -( 4 .4 7 ) Sessions (4 .4 3 ) attended 5 Sessions attended 78/79 (4.25) T each er/ - ( 3 .7 7 ) pupil ratio 6 Principal's qualifications (2.99) (3 .0 6 ) 7 Parent's socioeconomic status (2 .3 7 ) Not included 8 Teacher's experience (2.30) (2 .0 0 ) 9 Teacher* s qualifications (1.78) (1.42) 10 Overcrowding (1 .1 0 ) ( .9 8 ) 11 Years in school - ( .9 8 ) - ( .95) 12 Sex ( .37) ( -51) 13 Age - ( .0 9 ) - ( .35) KEY : Broken line ( ----- ) separates most important variables Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 153 TABLE 51 SUMMARY RANKINGS OF TEACHERS' OPINIONS ON SIXTEEN VARIABLES 1 Degree of overcrowding (school) 2 Teacher/pupil ratios (trained teachers) 3 Efficient principal 4 Index of malnutrition 5 Overall teacher/pupil ratio 6 Percentage pupils attending 90 percent of school sessions 7 Crade-by-grade achievement in English 8 Ratio enrollment to school age population 9 Crade-by-grade achievement in mathematics 10 Percentage overcrowded schools 11 Percentage unstable families 12 Percentage teacher turn-over 13 Percentage students achieving below 50 percent in English and mathematics 14 Percentage eligible group obtaining pass mark in CEE (not necessarily awards) 15 Number of library books per pupil 16 Number of hours of ETV Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 154 3. Socioeconomic status (if included) 4. Principals with less administrative experience 5. Higher grade level attained 6. Teacher1 s years of experience 7. Teacher's qualifications 8. Principal1 s qualifications. The factors which separated urban high performers from low performers were: 1. Higher socioeconomic status of parent (if included) 2. Sitting the CEE at higher grade levels 3. Schools where principals had fewer years of administrative experience 4. Schools with relatively higher teacher/pupil ratios 5-6. Teachers with higher qualifications and more years of experience 7-8. Less crowded schools and principals with higher qualifications 9-10. More sessions attended and higher percentage of trained teach ers. For rural areas, the factors separating high performers from low performers were: 1. Higher percentages of trained teachers 2. Principals with fewer years of administrative experience 3. Sitting the CEE when in a higher grade 4-5. More sessions attended and lower teacher/pupil ratios 6. Principals with better qualifications 7. Socioeconomic status (if included) 8. Grade teachers with higher qualifications and more years of experience. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 155 The teachers' survey produced the following high factors: 1. Degree of overcrowding 2. Teacher/pupil ratio— (trained teachers) 3. Efficient principal 4. index of malnutrition 5. Overall teacher/pupil ratio 6. Regular attendance 7. Achievement in English 8. Ratio enrollment to age cohort 9. Achievement in mathematics 10. Unstable family life . Teachers added eighty-nine items as important indicators. These were placed in ten categories. The four most important were: 1. Teachers and staff 2. School working conditions 3. Parents and home 4. Pupil behaviors. Finally, the main factors mentioned by principals as important contributing factors, negatively and positively, to performance in the Common Entrance Examination were: 1. The home dimension— stable families with adequate supervision and parental help 2. Attendance 3. Books 4. Pupil motivation and intelligence. It should be noted that the top ten indicators rated by the teachers Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 156 included five which emerged as important in the various segments of the correlational analyses. These were: 1. The principal and his efficiency, cf. principal's qualifications and years of service from the correlational study 2. The trained teacher input 3. The teacher/pupil ratio input 4. Regular attendance 5. Degree of overcrowding. The other five highly rated indicators as presented to them but not included in the correlational study as input/output variables were: 1. Degree of malnutrition 2. Achievement in English 3. Ratio enrollment to school-age population 4. Achievement in mathematics 5. Unstable families. Implications Certain findings in this study appear surprising and bear closer study. 1. Overcrowding had a low weighting in rural areas. This may reflect the fact that there are fewer overcrowded schools in rural areas. That it does not emerge more heavily weighted in urban areas is also surprising. 2. Principal's administrative experience (negative) may be a function of distribution in the sample rather than a contributing factor. 3. High teacher/pupil ratios (positive) led to the question of whether the effects are as negative as is generally supposed or whether Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 157 these negative effects are mitigated by other inputs. 4. Socioeconomic status did not emerge as important in rural areas. This is doubtessly because there was less differentiation along this variable in rural areas. In urban areas, where there was greater differentiation, the weight of this variable was more clearly seen . Comments In spite of the possible intervening variables the findings of this research point to the following implications: 1. A different composite of factors are contributing to school success in rural areas from those in urban areas. Higher percentage of trained teachers, longer and better school attendance, lower teacher/ pupil ratios are rural differentiating inputs. In fact, where socioeconomic status is more homogeneous schooling inputs may make a more marked difference than where there is greater differentiation in socioeconomic status. It must be pointed out, however, that rural success may be relatively lower overall than urban success. When socioeconomic status is introduced into the urban analyses, it becomes a major factor influencing school success. Nevertheless, even in overcrowded urban schools with high teacher/pupil ratios, some children with longer and regular school experience, and under better qualified and more experienced teachers, were performing better. Officials responsible for financing education should begin to look at these input variables in order to maximize their use and observe the effect on pupil performance. Although the sample of teachers was limited in number, their rating of indicators is interesting. Mention has already been made of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 158 the three factors considered by them to be most important: (1) degree of overcrowding, (2) trained teacher/pupil ratios, and (3) efficient principal. The factor which ranked next in importance to these is one which is more palpable to them than to non-school personnel, i.e., the factor malnutrition. This was ranked above achievement and above percentage of school-age population enrolled in school. The implication is that this factor needs direct assessment. Other ratings of indicators by the teachers are, by and large, what would be expected. Noteworthy, however, are factors which earned the lowest ratings, and, by virtue of this, are outstanding. These were (1) percentage of children achieving pass marks in the CEE examination, (2) use of the library, and (3) hours of educational television. This might imply low perception of the importance of these factors on the part of these teachers. The CEE is regarded, in many quarters, as a "necessary evil." Its virtues, in terms of being a standardized test, are not appreciated. The exam is perceived (1) as excluding children from an elitist type of education which is viewed as desirable, and (2) as producing negative self-concept in the children who fail the exam. School library and educational television services received low ratings. This may imply that these two resources are not perceived as important educational inputs. Teachers' and the Ministry of Education's supplies (books, etc.) are perhaps regarded as more important inputs. Family of Indicators What then may one put forward as the family of indicators or the critical indicator? The researcher submits that these must be determined not only by the conclusions of this research and previous Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 159 findings as cited in the literature, but also by the present activities in and needs of the Jamaican educational system. These include: 1. High absenteeism whether through irregular attendance or major periods of absence during the school year 2. The steps being taken by the Ministry of Education to address the problem through enforcement of the Compulsory Education Clauses of the Education Act 1965 3. The fact that more girls than boys in the age cohort may be participating in the educational system. More girls (22,979) than boys (12,718) took the CEE in 1980 (Education Statistics 1979/80). It is recommended that indicators for the Jamaican primary system be categorized on a systems/economics/time dimension into: 1. Context indicators/leading indicators 2. Input/coincident indicators 3. Process/contributing indicators 4. Output/lagging indicators. Indicator 1—A Context/Leading In d icato r The first and most critical indicator therefore should be the ratio of the age cohort attending school for a stated percentage of sessions each term or year broken down by gender (male; female). This is seen as a critical indicator. If pupils are not in school they cannot be exposed to the effects of schooling; the longer the pupil stays in formal education, the more likely it is that he/she will leave school literate and numerate. This was borne out by the findings of this research. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 160 It must be underscored that this statistic is distinctly different from the average attendance statistic currently being used. The average attendance statistic masks the attendance record of the indivi dual pupil. A prerequisite to the development of this statistic is know ledge of the schools catchment area— each school board managing a pupil account. It is also different from the male/female flows put forward by Johnstone (1981): This indicator could be expressed as f(n) + m(n) = n ; TOT) + tTTn T fT where f = female, m = male, n = numbers attending (say 90 percent) of the sessions, N = the number in the age cohort of the school catchment a re a . The need for and importance of this indicator cannot be over stressed. It would clearly point up when, where, and in what categories, grade-wise or sex-wise pupils were not participating in the education system. Disaggregation would be possible and diagnostic and remedial action could be undertaken on a timely basis. If other school conditions are propitious, watching this indicator would forecast how well pupils will perform over the period of primary school experience. Indicator 2—A Process/Contributing In d ic a to r The researcher submits that the second critical indicator should be along the teaching dimension. The findings of this research suggest that the percentage of trained teachers and the teachers' length of experience combined have definite co-varying relationships with achieve ment. Teacher concerns figured strongly in teachers' responses both to the survey questionnaire and on independent items. This finding also Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 161 suggests that the CEE test requirements are being understood better by experienced teachers. These grade teachers are known to give extra lessons to CEE pupils, and it must be admitted that the trained, experi enced teacher contributes significantly to the CEE successes. Therefore, if standards or objectives are set and enforced for the whole primary system, teachers could be encouraged to produce as they do against the CEE criteria. Two questions arise: First, could this "experience" skill be taught? Second, could there be a statistic combining both factors to provide a quick measure of expectation from the system? It is suggested that this formula could be the ratio of trained teachers with x years of experience over the number of pupils, by area of interest. This could be a secondary statistic to pupil/teacher ratios (now being collected) or percentage of trained teachers in schools. Further research would be necessary to ascertain what should be the base number of years or whether the relationship is curvilinear. Teaching style could also be studied. Indicator 3— A Output/Lagging In d ic a to r The third critical indicator must be an output/outcome indicator which measures the success of the system. This should be an assessment system which quantitatively describes success grade-by-grade or at certain predetermined critical levels in the system— say grades H and 6. To some extent this is accomplished by the Common Entrance Examination and a three-year assessment of all primary schools. This study has pointed to the fact that quotients on the mathematics and English test of the Common Entrance Examination may not be demonstrably different after twenty-one years of similar test administration. If this Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 162 is taking place among the select group offered for the CEE, what must then be happening to the larger group? Standardized tests should be introduced. The CEE results could also be monitored on this basis. Other Indicators There are other indicators suggested by this research but on which definitive conclusions have not yet been reached. Overcrowding— a contributing indicator. This was mostly negatively related to achievement, though not as highly as might have been expected. It was the foremost concern of the teachers in the opinionnaire. The findings of this study may have been influenced by the fact that mostly "brighter" pupils were sitting the Common Entrance Examination, and that overcrowding was most adversely affecting "duller" pupils who would not have been measured at all. That overcrowding received a ranking of No. 1 of the thirty-one opinionnaire statements shows that it is a nettlesome problem to teachers. It may also be that the effects of overcrowding are not as significant as the teachers feel them to be. Further study of this problem should be made. Statistics for this indicator exist and only need to be aggregated for monitoring as a contributing indicator. Teacher/pupil ratio—a contributing indicator. In urban areas higher ratios did not seem to affect achievement; in rural areas it did. Further research is necessary to see whether after-school classes, given by teachers in the system or individuals outside of it, are mitigating the adverse effect of high teacher/pupil ratios in the schools. Establishment of a definite relationship is necessary as this is a "process" — "contributing" indicator of prime importance in the educational system. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T63 An index of malnutrition—an input indicator. This was of fourth major concern to the teachers. A ratio based on assessed levels of student malnutrition could be developed. Research establishing the relationships between malnutrition and achievement could also be under taken, although the connection seems obvious. Its extent should be m easured. Principal's effectiveness—a process indicator. The negative relationship of administrative experience of the principal needs further investigation. Does this imply that, at a specific point in time there were better educated principals with fewer years of service whose output in terms of pupil achievement was better? An efficient principal was the third indicator on the teachers' list. However, the findings of the study may be a function of distribution in a small sample. School inputs as input indicators. Research on the contribution of media and textbooks is also needed before determination can be made as to measures of effectiveness for these inputs. Should it be per unit expenditures, or numbers of units, or combinations of units as inputs? These are questions to be answered. Enrollment projections. Usefulness of enrollment projections for forecasting purposes was briefly mentioned in the literature review. The need for these projections for planning purposes is obvious. They are recommended for inclusion in the family of indicators. Graphical representations would be useful for presentation. Attention is invited to earlier work by the researcher on statistical methods for forecasting enrollment at micro levels in Hanover, Jamaica (Marshalleck, 1976). Expenditures per pupil. Mention was made in the review of literature of the covarying relationship between per pupil expenditures Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 164 and achievement. It has been brought out in the research that expenditures may be more meaningfuBy looked at when they are made against inputs which maximize outputs. Projections of total available sums against presently efficient levels should provide some useful bench marks. Further work to refine indicator measures is indicated. Primary-school life-expectancy rates. This indicator arises from the literature but also rests on the findings that higher grade levels correlate with higher CEE scores. The indicator would be built on trend analyses of primary-school completion rates. It would be a critical indicator in manpower and human resource development planning. Schema for indicators. Finally, figure 3 presents a schema for the recommended and potential indicators. This figure summarizes the discussions above. Recommendations Recommendations flowing from the research: 1. Replication of this study should be made periodically. With the information gained from restudy of these school inputs, as well as those listed in 2-7 below it would be possible for the Ministries of Finance and Education to allocate resources more effectively. 2. Further study should be given to profiling the effective principal in terms of levels of training, years of experience, as well as other measures of effectiveness. 3. Further study should be given to teacher/pupil ratios to determine desirable levels and to determine the relationship in "pure" situations—i.e ., situations where no extra tutoring is provided. Co varying relationships should be examined in randomly selected and ability streamed classes. I r Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 165 education CONTEXT --Interest In Children's --Income Parents'-Education - Socioeconomic Factors Socioeconomic LEADING INDICATORS 3- 3- OUTPUT and at Prescribed Levels Tests--Means Variances--Distributions LAGGING INDICATORS " " Other Standardized " " Numbers Achieving Literacy " " Numbers Entering Next Level " Common " Entrance Common Measures "/Nos. Trained "/Pupil Ratios "Experience 2. 2. PROCESS Index of Overcrowding Media Books Books and Teaching Materials Principals- -Effectiveness Teachers' "Qualifications COINCIDENT CONTRIBUTING INDICATORS education to Next Level 1. 1. INPUT CONTEXT CONTEXT CONTEXT --Income --Interest In Children's 0 Numbers Graduating Attending 901 Sessions Cohort Male and Female Entrants Principals Books Books and Teaching Materials Accomodation Media ’ ’ Percentage Pupils in Age LEADING INDICATORS LEADING INDICATORS Fig. 3. Schema for Indicators developed by writer LEADING INDICATORS " " Estimations of New Socioeconomic FactorsSocioeconomic „ Numbor9 Completing Ftarents'--Educatlon x grades Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 166 4. The Ministry of Education should develop means of distribu ting the effects of the skilled teacher to areas of the system deficient in this effect. Rural areas are usually most affected since qualified teachers more often locate in urban areas. Periodic tutoring and super vision by master teachers and periodic extended teaching visits by teams of master teachers are two possible solutions. 5. The Ministry of Education should undertake a cost benefit analysis of Educational Television. It should be investigated whether the medium could be used to help reduce the increasing costs of teachers' salaries. 6. The Ministries of Health and Education should cooperate in establishing regular school health examinations. Levels for under nourishment and malnutrition should be established and data gathered (a) to examine the correlation between these conditions and achieve ment, and (b) to develop indices. 7. Study of the contribution of other educational inputs to achievement should be made, e.g., (a) provision of textbooks in differing combinations and numbers, (b) provision of school meals (breakfast and/or lunch), (c) provision of uniforms, and (d) any other input felt to be highly correlated with achievement. A random sample of typical schools should be used in these studies. 8. The Ministry of Education should develop plans for annual standardized achievement assessments grade-by-grade, across the school system. If even on a sampling basis, the out-turns should be studied for changes in averages, variances,and distribution. Recommendations for administrative action: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 167 1. The "teacher dimension" stood out in the responses from teachers. The Ministry of Education should address these concerns as best as possible. Despite financial constraints, efforts should be made to come to terms with those concerns which are uppermost in the minds of teachers. Detailed study may be required on this item. 2. The Ministry of Education should review its requirements for school record keeping to make these as economical as possible in terms of items, forms to be filled out, books kept,and returns to be made to headquarters. Principals complained of the number of "statistics" to be developed for the Ministry of Education and that no additional assistance was provided them. The class list should be reviewed by the Ministry of Education for reintroduction if considered advisable. On the other hand, there is need for certain schools to preserve registers more carefully and to send reliable returns promptly. The reliability of these data is necessary for deriving valid measurements for educational studies and international statistics. 3. Based on the findings of Reid's (1964) study and others showing similar results, the Ministry of Education should mount programs aimed at sensitizing parents to their responsibilities for their children's education as well as their own continuing growth, however late it may seem to them. 4. The Ministry of Education should not only use educational indicators for planning purposes but should publicize them appropriately. This could help to stimulate interest throughout the society on the condition of primary education, a critical input in the country's program of human resource development. Eight years of education (over ages 6j-15) might be the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 168 educational heritage for a major proportion of Jamaica's school age population; it may be all that the public purse can afford for them for some time to come. The outputs from this level of the system must therefore be maximized in terms of curricular outputs and standards reached. Primary education (grades 1-6, ages 6^-12) is a vital segment of this and all other levels of education. There can hardly be, therefore, an area more needful of early indicators of progress toward goals or retrogression from them than the primary education system . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A P P E N D IX A SAMPLE PACES FROM 198*1 ECONOMIC INDICATORS C o n te n ts TOTAL OUTPUT, INCOME, AND SPENDING ^ Gross National Product...... 1 Gross National Product in 1972 D ollars ...... 2 Im plicit Price Deflators for Gross National Product ...... 2 Changes in G NP. Personal Consumption Expenditure*, and Related Price Measures ...... - ...... ) Nonftnancial Corporate Business—O utput. Costs, and Profits ...... 5 National Income ...... * Personal Consumption Expenditures ...... 4 Sources of Personal Income ..... 3 Disposition o f Personal Incom e ...... 6 Farm Income ...... - ...... Corporate Proflts ...... 8 Gross Private Domestic Investm ent ...... - ...... 9 Expenditures for New Plant and Equipment—Nonfarm Business ...... 10 EMPLOYMENT, UNEMPLOYMENT. AND WAGES Status o f the Labor Force ...... n Selected Unemployment Rates ...... - ...... 12 Selected Measures o f Unemployment and UnemploymentInsurance Programs ...... 13 Nonagricultural Employment ...... t4 Average W eekly Hours and H ourly Eamtngs— Private Nonagricultural Industries ...... - ...... 13 Average W eekly Earnings— Private Nonagricultural Industries ...... 13 Productivity and Related Data, Business Sector ...... 16 PRODUCTION AND BUSINESS ACTIVITY Industrial Production and Capacity U tilization ...... 17 Industrial Production—M ajor M arket Groups and SelectedManufactures ...... IS New Construction - ...... 19 New Private Housing and Vacancy Rates ...... 19 Business Sales and Inventories— Manufacturing and Trade ...... ~...... 20 Manufacturers* Shipments. Inventories, and O rders ...... - ...... - ...... 2\ PRICES Producer Prices ...... 22 Consumer Pnces— 23 Changes in Producer Prices for Finished Goods ...... 24 Changes in Consumer Pnces ...... 24 Prices Received and Paid by Farmers ...... 23 MONEY. CREDIT. AND SECURITY MARKETS Money Siock, Liquid Assets, and Debt Measures ...... -...... ~ ...... 26 Components o f Money Stock and Liquid Assets ...... — 27 Consumer Installment Credit ...... 2" Bank Loans and Investments, and Reserves ...... 28 Sources and Uses o f Funds. Nonfarm Nonftnancial Corporate Business ...... - ...... 29 Current Assets and Liabilities o f Nonfinancial Corporations ...... 29 Interest Rates and Bond Yields...... 30 Common Srock Prices and Y ields...... 5 1 FEDERAL FINANCE Federal Receipts, Outlays, and Debt ...... „ ...... 32 Federal Budget Receipts by Source and Outlays by Function ...... - ...... 33 Federal Sector, National Income Accounts Basis ...... - ...... - ...... 34 INTERNATIONAL STATISTICS Industrial Production and Consumer Prices— M ajor Industrial Countries ______—...... 33 U.S. Merchandise Exports and Im ports ...... 33 U.S. International Transactions ...... — ...... 36 I General Note { Detail to ebese cables may not add to totals because o f rounding, i Unless otherwise noted, all dollar figures are in current dollars. 169 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission 170 CONSUMER PRICES In April, the consumer price index for afl urban consumers rose 0~5 percent seasonally adjusted and not seasonally adjusted. The index was 4.3 percent above its level in April 1983. M OO . 1 f * 7 - 100 (KATtO SCALE) MCCX. 1 9 *7 -1 0 0 (XAnO SCALE) no 300 300 290 290 260 240 220 230 200 200 180 160 140 ^ 1976 1977 1979 1980 Ml «on r««| W u K I « **•» « « ** C* l*SOi (OW<» 9 t M in (1967 ■ 100, ncrpt u M id ; eeethly dati eeesoeeBy adjtuttd, except u M id ] r Trias porta boo All Shelter items Appar lets ah Hoar- MirnTt- FWI M«Ccal Eaer- Prno4 F«o4 el urf food. ium i 1 •wa aaace a td New Motor nr* O* ' Total • Sem •placp T iu l 1 eneffr. TouJ en' rn ’ ud •dwr can bsd1 amd CtMU > (Ml 1 reprnn ihdtrr SSA SSA KM r "I RtL im p.*...... 100.0 18.7 37.6 21.3 7.0 14.0 0.5 H2 S3 21.8 3.5 5.9 6.1 ! 11.9 47.9 1976...... 170.5 180.8 174.6 179.0 199.6 182.7 147.6 165.5 135.7 177.9 184.7 189.3 159.9 1977...... 192.2 186.5 191.1 214.7 202.2 154.2 177.2 142.9 188.2 202.4 207.3 169.5 197M...... 211.4 202.8 210.4 ...... 233.0 216.0 159.6 185.5 153.8 196.3 219.4 i 220.4 179.1 1979 ...... , 217.4 234.5 227.6 239.7 256.4 239.3 166.6 212.0 IC6.0 265.6 239.7 2759 191.5 1980...... 246.8 254.6 263.3 281.7 ...... 285.7 278.6 178.4 249.7 179.3 369.1 265.9 361.1 208.3 I 9 * t ...... 272.4 274.6 293.5 314.7 ...... — 4...... 314.4 319.2 186.9 280.0 190.2 410.9 294.5 , 410.0 228.1 1982...... 289.1 265.7 314.7 337.0 ...... 334.1 350.8 191.8 291.5 197.6 389.4 328.7 416.1 245.6 1983...... 291.7 323.1 344.8 103.0 102.5 346.3 370.3 196.5 298.4 202.6 376.4 357.3 419.3 258.4 1983; Apr...... 295.5 291.5 320.3 341.4 101.7 101.7 343.6 366.6 194.7 293.7 201.4 368.7 353.1 413.3 255.7 ifa r ... 297 ! 292.1 321.1 342.3 102.2 102.0 344.3 368.4 195.7 296.4 201.1 379.8 354.8 420.5 256.3 June ..., 298.1 291.5 3219 343.3 102.5 102.2 345.1 369.3 196.3 2973 201.3 381.0 356.5 421.6 257.1 J u lr. ... 299.3 291.2 323.2 345.1 103.1 102.7 346.1 370.7 197.3 298.7 201.7 382.5 358.5 423.2 258.4 Auc- — 3003 291.7 324.1 346.3 103.6 103.0 347.9 371.9 197.9 300.8 203.3 383.7 360.3 424.6 259.7 Sept. .... 301.9 292.3 325.3 348.0 104.2 103.5 346.6 372.9 198.2 .702.8 204.9 383.7 361.7 425.1 260.9 O ct.. ... 302.6 293.5 326.0 349.3 104.6 103.9 351.1 372.2 198.2 304.4 205.5 383.2 362.9 424.2 262.2 s v . .... 303.1 294.1 327.4 350.7 105.0 104.3 353.4 374.4 198.5 305.5 205.3 381.6 364.7 424.5 263.5 Dec.. ... 303.5 295.4 328.1 351.8 105.3 104.5 354.7 373.8 198.5 306.1 205.7 379.8 366.0 423.3 264.2 19*4: Jan...... 305.2 300.2 329.6 353.0 105.7 104.9 356.7 378.2 199.0 306.7 205.6 375.9 368.6 421.7 265.8 Feb.. ..., 306.6 302.2 331.1 353.8 106.0 105.1 353.5 384.8 !9«.5 306.6 206.4 370.5 371.5 422.7 266.7 Mar . ..., 307.3 301.8 331.2 355.3 106.5 105.6 355.3 380.9 198.6 309.4 207.4 374.0 3735 I 421.8 267.H Apr...... ; 308.9 301.7 3338 357 6 107.4 106.2 356.3 383.9 1985 311.2 207.6 375.4 375.3 424.7 269.1 * ItrM n ivw M ■* i N u n .—JC5A data m M w iw m B t * rw^wWf it*;, im : tt? N tiw h r «4 wrWa rtHrr Im i «it br ««|» * W Im J m i t r m >nn«c «f t f v l m d p « U I * Fv4 «d. —•I ud MrtW pr. p« ifipil) , . Malar *4, ra d u l. Dili WfMMnc IW w w pm i i rratai tm t hr iM M tw nbf ««• w M d If ik tn im i m t «wtfr im u fillr m l fiprri 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX B JOHNSTONE (1981) SELECTED FORMULAS Gini Coefficient This statistic may best be described by reference to the attached diagram of the Lorenz curve. The coefficient ratios the area between the curve and the line of equality, to the total area under the line of equality. The formula is: n G = I (pi - 1qi - piqi - 1) i=1 where pi = the cumulative proportion of the characteristic whose equality is being investigated, and qi= the cumulative proportion of the variable which is acting as the criterion for the measurement. The values for the two curves in the diagram are: El S alvad o r G = 0.31 Nicaragua G = 0.41 Source: Johnstone (1981, pp. 93-94; 280-284). 171 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 172 60 100 Cumulative percentage of students enrolled Fig. 4. Lorenz curves measuring inequality of distribution of enrollments across grades at the first-level— El Salvador and Nicaragua, 1965 (Johnston, 1981, p. 93). T A B L E 52 PERCENTAGE DISTRIBUTION OF FIRST-LEVEL STUDENTS BY GRADE— EL SALVADOR AND NICARAGUA, 1965 G rade El Salvador Nicaragua 1 37 45 2 20 20 3 15 14 4 12 10 5 n 6 6 7 5 Source: UNESCO Statistical Yearbook, 1974, p. 192. The interpretation really implied by figure 4 is that, for El Salvador, 37 percent of the first-level enrollments receive 17 percent of the education offered at that level (i.e ., first grade), 57 percent of the total enrollment receives 33 percent of the total education offered (i.e ., first and second grades) and so on. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 173 T h e C Index 1. Calculates percentages of students in each educational band. 2. Substitutes values into the equation below. The frequently used measure of concentration is that known as the Herfindahl-Hirschman index. It is very simply defined as: k 2 C = z P i i=1 where Pi = proportion of total observations in the ith category, and k = total number of categories. If one category contains all observations, the C measure of concentration takes the value of 1.0. If all k categories in a defined scale have an equal proportion of the total amount— so no point of concentration exists—the C measure takes the value 1/k. When there are many categories, C is approximately 0.0 therefore. 3. The resulting value is interpretable in terms of whether there is a concentration of students in one band or whether enrollments are fairly evenly distributed across all bands. A high C value does not, however, show where the concentration exists. Reference would have to be made to th e basic d is trib u tio n s . Source: Johnstone (1981, pp. 97, 154). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 174 Expected Number of Years of Education to Be Received by an E n tra n t Expected number of years 1 ______P I2 ______of education to be = (1 - r ^ ) + (1 - r-j-j) (1 - rjj) received by entrant P 23 + (1 -r n )(l - r 22) (1 - r 33) + . . (Equation 3.15) The formula continues to develop as in the model evolving above. For six grades, the last term to be summed would be: ______PI 2 P23 P34 P45 P56 ______(1 - r n )(1 - r 22)(1 - r 33)(1 - r^M l - r 55)(1 - r g6) Substituting into Equation 3.15 can be time-consuming. When moderately sophisticated computing facilities are available, it is more efficient to use elementary matrix algebra. Note: This is based on the Continuation Coefficient expressed as: continuation = proportion of students promoted from grade i coefficient ______to g ra d e j______proportion of students not repeating grade i = Pii 1 - r-. ii Source: Johnstone (1981, pp. 109-110). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 175 Expected Number of Grades of Education Completed by an E n tra n t The formula for the number of grades expected to be completed by a grade 1 entrant is: number of grades pi2 + P 12 P23 expected to be = (1 - r n ) (1 - r n h l - r 22) completed by an e n tra n t P I 2 P23 P34______(1 - rn ) (l - r22) [ 1 - r23) (Equation 3.17) Source: Johnstone (1981, p. 112) P ro p o rtio n o f E n tran ts Who F in a lly G rad u ate The formula for the calculation of values is: proportion of entrants who g rad u ate = P I 2 P23 P3U Pg 1 " r 11 1 " r 22 1 " r 33 1 - r gg (Equation 3.19) where the symbols are as for formula 3.19 and 'g' is the last grade in the le v e l. Source: Johnstone (1981, p. 113). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 176 The Guttman Scale The Guttman Scale Analysis involves two stages: 1. It identifies whether the set of nominated system charac teristics forms a hierarchy. 2. It orders the systems into the identified hierarchy. An example is provided by Farrell ( 1969) in the attached d ia g ra m . Source: Johnstone (1981, pp. 134-136). Laspeyres Formula for Education Cost index for levels 1 - 3 is expressed as follows: 3 Cl = I E.. C.„ ib . it i=1 E -u-C *. ib ib i=1 where Cl = education cost index in year t; = enrollment at level i in the base year b; and C.^ = per student cost for level i in year t. When the pupil costs do not change from those operating in the base year, or they change at the different levels by amounts proportional to the base year enrollments at each level, the index number which is calculated is 100. The weakness is that when there is a fall in demand the Laspeyres formula will overstate price increases. Source: Johnstone (1981, p. 105). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. erdcdwt priso o h cprgt we. ute erdcin rhbtd ihu permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced FARRELL'S SCALOGRAM OF EDUCATIONAL STRUCTURAL DIFFERENTIATION: 49 NATIONS, 1960 oiia Republic Dominican R. .5 C. uig ny tm wt * miuu Data Ambiguous * with items only using . .R C .95; = . .R C 0 ecn i mdl aeoy .92 = category model in percent 80 hn (Taiwan) China ad Arabia Saudi Philippines Afghanistan ot Korea South l Salvador El ot Rica Costa Venezuela Nicaragua Gua tema 1 a Indonesia Colombia Pakistan Ethiopia Thailand Paraguay Cambodia Honduras Morocco Tunisia olivia B Lebanon Uruguay Eucador beria ib L Mexico Ceylon Malaya razil B Guinea Jordan Panama Chile Nepal Syria Ghana Lybia Egypt Sudan Yemen aiti H Burma India Peru Laos Iraq Iran — I 1 II li (A V) *5■5 S § 177 11 = -3 ■= - 5 i- c c 01 U U V _ . KXXXXXXXOX o o o X « w w t w ? « I w I \ I § § »• o o ■ 3 5 > 3 > 3 -5 -3 ?■ ! • ~ ~ ~ ? • ! • • • 2 2 2 2 2 J | 2 - * • • d u u £ 1 £ & * * & i w • - • • • C. Z J . C . C o. a — m a = 5 B> e 5 -= Is 5 1 xI a x I I - • £ w • 2 * 2 ■ * _ _ H 1 3 / 3 i/l * * 5 a - I ? i 5 i 2 Source: Farrell ( 1969, p. 309), illustrated in J. Johnstone. Indicators of Education Systems. Paris: UNESCO, International Institute for Educational Planning, 1981, p. 135. (ERIC Document Reproduction Service No. Ed 212 091). 178 Selected Indicator Formulas from UNESCO (1~974T 1. Equality of Acess I = (a .b ) ru ra l (a.b) urban where a = first grade enrollment ratio for new enrollments, and b = first level education completion rate. UNESCO (1974) points out that it is (theoretically) possible that rural education might be more freely attended than urban education, in which case the value for this indicator might exceed 1. Should such an occasion arise, UNESCO suggests inverting the value. Values less than 0.5 form a "red zone" requiring remedial action (Johnstone, 1981, p. 144). 2. Minimum Educational Attainment: regarded as being four years of first level education. where N^ = the proportion of the total population of an age theoretically corresponding to grade 4 of first-level which is enrolled in grade 4 or above in year t, and z = a dummy variable to ensure the value for I is between 0.0 and 1.0. "There seems to be reasonable consensus that four' grades of primary education represents the true requisite for achieving basic literacy" (Johnstone, 1981, p. 146). 3. Availability of Qualified Teachers: measures availability of trained nationals as against expatriates in school system. I = ( Ul (1 - s ^)) . ( u 2 (1 - s2)) where Uj = proportion of teachers with qualifications corresponding to standard requirements or higher for ith level of education, and s. = proportion of teaching force at the ith level who are expatriates. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 179 Weakness: indicator makes assumptions that distributions of qualifications of nationals and expatriates are identical. Johnstone (1981, p. 158) suggests modifications—a division of the teacher qualification spectrum into unqualified, semiqualified,and qualified. 4. Educational Flow Differential: measures differences in flow patterns between males and females. (a.b.c.d) females (a.b.c.d) males where a = first-time first grade enrollment ratio; b = first level education completion rate; c = transfer rate from first to second level; and d = second level education completion rate. To ensure a value between 0.0 and 1.0 UNESCO (1974) suggests that when the value for males is smaller than that for females the inverse of the value calculated should be used. Weaknesses include (1) confusion in cross-national interpretations and ( 2) examination of gross flows only (Johnstone, 1981, p. 160). 5. Educational Profile of the Unemployed: indicator showing congruence between the education system and the labor force. I = R. " 1 where R j = unemployment rate of the population aged 15-64 with education level i, and R^ = national average of unemployment rate of population 15-64. The dummy variable z ensures that the value for I lies between 0.0 and 1.0 by reducing all values for 1 above 1.0 to the maximum value. Johnstone (1981, p. 160) shows that the indicator is limited in scope as well as being insensitive to differential unemployment rates at various educational le v e ls . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX C MCGUIRE-WHITE SOCIAL CLASS INDEX The McGuire-White Index of Socioeconomic Status is divided into three areas—occupation of family head, source of the family income, and education of the family head. The first rating, the rating of occupation, is presented in table 54 . If the respondent reports the family head is president of a corporation, the occupation is g iv e n a ra tin g o f " 1"; if the respondent reports the family head to be a high-school teacher, the occupation is given a rating of " 2"; if the respondent reports the family head to be a truck driver, the occupation is given a rating of " 6," etc . Whatever the respondent checks as being the primary source of income is simply recorded as that number. For example, if the respondent checks profits and royalties on the questionnaire, the score o f " 3" is re c o rd e d . The same is true for education. If the respondent indicates the family head completed graduate work, a score of " 1" is recorded. To calculate the total socioeconomic score for each subject, the rating for occupation is multiplied by 5; the rating for source of income is multiplied by 4; the rating for education is multiplied by 3—then the three scores are added together for total score. The total score is checked against the listed five categories to see which level 180 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 181 the respondent is in. For example, if the total score is 15, it would fall in the upper-class category and the respondent would receive a socioeconomic code of "1." Category Socioeconomic Level Score (1) Upper Class 12-22 (2) Upper Middle 23-37 (3) Lower Middle 38-51 (4) Upper Lower 52-66 (5) Lower Lower 67-84 Questionnaire items 1. What is the occupation of the head of your family of orienta tion household (teacher, minister, etc)? ______. 2. what is the primary source of the family income? ______1. Inherited savings and investments ______2. Earned wealth, transferable investment ______3. Profits, royalties, fees ______4. Salary, Commissions (regular, monthly, or yearly) ______5. Hourly wages, weekly checks ______6. Odd jobs, seasonal work, private charity ______7. Public relief or charity 3. What is the highest educational attainment of the principal earner of the family income? ______1. Completed graduate work for a profession ______2. Graduated from a 4 year college ______3. Attended college or university for two or more years ______4. Graduated from high school ______5. Attended high school, completed grade 9, but did not graduate ______6. Completed grade 8, but did not attend beyond grade 9 ______7. Less than grade 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Farm People landowner who do not leased property who their property. Land o perators supervise who properties £ have an activel i f e . urban farms who "hire out" supervise directly Farm owners w ith "h ired supervises "squatters-nesters" helpi" operators of foreman! owners of Small landowner,of rented operator hiring property "hands" lished farmsubs'ce laborers farmers. Gentlemen farmer or Sharecroppers! estab Tenants on good farms Migrant workers t a i l o r , RR Policeman P o lice conduct. barber, prac.nurse, brake- C aptain, to r h elp er T axi £ trk qas s tn . a t . watchmaker d rlv e rs i w a ite r, w a itre s s , Domestic h e lp , bus boy, scrub men, men, e t a l. women, women, Ja n i Blue Collar Service Small contr. trician, RR engineer et al skilled trades who who works at or vises super his job Foreman! mstr carpenter, elec skilled wrks. repairm oni med. Apprentice to i TABLE 54 exec. sec. status of orqan. insurance, real estate, salesmen stock edit, writer auto salesman RR RR or t e l . tants to skilled tradei CPSi CPSi e d itoof r newspaper, Accountant! Bank c le rk , magazine! Semi-skilled factory and postal clerk, agt. or supv. Heavy la b o r, odd-job skilled mem workers production workersi assiswarehousemen, watchmen mine or m ill hands, un , grocery £ £ beauty executives, (Stenographer, bookkeeper! (Dime s to re c le rk s u t i l i t y turing agents small branches salesman of supervisors! ticket in agent, dept sales stores, people et al President, et al bank, public some some manufac Top of corporation, Manager o f Asst, office £ known mchdse. or buyers £ dept. mgrs. or clerks! telephone operators, et al. or MCGUIRE-WHITE SOCIAL CLASS INDEX--OCCUPAT IONS* ses valued $100,000 at from $10,000 $100,000 from $5,000 ProprietorsLarge busines Businessmen White Collar to $50,000 Dusiness value equity value to $10,000 from $2,000 $2,000 Business or Business or Business or to $5,000 equity value more depending on community at less than equity value at $50,000 to equity value "Reputed Lawbreakers" teacher, libra ineer, school professor supt. teacher, tered regis nurse, e t a l r ia n , and others physician, eng with 4-yeardeqree out 4-yr deqree minister with High school table represent revisions made after interviewing in communities and are types to guide other ratings. *For an original table, consult Warner's revised scale (12, pp. 140-141). Modifications in the present 1. Lawyer, judge 3. Grade school 2. Rate Professionals 7. 4 .” " 5. 6. Business or Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A P P E N D IX D PERFORMANCE OF ALL SCHOOLS IN CEE THREE PARISHES 1979 List of Schools in Kingston, St. Andrew, and St. Thomas Showing (1) Enrollments (Grades 4-6), September, 1978, (2) CEE Entries, (3) Common Entrance Awards, and (4) Awards as Ratios o t bnroiimeht (Grades 4-b), January, 1979 Column 4 over Schools Enrollment Entries Awards Column 2 Parish of Kingston Prim ary 1. All Saints' 94 22 3 .03 2. Allman Town 209 118 25 .12 3. A lpha 465 356 133 .29 4. Chetolah Park 396 141 61 -15 5. Clan Clarthy 358 156 32 .0 9 6. Denham Town 751 220 43 -06 7. Franklyn Town 372 225 42 .11 8 . North Street Congregational 115 80 24 .21 9. Ormsby Hall 140 45 6 -04 10. Rollington Town 379 270 50 .13 11. St. Alban's 254 122 22 -09 12. St. Andrew 319 168 21 .07 13. St. Anne's 289 75 17 -06 14. St. Aloysius 386 234 82 -21 183 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 184 Column 4 over Schools Enrollment Entries Awards Colum 2 Parish of Kingston Primary & Infant 15. Holy Family 282 88 24 -09 All Age 16. All Saints' 314 104 14 -04 17. Boys Town 113 ! - .00 18. Central Branch 394 205 31 .08 19. Elletson 637 161 16 .03 20. Norman Cardens 443 155 25 .06 21 . Rennock Lodge 110 11 - .00 22. St. Michael's 311 61 13 .04 23. Wesley 97 6 - .00 24. Windward Road 748 163 83 .11 All Age & Infant 25. Calabar 570 214 44 .08 26. St. George's Girls 270 225 77 .29 27. Port Royal 74 13 4 .05 'arish of St. Andrew Primary 1. August Town 151) 43 1 - .00 126) 2. Balmagie 668 176 40 .06 3. Duhaney Park 435) 649 149 .16 498) 4. Dun robin 478 428 152 .32 5. George Headley 206 107 14 .07 6. Greenwich 373 184 27 .07 7. Grove 97 11 - .00 8. Half-Way-Tree 385 277 41 .11 9. Hope Valley 169 70 34 .20 10. Iris Gelley 123 27 1 -01 11. Jones Town 454) 287 52 .06 385) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 185 Column 4 o ver Schools Enrollment E n trie s Awards Column 2 Parish of St. Andrew Primary - cont'd 12. Lawrence Tavern 417 170 46 .11 13. Maxfield Park 350) 323 29 .05 615 265) 14. Mona Heights 394 191 95 .24 15. Mountain View 332 199 17 .05 16. New Providence 448) 465 108 .12 913 465) 17. Pembroke Hall 485 329 139 .29 18. Rousseau 424 176 91 .21 19. St. Martin de Porres 140 46 3 .02 20. St. Patrick's 714 273 63 .09 21. St. Peter Claver 338 178 8 .02 22. St. Jude1 s 479 321 69 .14 23. Tarrant 520 237 123 .24 24. T rench Town 343 57 9 .03 Primary & Infant 25. Dupont 278 163 7 .03 26. Unity Primary 233 46 2 .01 All Age 27. Allman Hill 221 36 10 .05 28. Balcombe Drive 339 85 1 5 .04 29. Bito 14 3 - .00 30. Bloxburgh 31 2 1 .03 31. Bowden Hill 39 -- .00 32. Brandon Hill 91 4 - .00 33. Cavaliers 147 10 2 .01 34.Clifton 67 12 - .00 35. Cockburn Cardens 411 65 8 .02 36. Constant Spring 525 150 41 .08 37. Constitution Hill 86 14 2 .02 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 186 Colum n 4 o v e r Schools Enrollment Entries Awards Column 2 Parish of St. Andrew All Age - cont* d 38. Content Cap 63 19 - .00 39. Craighton 66 19 4 .06 40. Dallas 132 25 7 .05 41. Drews Avenue 395 178 27 .07 42. Essex Hall 117 13 - .00 43. Friendship Brook 59 8 - .00 44. Cordon Town 168 20 1 .01 69 18 .03 45. Greenwich Town 260) 562 302)562 46. Halls Delight 115 4 1 .01 47. Harbour View 565 391 106 .19 48. Hugh Sherlock 176 19 4 .02 49. King Weston 61 8 1 .02 50. Mannings Hill 121 6 - .00 51. Maverley 616 180 44 .07 52. Melrose 292 217 43 .15 53. Mico Practising 398 272 73 .18 54. Mt. Fletcher 170 17 2 .01 55. Mt. James 107 8 - .00 56. New Day 472 233 83 . 1 8 57. Padmore 58 4 - .00 58. Red Hills 166 18 6 .04 59. Rock Hall 266 76 17 .06 60. Seaward 467 195 44 .09 61. Somerset 58 11 - - 62. St. Richard's 339 239 115 . 34 63. St. Theresa's 121 5 1 .01 64. St. Benedict's 307 138 29 .09 65. Swallowfield 600 170 65 .11 66. Tavares Cardens 450 95 16 .04 67. Tower Hill 47 10 - .00 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 187 Column 4 over Schools Enrollment Entries Awards Column Parish of St. Andrew All Aqe - cont1 d 68. Westphalia 41 6 - .00 69. Whitfield 636 146 23 .04 70. Woodford 91 11 - .00 All Age & Infant 71 . Colden Valley 96 16 - .00 72. Jack1 s Hill 64 8 - .00 73. John Mills 652 266 55 .08 74. New Carden 50 - - .00 75. Shortwood Practising 245 224 42 .17 76. St. Francis 332 222 46 .14 77. Stony Hill 382 76 9 .02 Parish of St. Thomas Primary 1. Barking Lodge 57 6 1 .02 2. Dalvey 86 21 1 .01 3. Duckenfield 299 79 33 .11 4. Font Hill 78 16 7 .09 5. Hillside 44 2 - .00 6. Lystra 98 6 2 .02 7. Middleton 157 32 13 .08 8. Mt. Vernon 40 4 - .00 9. Prospect 36 10 2 .06 10. Seaforth 412 50 13 .03 - * Cp. in^ Carden 76 19 3 .04 12. T rinityville 252 45 9 .04 13. White Hall 173 16 - .00 14. Winchester 45 7 - .00 15. Yallahs 303 107 40 .13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 188 Column 4 o v e r Schools Enrollment Entries Awards Column 2 Irish of St. Thomas Primary £ Infant 16. Amity Hall 36 10 1 .0 3 17. Colden Grove 141 25 13 .0 9 18. Grants Pen 161 33 4 .02 All Age 19. Aelous Valley 157 34 10 .06 20. Airy Castle 121 15 1 .01 21. Arcadia 100 10 - .00 22. Bath 288 37 10 .03 23. Bethesda 95 19 1 .01 24. Bull Bay 210 24 1 .00 25. Easington 129 18 3 .02 26. Hayfield 47 8 - .00 27. Johns Town 64 12 - .00 28. Johnson Mt. 47 2 - .00 29. Lyssons 324 120 43 .13 30. Minto 90 21 1 .01 31. Morant Bay 670 188 70 .10 32. Mt. Felix 29 2 - .00 33. Old Pera 37 7 - .00 34. Penlyne Castle 55 9 1 .02 35. Port Morant 230 18 11 .05 36. Richmond Gap 48 14 2 .04 37. Rowlandsfield 35 8 - .00 38. Thornton 94 12 - -00 39. White Horses 168 15 - -00 40. Wilmington 74 11 1 .01 41. Woburn Lawn 127 17 - -00 All Age S Infant 42. Cedar Valley 192 30 5 -03 43. Pear Tree River 72 7 5 .07 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A P P E N D IX E PERFORMANCE OF ALL SCHOOLS IN CEE THREE PARISHES 1980 List of Schools in Kingston, St. Andrew, and St. Thomas Showing (1) Enrollments (Grades 4-6), September, 1979, (2) CEE Entries, (3) Common Entrance Awards, and (4) Awards as Ratios of Enrollment [OraggTTR 6),'----- January, 1980 Column 4 over Schools Enrollment Entries Awards Column _ 2 Parish of Kingston Primary * 1. All Saints' 80 22 - 00 2. Allman Town 353 126 13 04 3. Alpha 594 215 45 08 4. Chetolah Park 566 1 58 55 10 5. Clan Clarty 408 158 30 07 6. Denham Town 798 217 41 05 7. Franklyn Town 378 167 28 07 8. North Street Congregational 177 57 15 08 9. Ormsby Hall 165 45 6 04 10. Rollington Town 487 241 59 12 11. St. Alban's 328 123 18 05 12. St. Andrew 354 149 18 05 13. St. Anne's 394 97 24 06 14. St. Aloysius 496 237 76 15 15. Alpha II 210 192 74 35 * School classifications changed during 1979/80. 189 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 190 Column 4 o v e r Schools Enrollment Entries Awards Column 2 Parish of Kingston Primary S Infant 16. Holy Family 407 90 25 .06 All A ge 17. A ll Saints 297 101 15 .05 18. Boys Town 43 4 1 .02 19. Central Branch 392 180 40 .10 20. Elletson 465 122 7 .02 21. Norman Cardens 437 154 21 .05 22. Rennock Lodge 104 11 - .00 23. St. Michael 1 s 508 98 7 .01 24. Windward Road 625 173 77 .12 All Age & Infant 25. C alabar 609 204 57 .0 9 26. St. George's Girls 298 197 62 .21 27. P o rt Royal 70 19 2 .03 arish of St. Andrew P rim ary 1. August Town 235 38 5 .02 2. Balmagie 610 167 45 .07 3. Duhaney Park 481) 674 134 .16 361) 4. Dunrobin 599 420 137 .23 5. Excelsior 35 29 10 .2 9 6. G eorge Headley 260 193 33 .1 3 7. Greenwich 377 158 36 .10 8 . G rove 73 11 - .00 9. H a lf-W a y -T re e 385 268 53 .14 10. Hope Valley Experimental 185 109 35 .19 11. Iris Gelley 120 30 2 .02 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 191 Column 4 o v e r Schools Enrollment Entries Awards Column 2 Parish of St. Andrew Primary - cont 1 d 12. Jones Town 740 281 56 .08 13. Lawrence Tavern 422 116 19 .05 14. Lister Mair Cilbey 1 1 15. Maxfield Park 547 262 53 .10 16. Mona Heights 452 204 56 .12 17. Mountain View 317 166 16 .05 18. New Providence 432 343 112 .26 19. Pembroke Hall 420 327 126 .30 20. Rosseau 616 171 83 .13 21. St. Martin de Porres 177 40 11 .06 22. St. Patrick's 558 285 56 .10 23. St. Peter Claver 334 198 11 .03 24. S t. Ju d e's 470 343 89 .1 9 25. T a r r a n t 468 225 85 .18 26. Trench Town 292 53 20 .07 27. D up o n t 332 181 7 .0 2 Primary & Infant 28. U n ity 205 60 5 .02 All A ge 29. Allman Hill 220 57 6 .03 30. Balcombe Drive 306 64 16 .05 31. B ito 12 5 - .00 32. B lo xb u rg h 32 1 - .00 33. Bowden Hill 48 5 - .00 34. Brandon Hill 87 12 - .00 35. C a v a lie rs 127 10 1 .01 36. Cockburn Gardens 593 73 15 .31 37. Constant Spring 498 183 21 .0 4 38. Constitution Hill 72 8 - .00 39. Content Cap 65 17 - .00 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 192 Column 4 o v e r Schools Enrollment Entries Awards Column 2 Parish of St. Andrew All Age - cont1 d 40. Craighton 61 18 1 .02 41. Dallas 168 20 9 .05 42. Drews Avenue 373 197 26 .07 43. Essex Hall 97 16 3 .03 44. Friendship Brook 66 9 - .00 45. Cordon Town 1 54 11 2 .01 46. Greenwich Town 485 67 6 .01 47. Halls Delight 95 10 - .00 48. Harbour View 565 383 73 .13 49. Hugh Sherlock 100 20 - .00 50. King Weston 67 20 2 .03 51. Mannings Hill 111 16 - .00 52. Maverley 586 156 32 . 05 53. Melrose 542 207 49 .09 54. Mico Practising 379 267 71 .19 55. Mt. Fletcher 1 98 18 2 .01 56. Mt. James 116 8 - .00 57. New Day 649 257 65 .10 58. Padmore 53 6 1 .02 59. Red Hills 173 30 10 .06 60. Rock Hall 369 79 13 .04 61. Seaward 766 171 38 .05 62. Somerset 58 16 1 .02 63. St. Richard's 359 224 122 .34 64. St. Theresa's 117 8 1 .01 65. St. Benedict's 308 95 12 .04 66. Swallowfield 557 212 54 .10 67. Tavares Cardens 362 111 7 .02 68. Westphalia 32 4 - .00 69. Whitfield 528 123 14 .03 70. Woodford 102 15 - .00 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 193 Column 4 o v e r Schools Enrollment Entries Awards Column 2 Parish of St. Andrew All Age S Infant 71. Golden Valley 91 8 1 .01 72. Jacks Hill 59 6 1 .02 73. John Mills 719 310 46 .06 74. Shortwood Practising 304 233 44 .14 75. St. Francis 265 225 35 .13 76. Stony Hill 452 56 12 .03 Parish of St. Thomas Primary 1. Barking Lodge 53 8 - .00 2. Dalvey 90 18 5 .06 3. Duckenfield 275 78 26 .09 4. Font Hill 69 23 - .00 5. Hillside 35 4 - .00 6. Lystra 55 5 - .00 7. Middleton 139 45 7 .05 8. Mt. Vernon 32 2 - .00 9. Prospect 64 13 2 .03 10. Seaforth 326 55 12 .04 11. Spring Garden 60 18 1 .02 12. Trinity ville 230 43 11 .05 13. Whitehall 175 20 1 .02 14. Winchester 49 7 - .00 15. Yallahs 462 129 4£ .11 Primary & Infant 16. Amity Hall 30 13 - .00 17. Golden Grove 153 36 15 .10 18. Crant's Pen 136 26 5 .04 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 194 Column 4 over Schools Enrollm ent E n trie s Aw ards Column 2 Parish of St. Thomas All Age 19. Aelous Valley 142 18 1 .01 20. Airy Castle 98 23 4 .0 4 21. A rc a d ia 76 7 1 .01 22. B ath 265 46 5 .02 23. B ethesda 73 8 - .00 24. B u ll B ay 148 21 8 .05 25. E asington 112 24 3 .0 3 26. H a y fie ld 41 6 - .00 27. Johns Town 62 10 4 .06 28. Johnson Mt. 31 6 - .00 29. Lyssons 287 127 27 .0 9 30. M into 76 12 - .00 31 . Morant Bay 623 163 50 .0 8 32. Mt. Felix 37 6 - .00 33. O ld Pera 37 4 - .00 34. Penlyne Castle 57 9 1 .02 35. Port Morant 287 16 6 .02 36. Richmond Cap 55 16 - .00 37. Rowlandsfield 79 7 - .00 38. Thornton 81 25 - .00 39. White Horses 110 10 - .00 40. W ilm ington 49 11 - .00 41 . W oburn Lawn 125 15 1 .01 42. C e d a r Valley 175 43 4 .02 43. Pear Tree River 47 8 3 .06 Notes 1. Schools not submitting entrants were: Clifton, New Carden, Tower Hill, and Wesley. 2. Additional entries not included above = 16. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A P P E N D IX F T A B L E 55 CEE AWARDS 1979 AND 1980 AS PERCENTAGE OF PUPILS ENROLLED IN GRADES 4-6—TEN BEST PERFORMING URBAN SCHOOLS Percentages School 1979 1980 St. Richard's Primary 34 34 Dunrobin Primary 32 23 Pembroke Hall Primary 29 30 St. George's Girls All-Age 6 In fa n t 29 21 A lpha II 15 Tarrant Primary 24 18 St. Aloysius Primary 21 15 Hope Valley Primary 6 Infant 20 19 New Providence Primary 12 26 Mico Practicing All-Age 18 19 * A djusted T A B L E 56 CEE AWARDS 1979 AND 1980 AS PERCENTAGE OF PUPILS ENROLLED IN GRADES 4-6—TEN BEST PERFORMING RURAL SCHOOLS Percentages School 1979 1980 Dallas All-Age 5 5 Lawrence Tavern Primary 11 5 Red Hills All-Age 4 6 Dalvey Primary 1 6 Duckenfield Primary 11 9 Golden Grove Primary 6 In fan t 9 10 Johns Town All-Age 0 6 Lyssons All-Age 13 9 Middleton Primary 8 5 Trinity ville Primary 4 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX C TABLE 57 DETAILS OF SCHOOL SAMPLE DRAWN, URBAN (MOST AWARDS) Urban Schools Selected for Analysis with the Total Number of Students (1) Enrolled in Grades 4-6 in 1979/80, (2) Entered in the Common Entrance Examination, 1980, (3) Securing Awards, and and (4) Selected for Sample School Enrollment Entries Awards Sample St. Richard's 359 224 122 44 Dunrobin 599 420 137 53 Pembroke Hall 420 327 126 47 St. George's Girls 298 197 62 44 Alpha II 210 192 74 42 Tarrant 468 225 85 45 St. Aloysius 496 237 76 44 Hope Valley 185 109 35 35 New Providence 432 343 112 67 Mico Practising 379 26 7 71 45 Total 3, 846 2,541 900 466 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 197 T A B L E 58 DETAILS OF SCHOOL SAMPLE DRAWN, URBAN (FEWEST AWARDS) Urban Schools Selected for Analysis with (1) Nos, Enrolled C rad es 4 -6 in 1979/80 (2 ) Nos. of E n trie s in Common Entrance Examinations, 1980, (3) Nos. of Awards and (4) Sample Selected School Enrollment Entries Awards Sample Cordon Town 154 11 2 8 Allman Town 353 126 13 45 Dupont 332 181 7 31 B o y 1 s Town 43 4 1 4 Tavares Garden 362 111 7 26 B a (combe D rive 306 64 16 30 St. Peter Claver 334 198 11 33 M anning 1 s Hill 111 16 - 8 M ountain View 317 166 16 35 Seaward 766 171 38 38 Total 3, 078 1,048 111 263 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX H TABLE 59 DETAILS OF SCHOOL SAMPLE DRAWN, RURAL (MOST AWARDS) Rural Schools Selected for Analysis with (1) Nos. Enrolled Crades 4-6 in 1 979/80, (2) Nos. of Entries in Common Entrance Examination, 1980, (3) Nos. of Awards and (4) Sample Selected School Enrollment Entries Awards Sample Duckenfield 275 78 26 37 Golden Grove 153 36 1 5 17 Lyssons 287 127 27 58 Johns Town 62 10 4 5 Red Hills 173 30 10 14 Lawrence Tavern 422 116 19 55 Dallas 168 20 9 9 Middleton 139 45 7 20 Dalvey 90 18 5 7 T rinityville 230 43 11 21 Total 1,999 523 133 243 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 199 T A B L E 60 DETAILS OF SCHOOL SAMPLE DRAWN, RURAL (FEWEST AWARDS) Rural School Selected for Analysis with (1) Nos. Enrolled Crades 4-6 in 1979/80, (2) Nos. of Entries in Common Entrance Examination, 1980, (3) Nos. of Awards and (4) Sample Selected School Enrollment E ntries Aw ards Sample H a y fie ld 41 6 - 5 B ethesda 73 8 - 7 G rove 73 11 - 9 Halls Delight 95 10 - 8 M t. James 116 8 - 6 T h o rn to n 81 25 - 19 M t. F e lix 37 6 - 5 Bowden Hill 48 5 - 4 W estphalia 32 4 - 5 White Horses 110 10 - 8 Total 706 93 - 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ■St. akt C tth . u n k L D at irthday ctitaxht o tal B kavk * T L D Act Act don if u n CUrcm ( D t t a c h id A ulh kak m T Farentt Stgnaturt Y incb ast M en* in c i rheeter S E l m utt L S to St. ii G ilt. “ w 1 j M itt. | " a y t t t n d in o Hen. We*. 4. D A T C O filK IH t t in d in o D i m i h c a t i A A C ttindco y ) h o w A H o w CU ir t h B fa Q u a S From Pmmnr School) F ro m P ju rc n t S c h o o l) St. Jamc* T r*l*w n ok chool c h o o l c ho o l S S . S B ov a I J. S U U IP 7. Occupation of Pakent/Ouaildun It, It, 9b. Soiool Arm voio up to Lajt Tom (ip Dirm iNT 9 a . * a a a i a St. A na St. M a ry V V - J. Sii? h u k T Sl. o n t h o n t h F o n .f Th om . M M a y a y D D SSi APPENDIX I Ljh (SLOCK LETfUtS) ■p o SI ni O t S TAE A •CHOUAO*HIP OH PIIKK PLACK MAT HOT OK HK>KNTCHKO r a v i L Of 3 A ndrew l S ; fi-Sls •3 w h o f i COMMON ENTRANCE SCORE CARDS a t i ati COMMON ENTRANCE EXAMINATION ENTRY FORM ENTRY EXAMINATION ENTRANCE COMMON D D KUtploa end .) a n d and C liru > ktc c ic .) a ., ., .NO T! THK HOLDCR u pair E iE .-i-it r u f . , , , , AmM>u AmM>u ' <.d|i3U £5 it i i n a n u io a t i C pr im a r y primary , , , ___ Saieou AntM iro ant c h o o l ipilut ipilut ? ? F o rm H o . 2a S in in f a n t 4 ( ( (See overleaf) rjvio ur Pervious Scxoou P r im a r y (b) Either i» now in Pnmary School or » u in Pnoury School up to La*1 Term (End ofSummer Tens), («) He* aurnJcJ Primary School for at ka il two yc*n; end M u k School Preferred . Schooli Picfcrred . Schooli I. School Code No, 1 0 2. CANDIDATE'S SUltNAME6, AND CHKiSlUN NAME AND NAM ADDRESS OF I.’ PARENT/Gl'ARDIAN (TO Jik COMl'LElED DY ‘ PRIMARY SCHOOL CAM)u5aTES) U . RtAiOM FOR FlPTT PfJIlXlN C l I l l PRIFCAINCB 9. TO DE COMPLETED UY OTHER CANDIDATES 12. 12. NAME OF NEAREST EXAMINATION CENT.IE « A P 2b J2b P u n Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A P P E N D IX J SOCIOECONOMIC STATUS—CATEGORY 1 LOWEST INCOMES—PRIMARY EDUCATION ONLY 1. B arm aid 28. Laundress 2. Bar Tender 29. Messenger 3. Caretaker 30. Maid 4. Casual Worker 31. Meter Reader 5. Char Woman 32. Meter Repairman 6. C h e f 33. Mosquito Eradication Worker 7. C lean er 34. Office Assistant 8 . Clinic Attendant 35. Office Attendant 9. Cook 36. Office Helper 10. D ay's W orker 37. Office Maid 11. Delivery Man 38. Pants Renovator 12. Domestic Worker 39. Peddler 13. Female Attendant 40. P orter 14. Fisherman 41. Pump Operator 15. Fish V e n d o r 42. Self-Employed 16. G u ard 43. Shop Assistant 17. H elp er 44. Sprayman,Ministry of Health 18. Higgler 45. Surveyor Help 19. Hospital Attendant 46. Time Keeper 20. Hospital Cook 47. Unemployed 21. Hospital Helper 48. Unskilled Laborer 22. Hospital Orderly 49. Upholsterer 50. Vendor 23. Household Helper 24. Impact Worker 51. Waitress 25. Janitor 52. Ward Assistant 26. Lab Attendant 53. Ward Maid 27. Laborer 54. Watchman SOCIOECONOMIC STATUS—CATEGORY 2 LOWER MIDDLE AND MIDDLE-MIDDLE INCOMES—SECONDARY EDUCATION LEVELS 1. A rtis t 5. Basic School Teacher 2. Assistant Nurse 6. Beautician 3. Assistant Security Officer 7. B inder 4. B a ker 8 . Brakeman, Jamaica Railway Corporation 201 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 202 9. Bus Conductress 47. Motor Mechanic 10. B us D riv e r 48. Needle Worker 11. B u tc h e r 49. Nurses Aide 12. Cabinet Maker 50. Office Clerk 13. Canteen Operator 51. Operator 14. C a rp e n te r 52. Painter 15. C a s h ie r 53. Plumber 16. Chauffeur 54. Police Constable 17. Clerical Officer 55. Police Corporal 18. C le rk 56. Policeman 19. Confectioner 57. Postal Clerk 20. Counter Clerk 58. Postal Employee 21. Craft Worker 59. Postman 22. D e c o ra to r 60. Printer 23. Dressmaker/Seamstress 61. Psychiatric Aide 24. D r iv e r 62. Restaurant Operator 25. Electrician 63. Sales Clerk 26. F a b ric a to r 64. Salesman 27. Factory Worker 65. Saleswoman 28. F a rm e r 66. Security Guard 29. Field Assistant—Mosquito Control 67. Service Manager 30. Firem an 68 . Shoe Maker 31. Fruit Inspector 69. Shop Keeper 32. G ro c e r 70. Store Clerk 33. Headman 71. Store Employee 34. Headman (Kingston and St. Andrew 72. Student Corporation) 73. Student Typist 35. Hom eguard 74. Supervisor/Corporation Workers 36. Housekeeper 75. Tally Clerk 37. Housewife 76. Taxi Operator 38. Hotel Worker 77. Technician 39. IBM Key Punch Operator 78. Telegraph Clerk 40. Insurance Clerk 79.- T e le g ra p h O perator 41. Jamaica Public Service Worker 80. Telephone Operator 42. L an d la d y 81. Telephone Technician 43. Line Supervisor (Agricultural 82. Trailer Driver Marketing Corporation) 83. Transport Operator 44. Machine Operator 84. Truck Operator 45. Mason 85. T y p is t 46. Mechanic 86 . W elder Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 0 3 SOCIOECONOMIC STATUS—CATEGORY 3 UPPER-MIDDLE AND LOWER UPPER CLASS INCOMES—WELL TRAINED PERSONNEL 1. A ccountant 28. Machine Shop Engineer 2. Advertising Agent 29. Manageress 3. Agricultural Officer 30. Medical Representative 4. Assistant Service Manager 31. Minister of Religion 5. Bank Clerk 32. Nurse 6. B u ild e r 33. Planter 7. Businessman 34. Price Investigator 8 . Civil Servant 35. Public Health Inspector 9. Contractor 36. Rehabilitation Officer 10. Correctional Officer 37. R e g is tra r 11. Crop Control Officer 38. Salesman 12. Detective 39. Sales Manager 13. Dress D esigner 40. Sales Representative 14. Druggist 41. Sales Supervisor 15. Elevator Engineer 42. S ecretary 16. Field Officer - JAMAL 43. Sergeant of Police 17. F lo ris t 44. Service Manager 18. Foster Parent Home Supervisor 45. State Registered Nurse 19. G arm ent M an u factu rer 46. Stenographer 20. Internal Auditor 47. Stenotype— writer 21. JAMAL Area Officer 48. Statistician Ministry of Labour 22. Jeweller 49. Supervisor 23. Lab Technician 50. Supervisor JTC 24. Lecturer 51. Supervisor Telephone Operator 25. Librarian 52. Teacher 26. Life Underwriter 53. Travel Consultant 27. Lithographer 54. Verbatim Reporter JAMAL = Jamaica Movement for the Advancement of Literacy JTC = Jamaica Telephone Company Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. erdcd ih emsin fte oyih onr Frhrrpouto poiie ihu permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced EXAMINATION CLASS LIST ( Illustration) eieH oj g ja jo a -euimexH s| e epueis u p je p ;e tse| o ja u a6v 6 a ui sjeaA uo Date jg jsen jjig q p e X of u j aea aje jo -ug m OLD SJB3A |OOL|DS 0 uoi;euiuiex Buipjoooe } Admission 10 V S S oe v u a aouep u uoissiiupv in a SUOISS9S NAMES This First UOI* School School 1 3 J N C m in C APPENDIX L SAMPLE OF QUESTIONNAIRE WITH ACCOMPANYING LETTER Dear Teacher: You have often thought, no doubt, of the many areas of educa tional operations in our island which either describe the health of our educational systems or are directly related to good or poor academic performance. These would be critical factors or indicators to be watched and possibly controlled. The question arises as to which are important indicators and w hether there is agreem ent in research and among educators on these. I am currently doing a doctoral dissertation on just such a problem, trying to isolate a set of indicators of the inputs into primary education which relate to performance. This is being done, first by study of the Common Entrance results of 1 980, alongside certain school and pupil variables. Next, I wish to secure the opinions of teachers as to what they consider to be significant factors in primary education, both in terms of causal and descriptive factors/indicators. Would you kindly assist me in this study by completing the attached form? First, please rate the importance of the indicators in the attached list on a five point scale from 1-5 graduated as follows: 1. = Extrem ely important 2. = Very important 3. = O f medium importance 4. = of less than medium importance 5. = O f little importance Next, please add in the space provided, any important indicators which I may have omitted, and similarly rate these. Your input is critical to this study so please oblige by returning this form at your earliest. A stamped self-addressed envelope is enclosed. This study had the approval of the Ministry of Education, as proof of which it has been countersigned by the Assistant Chief Education Officer (Primary Education' . 205 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 206 I look forward to an early response. Yours gratefully. Jean Marshalleck c/o Ministry of Finance 6 Planning 30 National Heroes Circle, Kingston 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 207 RATINC OF PERFORMANCE INDICATORS Primary School System in Jamaica PLEASE PROVIDE YOUR RATINC OF THE FOLLOWING MEASURES OF EDUCATIONAL PERFORMANCE BY ASSIGNING A NUMBER AS GIVEN IN THE SCALE BELOW e.g. Extremely Very Medium Less than Little =1 =2 =3 Medium=4 =5 High Performance in Common Entrance Examination 2 INDICATORS IMPORTANCE - RATING SCALE Extremely Very Medium Less than Little =1 =2 =3 Medium=4 =5 Ratio Enrollment to School Age Population a) By Parish b) For the Island 2 . Degree of Over crowding by School Percentage Over crowded Schools a) By Parish 4 b) For the Island 5 Teacher/Pupil Ratio-T rained Teachers a) By School 6 b) By Parish 7 5. Teacher/Pupil Ratio Overall a) By School b) By Parish Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 208 INDICATORS IMPORTANCE - RATINC SCALE Extremely Very Medium Less than Little =1 =2 =3 Medium=4 =5 6. Percentage Teacher T u rn -O v e r Annually a) By School 10 b) By Parish 11 c) For the Island 12 7. Percentage Pupils Attending 90% of School Sessions Termly or Yearly a) By School 13 b) By Parish 14 c) For the Island 15 8. C rade by Crade Average Achieve ment in English a) By School 16 b) By Parish 17 c) For the Island 18 9. C rade by Crade Average Achieve ment in Mathe matics a) By School 19 b) By Parish 20 c) For the Island 21 10. Percentage of Students achiev ing Below 50% in English and Mathematics a) By School 22 11. Percentage Eli gible Age Croup Obtaining Passing M ark in Common Entrance Exami nations (not neces sarily awards) a) By School 23 b) By Parish 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 209 INDICATORS IMPORTANCE - RATINC SCALE Extrem ely V ery Medium Less than Little =1 =2 =3 Medium =4 =5 12. Na of L ib rary Books Per Pupil a) By School 25 1 3. No. of Hours of Educational T.V. Viewed a) By School 26 14. index o f Mal n u tritio n a) By D istrict 27 b) By Parish 28 15. Percentage Un stable Families a) By School 29 b) By Parish 30 16. Efficient Principal a) By School 31 17. Others (Please add any other indicators you consider impor tant and rate them similarly.) 32 18. 33 19. 34 20. 35 Kindly return to: Mrs. Jean Marshalleck c/o Ministry of Finance and Planning 30 National Heroes Circle, Kingston 4 Please give your name and address if you wish a copy of the results of this study. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A P P E N D IX M ADDITIONAL INDICATORS AS SUCCESTED BY TEACHERS Teachers and Staff 1. Efficient staff and good staff relationship 2 . Efficient staff 3. Efficient teacher by school 4. Teachers' interest 5. Teachers' salary 6 . Teacher/pupil relationship 7. Refresher courses for all teachers 8 . Degree of backwardness among teachers 9. Teachers' salary vs. pupils' performance 10. Teacher/parent communication 11. Teachers' personal interest in children 12. Trained teacher of Grade 6 13. Give all Common Entrance teachers insight of examinations, not some 14. Teacher/parent relationship 15. Teacher/teacher relationship 16. Pupil relationship set 17. Staff relationship 18. Attitude of teachers towards pupils School Working Conditions 1. W orking conditions 2. School equipment 3. Availability of textbooks 4. Proper sanitary conveniences 5. Furniture 6 . Security of building 7. Size of buildings and accomodation by school 8 . Environment school and home 9. Health facilities by school 10. Health and Medical Care 11. How poor classroom equipment affects learning 12. Proper security of school, community facility and properties 13. Books 14. Pleasant learning situations 210 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 211 15. Protection of school property 16. Working conditions Parents and Home 1. Father image should be strong and positive 2. Parental supervision/homework 3. Educational facilities at home 4. Nutrients of pregnant mothers 5. (Q and literacy of parents in aiding children 6. Parents' attitude to pupils 7. Low or no income 8. Most effort spent on other children who are at other schools 9. Parental guidance 10. Family help and guidance 11. Home supervision and discipline 12. Parents' attitude 13. Environment at school and home Pupil Behaviors 1. Pupil/pupil relationship 2. Pupils' interest 3. Pupils' and parents' interest 4. Children should learn to select priorities study before recreation (Games, T .V .) 5. Number of mentally handicapped pupils who need special treatment by school 6. Children's interest 7. Security of child up to age 6 y e ars 8. T .V . programs viewed after school hours 9. Social functions attended after school hours 10. Pupils' attitude to work 11. Pupils attending church and not roaming the streets 12. Pupils should attend selected shows Curriculum /Achievement 1. Availability of textbooks and resource materials 2. Guidance services 3. Emphasis on practical work later easily applied to life 4. Printed textbooks relevant to Jamaica 5. National syllabus for all primary schools 6. Relevance of curriculum to communities 7. Level of reading D iscipline 1. Discipline regarding independent study 2. Amount of play period 3. B eh avio r 4. Discipline Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 212 5. indiscipline in schools 6. Home supervision and discipline Diet and Nutrition 1. Number of free lunches per pupil by school 2. Consistent school feeding program 3. Feeding 4. Diet 5. Nutrition Ministry of Education Inputs 1. Supervisors by (1) constituency, and (2) district 2. Efficient Ministry staff for Island 3. Public relations training by parish/by Island Parent/Teachers Associations 1. P.T.A. involvement 2. Place of P.T.A. in school life O th er 1. Cooperation 2 . Distance from home to school 3. Clothing 4. Take politics out of schools 5. School Boards—quality of members 6 . School transportation 7. Community school relationship Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX N PRINCIPALS' OBSERVATIONS ON FACTORS INFLUENCING PERFORMANCE IN THE COMMON ENTRANCE EXAMINATION Urban Schools—Most Awards St. Richard's Economic factors; unhappy or unstable parental/guardian relationships; parental and teacher pressures. Dunrobin Consistent work; teacher and children; motivation of pupils; regular attendance. Pembroke Hall Regular attendance; supply of learning materials; pupil's interest and attitude; parent's interest and attitude. St. George's Innate intelligence; stable home; loss of parent detrimental, socioeconomic back ground; home with books. Alpha II Not supplied. T a rra n t Not supplied. St. Aloysius Not supplied. Hope V alley Economic background; parents' interest in child's progress; home environment; availability of reading material; ability to help p u p ils, etc. New Providence Not supplied. Mico Practising Two parent family. Urban Schools—Fewest Awards Gordon Town Intelligence; good school adjustm ent; extra lessons; irregular attendance; loss of books. 213 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 214 Urban Schools—Fewest Awards - cont'd Boy1 s Town Not supplied. Tavares Cardens Pupils fail to purchase text books. Physical conditions at home impede " Home W ork." Some pupils do not have even tables to sit around. Parents not capable of helping children with the "Prep" work. Balcombe Drive Irregular attendance; depressing environ mental factors. St. Peter Claver Most of the students are late developers. Manning's Hill Not supplied. Mountain View Unstable environment; lack of personal books; poor attendance; interference by unemployed youths of the community. Seaward One big factor is that many students are anxious to do well so as to be able to get out of their depressed community and go to other schools. Teachers responsible for preparing students are very keen. Hard w ork. Dupont Cood home background; proper food and parental guidance. Rural Schools—Most Awards Duckenfield Not supplied. Golden Grove Regular attendance; regular speed tests; manipulation of tables; good use of library facilities. Lyssons Attendance; home influence; parents who can help; having text books; children who read a lot; ability grouping; do given assignments and corrections regularly. Rural Schools—Fewest Awards Halls Delight - Irregularity of attendance; lack of teaching facilities; lack of exposure. Mt. James - Adequate text books; good teaching skills; a rounded curriculum, especially at Grades Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 215 Rural Schools- -Fewest Awards - cont1 d 5 and 6 levels; use of library books for information and enjoyment; correct usages in oral expression followed by written expression. Thornton Not supplied. Mount Felix Not supplied. Bowden Hill Comprehension skills (poor); computation (fair); some pupils had to travel long distances over hilly terrain to school; others attended poorly; all these pupils got no help with their studies at home. Westphalia Regular checks with past papers at the Ministry of Education. White Horses Not supplied. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission BIBLIOGRAPHY Alexander, L., and Simmons, J. The Determinants of School Achievement in Developing Countries: The Educational Production Function. Washington, D.C. : World Bank, 1975. (ERIC Document Repro- duction Service No. Ed 128 941). American Library Association. Standards for School Library Programs. Chicago: American Library Association, 1960. Andrews, Frank M ., and Withey, Stephen B. Social Indicators of Well- Being. New York: Plenum Press, 1976. Bagley, Christopher. "A Comparative Perspective on the Education of Black Children in Britain." Comparative Education 15, 1 (March 1979):63-81. Baker, Harold E. Educational Indicators of Development, Monograph III. Saskatoon, Canada: Saskatchewan University, 1978. Bargen, Mark,and Walberg, Herbert J. "School Performance." In Evaluating Educational Performance, pp. 239-254. Edited by Herbert J. Walberg. Berkeley, California: McCutchan Publishing Corp., 1974. Bauer, Raymond A. "Detection and Anticipation of Impact: The Nature of the Task." In Social Indicators, pp. 1-67. Edited by Raymond A. Bauer. Cambridge, Massachusetts: M .I.T. Press, 1966. Bell, Daniel. "The Idea of a Social Report." The Public Interest 15 (Spring 1 969) : 72-84. Berdie, Douglas R ., and Anderson, John F. Questionnaires Design and Use. Metuchen, New Jersey: The Scarecrow Press, 1974. Bernard, Harold W. Psychology of Learning and Teaching. New York: McGraw-Hill Book Co., 1972. Biderman, Albert D. "Social Indicators and Goals." In Social Indicators, pp. 68-153. Edited by Raymond A. Bauer. Cambridge, Massachusetts: M .I.T . Press, 1966. Biehler, Robert F. Psychology Applied to Teaching. Boston: Houghton Mifflin Co., 1974. 216 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 217 [Blaschke.] Education Turnkey Systems, Inc. Summary Cost Effective ness Study of Compensatory Education. Lansing : Michigan Department of Education, March 1975. Cited by Selma J. Mushkin and Bradley B. Billings, "Measures of Educational Outcomes in Developing Countries," p. 68. In Educational Indicators: Moni toring the State of Education, pp. 65-78. Proceedings of the 1 975 Educational Testing Service Invitational Conference. Princeton, New Jersey: Educational Testing Service, 1976. (ERIC Document Reproduction Service No. Ed 150 147). Bloom, Benjamin S. Stability and Change in Human Characteristics. New York: John Wiley and Sons, 1964. Brookover, W.; Gigliotti, R.; Henderson R. ; and Schneider, J. "Elemen tary School Climates and School Achievement." Paper presented at the Annual Meeting of the American Sociological Association, New York, 1973. [Burt.] 1940. Cited by L. H. E. Reid, "The Effects of Family Pattern, Length of Schooling, and Other Environmental Factors on English and Basic Arithmetic Attainment of Jamaican Primary School Children." Ph.D. dissertation. University of London, 1964. Campbell, Agnus, and Converse, Philip. The Human Meaning of Social Change. New York: Russell Sage Foundation, 1972. Carlisle, Elaine. "The Conceptual Structure of Social Indicators." In Social Indicators and Social Policy. Edited by Andrew Schonfield and Stella Shaw. London: Heinemann Educational Books, 1972. Carpenter, C. R. "Review of Research on Instructional Television." In Proceedings of the Conference on the Economics of Educational Television, pp. 1-28. Watham, Massachusetts: Brandeis University, 1963. Cassirer, Henry R. Television Teaching Today. Paris: UNESCO, 1962. Clemmer, Robert; Fairbanks, Dwight; Hall, M.; Impara, J.; and Nelson, C. Indicators and Statewide Assessment. Salem, Oregon: Oregon State Department of Education, 1974. (ERIC Document R eproduction S ervice N o. Ed 091 869). Clunie, Enid. "An Exploratory Study of the Ways in Which the Home, the School, and the Society Affect Absenteeism: A Sample of Crade 6 Students from Selected Primary and All-Age Schools in Rural Jamaica." B. Ed. study. University of the West Indies, Jamaica, 1983. Cohen, Jacob. Statistical Power Analysis for the Behavioral Sciences. New York: Academic Press, 1977. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 218 Cohen, Wilbur J. "Educational Indicators and Social Policy." In Educational Indicators: Monitoring the State of Education, pp. 57-62. Proceedings of the 1975 Educational Testing Service Invitational Conference. Princeton, New Jersey: Educational Testing Service, 1976. (ERIC Document Reproduction Service No. Ed 150 147). Coleman, J. S.; Campbell, E .Q .;etal. Equality of Educational Opportunity. U.S. Office of Education. Washington, D.C .: U.S. Government Printing Office, 1966. Comber, L.C ., and Keeves, J.P. Science Education in Nineteen Countries. New York: Wiley-Halstead Press, 1973. Coombs, Philip H., and Hallak, Jacques. Managing Educational Costs. Paris: International Institute for Educational Planning, 1972. Council of Economic Advisers. Economic Indicators. 1984. Washington, D .C .: Government Printing Office, 1984. .Drewnowski, Jan. On Measuring and Planning the Quality of Life. The Hague, Holland: Internationaal Instituut VOOR SOCIALE STUDIEN, 1974. Duncan, B. "Trends in Output and Distribution of Schooling.” In Indicators of Social Change: Concepts and Measurement, pp. 601-672. Edited by Eleanor B. Sheldon and Wilbert E. Moore New York: Russell Sage Foundation, 1968. Duncan, O. D. "Toward Social Reporting: Next Steps." Social Science Frontiers Series, No. 2. New York: Russell Sage Foundation, 1969. Dunkin, M. and Biddle, B. The Study of Teaching. New York: Holt, Rinehart and Winston, 1974. Dwyer, A. "A Proposal for Improving Teaching of English in the Lower Primary Grades in Jamaican Schools." Ed. D. dissertation. University of Columbia, 1971. Educational Testing Service. Educational Indicators: Monitoring the State of Education. Proceedings of the 1975 Educational Testing Service Invitational Conference. Princeton, New Jersey : Educational Testing Service, 1976. (ERIC Document Reproduction Service No. Ed 150 147). Available from Invitational Conference, Educational Testing Service, Princeton, New Jersey, 08541. Edwards, Allen L. Technigues of Attitude Scale Construction. New York: Appleton-Century-Crofts, 1957. Erickson, Donald A ., ed. Educational Organization and Administration. Berkeley, California: McCutchan Publishing Corp., 1977. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 219 Erskine, Cleveland C. "The Relationship between Certain Home Environ mental and School Factors and Lateness and Absenteeism in Crade 8 of Selected All-Age Schools in Rural and Urban Jamaica." B. Ed. study. University of the West Indies, Jamaica, 1979. Farrell, J. P. "The Structural Differentiation of Developing Educational Systems: A Latin-American Comparison." Comparative Education Review 13 (1969) : 294-311. Fenton, Sylvia. "Influences and Constraints on Decision Making in the Primary School Curriculum and their Effects on the Achievement of Aims and Objectives—A Comparative Study of Jamaica and Montserrat." B.Ed. study. University of the West Indies, Jamaica, 1979. Ferriss, Abbott L. Indicators of Trends in American Education. New York: Russell Sage Foundation, 1969. Frane, James W. "Missing Data and BMDP: Some Pragmatic Approaches." Technical Report #45, Health Sciences Computing Facility, University of California— Los Angeles, 1978. Garner, William T . "Inputs and Outputs in the Educational Process." Speech given before the American Educational Research Association Annual Meeting. New York, February, 1971. (ERIC Document Reproduction Service No. Ed 075 948). Gaver, Mary Virginia. Effectiveness of Centralized Library Service in Elementary Schools, 2nd e5~! New Brunswick, New Jersey: Rutgers University Press, 1963. Girling, S. "Teacher Evaluation and Student Participation in the Jamaican Primary School." Ph. D. dissertation. University of Stanford, 1974. Golladay, Mary A. The Condition of Education, 1976 ed. Washington, D.C.: U. S. Government Printing Office, 1976. (ERIC Docu ment Reproduction Service No. Ed 120 910). Good, Thomas L.; Biddle, Bruce J.; Brophy, Jere E. Teachers Make a Difference. New York: Holt, Rinehart and Winston, 1975. Gooler, Dennis D. "The Development and Use of Educational Indicators." In Educational Indicators: Monitoring the State of Education, pp. 11-T T. Proceedings of the 1975 Educational Testing Service Invitational Conference. Princeton, New Jersey: Educational Testing Service, 1976. (ERIC Document Reproduction Service No. Ed 150 147). Gordon, George N. Educational Television. New York: The Center for Applied Research in Education, 1965. Green, R.; Hoffman, L.; Morse, R .; and Morgan R. "The Educational Status of Children During the First Year of School Following Four Years of Little or No Schooling." East Lansing, Michigan: College of Education, Michigan State University, 1966. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 220 Cross, Bertram M. "Preface: A Historical Note on Social Indicators." In Social Indicators, pp. ix-xviii. Edited by Raymond A. Bauer. Cambridge, Massachusetts: M .I.T . Press, 1966a. ______. The State of the Nation: Social Systems Accounting. London: Tavistock Publications Associated Book Publishers, 1966b. Hammond, Robert L. "Evaluation at the Local Level." In Educational Evaluation: Theory and Practice, pp. 157-168. Edited by Blaine R. Worthen and James R. Sanders. Worthington, Ohio: Charles A. Jones Publishing Co., 1973. Hawkridge, David, and Durbridge, Nicola. "Making the Most of Screen Education." Inter Media II. 4/5 (July-Sept. 1 983) :34— 37. London: International Institute of Communications. Hawthorne, Phyllis, comp. Annotated Bibliography of the State Educational Accountability Repository, Report No. 1. Denver, Colorado: State Department of Education, C .A .P ., 1974. (ERIC Document Reproduction Service No. Ed 098 655). Hebb, D. O. The Organization of Behavior. New York: John Wiley and Sons, 1949. Heim, John, and Perl, Lewis. The Educational Production Function: Implications for Educational Manpower Policy. Institute of Public Employment Monograph No. 4. Ithaca, New York: Publications Divisions N .Y ., State School of Industrial and Labor Relations, Cornell University, 1974. Immegart, Glenn L., and Pilecki, Francis J. An Introduction to Systems for the Educational Administrator. Reading, Massachusetts: Addison-Wesley Publishing Co., 1 973. International Institute for Educational Planning. Educational Cost Analysis in Action: Case Studies for Planners—I. Paris: UNESCO, 1972. Jamaica. Department of Statistics. The Labor Force, 1980. Kingston: Department of Statistics, 1980. Jamaica. Ministry of Education. Annual Report of the Ministry of Education 1975-76. Papine, Jamaica: Vocational Training Development Institute, 1976. ______. Directory of Educational Institutions 1976-77. Kingston: Uniprint, 1977. ______. "Education Broadcasting Service ETV Reports, 1978-80." Kingston, Jamaica. (Mimeographed.) ______. Education Statistics 1 975-76. An Annual Review of the Education Sector. Papine, Jamaica: Vocational Training Development Institute, 1976. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 221 ______. Education Statistics 1979-80. An Annual Review of the Education Sector. Papine, Jamaica: Vocational Training Development Institute, 1980. Jamaica. "Schools Library Service Report for Period April 1979 to March 1980." Kingston, Jamaica. (Mimeographed.) Jensen, Arthur. "Equality for Minorities." In Evaluating Educational Performance, pp. 175-222. Edited by Herbert J. Walberg. Berkeley, California: McCutchan Publishing Corp., 1974. Johanson, Richard K. "The Development of an Information System for the Ministry of Education, Jamaica." Ph. D. dissertation. Harvard University, 1969. Johnstone, James N. Indicators of Performance of Educational Systems. Paris: UNESCO, IIEP Publications, 1981. (ERlt Document Reproduction Service No. Ed 212 091). Kean, Cleveland H.A. "An Investigation of Some Environmental and School Organizational Factors Influencing the Performance of St. Thomas Pupils in the 11+ Common Entrance Examination." B. Ed. study. University of the West Indies, Jamaica, 1974. Kendall, M. Multivariate Analysis. New York: Hafner Press, McMillan Publishing Co., 1975. Kerlinger, Fred N., and Pedhazur, Elazar. Multiple Regression in Behavioral Research. New York: Holt, Rinehart and Winston, 1973. Kiros, Fassil G .; Mushkin, S .; and Billings, B. Educational Outcome Measurement in Developing Countries. Washington, D .C .: Georgetown University, 1975. (ERIC Document Reproduction S e rv ic e No. Ed 120 253) . Krathwohl, David. "Preface." In Educational Indicators: Monitoring the State of Education, pp. xi, xii. Proceedings of the 1975 Educational Testing Service Invitational Conference. Princeton, New Jersey: Educational Testing Service, 1976. (ERIC Document Reproduction Service No. Ed 150 147). Kravetz, Nathan. The Evaluation of Educational Systems Outputs: An Exploratory Study. A Research Project, 1970. Paris: International Institute for Educational Planning, 1970. (ERIC Document Reproduction Service No. Ed 055 321). Krejcie, Robert V ., and Morgan, Daryle. "Determining Sample Size for Research Activities." Educational and Psychological Measurement 30 (Autumn 1970) :607-610. Kuhns, Rodney J. "Input Ouput Analysis of Secondary Schools in Pennsylvania." Ph. D. dissertation, Pennsylvania State University, 1972. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 222 Land, Kenneth C. "On the Definition of Social Indicators." The American Sociologist 6 (November 1971) :322-325. Land, Kenneth C ., and Spilerman, Seymour, eds. Social Indicator Models. New York: Russell Sage Foundation, 1975. Lazarsfeld, Paul F. "Notes on the History of Quantification in Sociology— Trends, Sources and Problems." In Quantification: A History of the Meaning of Measurement in the Natural and Social Sciences^ pp. 147-203. Edited by Harry Woolf. Indianapolis: BobbsrMerrill, 1961. Lewis, Arthur J. "Forecasting Social Trends as a Basis for Formulating Educational Policy." Paper presented at World Future Society Conference. Houston, Texas, October 1978. (ERIC Document Reproduction Service No. Ed 161 786). Manley, Douglas R. "Mental Ability in Jamaica: An Examination of the Performance of Children in the Jamaican Common Entrance Exami nation, 1959." Social and Economic Studies 12, 1 (March 1963) : 5 1 -7 1 . Marshalleck, E. C. "Micro-Level Enrollment Projections—A Study Involving the Parish of Hanover, Jamaica." M. A. thesis, Andrews University, Berrien Springs, Michigan, May 1976. McCall, Chester H ., Jr. Sampling and Statistics Handbook for Research in Education. Washington, D .C .: National Educational Association, 1980. McCuire, Carson, and White, Ccorge D. "The Measurement of Social Status." Report No. 3, Laboratory of Human Behavior, Department of Educational Psychology, University of Texas, 1955. McVeigh, Thomas, comp. Social Indicators: A Bibliography. Monticello, Illinois: Council of Planning Librarians, 1971. (ERIC Document Reproduction Service No. Ed 106 951). Meyer, W. J., and Thompson, G. C. "Sex Differences in the Distribution of Teacher Approval and Disapproval among Sixth Crade Children." Journal of Educational Psychology 47 (1956) :385-396. Miller, James G. "Toward a General Theory for the Behavioral Sciences." The American Psychologist 10 (1955) : 513- 531 . Miller, William H ., and Murray, Ross M., eds. Jamaica Education Sector Survey. Kingston: Ministry of Education, 1977. (ERIC Document Reproduction Service No. Ed 152 768). Moats, Calvin W. "Validation of Ten Indicators of School Effectiveness Identified by the Intermediate Unit Planning Study." Ed. D. dissertation, Lehigh University, 1970. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 223 Monteith, Sylvan I. "An Investigation into the Equitableness of Trained Teacher Distribution and Factors Affecting Teacher Retention in a Sample of Urban, Semi-Urban, Rural, and Deep Rural Primary Schools in Jamaica." B. Ed. study. University of the West Indies, Jamaica, 1978. Moore, W. E., and Sheldon, E.B. "Monitoring Social Change: A Conceptual and Programmatic Statement." In the American Statistical Association, Proceedings of the Social Statistics Section, pp. 144-149. 1965. Morgenstern, Oskar. Accuracy of Economic Observations. Princeton, New Jersey: Princeton University Press, 1963. Mushkin, Selma J., and Billings, Bradley B. "Measures of Educational Outcomes in Developing Countries." In Educational Indicators: Monitoring the State of Education, pp. 65-78. Proceedings of the 1975 Educational Testing Service Invitational Conference. Princeton, New Jersey: Educational Testing Service, 1976. (ERIC Document Reproduction Service No. Ed 150 147). Noah, Harold Jr., "Education Needs Rational Decision Making." In Emerging Patterns of Administrative Accountability, pp. 197-212. Edited by Lesley H. Browder. Berkeley, California: McCutchan Publishing Co., 1971. Olson, Mancur, Jr. "Measurement and Efficiency in Education." In Educational Indicators: Monitoring the State of Education, pp. 41-53. Proceedings of the 1975 Educational Testing Service Invitational Conference. Princeton, New Jersey: Educational Testing Service, 1976. (ERIC Document Reproduction Service No. Ed 150 147). ______. "The Plan and Purpose of the Social Report." The Public In te re s t 15 (S p rin g 1969) :85-97. Organization for Economic Cooperation and Development. . Indicators of Performance of Educational Systems. Paris: Organization for Economic Cooperation and Development, 1973. (ERIC Document Reproduction Service No. Ed 087 677) . Pedhazur, Elazar. Multiple Regression in Behavioral Research, Explanation and Prediction. New York: Holt, Rinehart and Winston, 1982. Platt, William J. "Policy-Making and International Studies." Paper presented at Conference on Educational Achievement. Boston, Massachusetts: Harvard University, November 1973. Psacharpopoulos, George, and Hinchliffe, Keith. Returns to Education— An International Comparison. San Francisco, California: Jossey- Bass, 1973. Purves, Alan C. Literature Education in Ten Countries. New York: Wiley-Halstead Press, 1973. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 224 Ragbir, Frank. "Academic Achievement Motivation in Jamaican High School Students." Caribbean Journal of Education 4, 3 (September 1977) : 146-63. Reid, J. Christopher,and MacLennan, Donald W. Research in Instruc tional Television and Film. Washington, D.C.: Office of Education, U.S. Government Printing Office, 1967. Reid, L. H. E. "The Effects of Family Pattern, Length of Schooling, and Other Environmental Factors on English and Basic Arithmetic Attainments of Jamaican Primary School Children." Unpublished Ph. D. dissertation. University of London, 1964. Roach, D. A. "Predictors of Mathematics Achievement in Jamaican Elementary School Children." Perceptual and Motor Skills 52, 3 (1981):785, 786. Rosenshine, B. "New Directions for Research on Teaching." In U. S. Department of Health, Education and Welfare, Office of Education, How Teachers Make a Difference, pp. 66-95. Washington, D .C .: U.S. Government Printing Office, 1971. Rosenshine, B.. and Furst, N. "The Use of Direct Observation to Study Teaching." In Second Handbook of Research on Teaching. Edited by R. Travers. Chicago: Rand McNally, 1973. Runyon, Richard P., and Haber, Audrey. Fundamentals of Behavioral Statistics, 3rd ed. Reading, Massachusetts: Addison-Wesley Publishing Co., 1977. Sawhill, Isabel V. "The Role of Social Indicators and Social Reporting in Public Expenditure Decisions." In The Analysis and Evaluation of Public Expenditures: The PPB System, pp. 473-485. A compendium of papers submitted to the Subcommittee on Economy in Government of the Joint Economic Committee, 91st Congress of the United States. Vol. 1. Washington, D.C. : U.S. Government Printing Office, 1969. Schiefelbein, Ernesto. Private 1974 conversation. Cited by Selma J. Mushkin and Bradley B. Billings, "Measures of Educational Out comes in Developing Countries," p. 68. In Educational Indicators: Monitoring the State of Education, pp. 65-78. Proceedings of the 1975 Educational Testing Service Invitational Conference. Princeton, New Jersey: Educational Testing Service, 1976. (ERIC Document Reproduction Service No. Ed 150 147). Schiefelbein, Ernesto, and Joseph Farrell. "Factors Influencing Academic Performance among Chilean Primary Students." Paper presented at the Annual Meeting of the International Society of Educational Planners and the American Association for the Advancement of Science. Mexico City, June 1973. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 225 Schmidt, W. "Socioeconomic Status, Schooling, Intelligence, and Scholastic Progress in a Community in Which Education Is Not Yet Compulsory." Paedagogica Europa 2 ( 1966):275-286. Schonfield, Andrew, and Shaw, Stella, eds. Social Indicators and Social Policy. London: Heinmann Educational Books, 1972. Sheldon, Eleanor B. "Discussion" and "The Social Indicators Movement." In Educational Indicators: Monitoring the State of Education, pp. 82-84 and pp. 3-9. Proceedings of the 1975 Educational Testing Service Invitational Conference. Princeton, New Jersey: Educational Testing Service, 1975. (ERIC Document Reproduction Service No. Ed 150 147) . Sheldon, Eleanor B ., and Freeman, Howard E. "Notes on Social Indicators: Promises and Potential." Policy Sciences 1 (1970) :97— 111. Sheldon, Eleanor B., and Land, K. C. "Social Reporting for the 1970s, a Review and Programmatic Statement." Policy Sciences 3 (1972) : 137-151. Sheldon, E. B., and Moore, W. E., eds. Indicators of Social Change: Concepts and Measurements. New York: Russell Sage Foundation, 1968. Shrestha, Gambhir. "Second Order Regression Model Applied to 1972-73 Florida Statewide Assessment Program." Paper presented at the Annual Meeting of the American Educational Research Association, New York, April 1977. (ERIC Document Reproduction Service No. Ed 137 3 8 2 ). Silvernail, David L. Teaching Styles as Related to Student Achievement— What the Research Says to the Teacher. Washington, D .C .: National Educational Association, 1979. Simmons, J. "Retention of Cognitive Skills Acquired in Primary School." Comparative Education Review 20 (1976) :79-93. Spencer, H. The Study of Sociology. New York: Appleton & Co., 1914. Summers, Anita A ., and Wolfe, Barbara L. "Equality of Educational Opportunity Quantified: A Production Function Approach." Paper presented for presentation at the Econometric Society Winter Meeting, December, 1974. Philadelphia: Federal Reserve Bank of Philadelphia, Pennsylvania. (ERIC Document Reproduction Service No. Ed 107 736). Tatsuoka, Maurice M. Multivariate Analysis: Techniques for Educational and Psychological Research. New York: John Wiley and Sons, 1971. Teaching by Television. 2nd ed. A Report from the Ford Foundation and the Fund for the Advancement of Education. New York: Ford Foundation, Office of Reports, 1961. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 226 Thorndike, Robert L. Reading Comprehension Education in Fifteen Countries. New York: Wiley-Halstead Press, 1973. Torgerson, Warren S. Theory and Methods of Scaling. New York: John Wiley and Sons, 1958. U.S. Congress. Joint Economic Committee. Economic Indicators. Prepared by The Council of Economic Advisers. Washington, D .C.: U.S. Government Printing Office, 1984. U.S. Department of Health, Education, and Welfare. The Condition of Education. A Statistical Report on the Condition of American Education, 1975. Washington, D .C .: U.S. Government Printing Office, 1976. ______. Office of Education, Bureau of Educational Personnel Develop ment. How Teachers Make a Difference. Washington, D.C.: U.S. Government Printing Office, 1971. ______. Toward a Social Report. Washington, D.C. : U.S. Government Printing Office, 1969. (ERIC Document Reproduction Service No. Ed 054 039). U.S. Office of Management and Budget. Social Indicators 1973. Washington, D .C .: U.S. Government Printing Office, 1 975. ______. Social Indicators 1976. Washington, D.C.: U.S. Government Printing Office, 1976. ______. Social Indicators 1980. Washington, D.C.: U.S. Government Printing Office, 1980. Vernon, P. E. Selection for Secondary Education in Jamaica. A Report to the Minister of Education. Kingston, Jamaica: The Government Printer, 1961. Walberg, Herbert J., ed. Evaluating Educational Performance. Berkeley, California: McCutchan Publishing Corp., 1974. Walker, D. and Schaffarzick, J. "Comparing Curricula." Review of Educa tional Research, 44 (1974):83-111. Warner, Paul, and Dublin, Lewis. The Bluebook of Economics. New York: Regents Publishing Co., 1965. Warner, W. L.; Meeker, M.; Eells, K. Social Class in America. Chicago: Science Research Associates, 1949. Welkowitz, Joan; Ewen, Robert B.; Cohen, Jacob. Introductory Statistics for the Behavioral Sciences. 3rd ed. New York: Academic Press, 1982. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 227 Wilson, J. "Pupil Achievement in Northern-1 reland Primary Schools: Twenty-five Years of Findings and Issues." Irish Journal of Education 7 (1973) : 102-116. Withey, Stephen. "Quality of Life as an Educational Outcome." In Educational Indicators: Monitoring the State of Education, pp. 29-40. Proceedings of the 1975 Educational Testing Service Invitational Conference. Princeton, New Jersey: Educational Testing Service, 1976. (ERIC Document Reproduction Service No. 150 147) . Winer, B. J. Statistical Principles in Experimental Design. New York: McGraw-Hill, 1962. Woolf, Harry, ed. Quantification. New York: Bobbs-Merrill Co., 1961. Young, Louise B., ed. Population in Perspective. New York: Oxford University Press, 1968. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. VITA Personal Data: Name: Edith Geraldine Marshalleck Date of Birth: June 16, 1930 Place of Birth: Darliston, Westmoreland, Jamaica Degrees Awarded: 1956 B.A. (Ceneral) University of London (University College of the West Indies, Mona, Jamaica) 1957 Diploma of Ed. University College of the West Indies, (Distinction, Mona, Jamaica in Teaching) 1972 Higher Diploma University of the West Indies, Mona, Jamaica of Education 1976 M.A. Education Andrews University, Berrien Springs, Michigan 1984 E d .D . Andrews University, Berrien Springs, Michigan Educational Administration Positions Held: 1957-58 Acting Principal, Bahamas Academy, Nassau, Bahamas 1958-61 Clerk to the Privy Council, Governor's Secretary's Office, Jamaica 1961-66 Administrative positions. Ministry of Finance and Planning and Ministry of Agriculture, Jamaica 1978-80 Projects Coordinator and Acting Director, Construction and Maintenance Division, Ministry of Education, Jamaica 1981-82 Director, Budget Division, Ministry of Finance and Planning, Jamaica 1982- Deputy Financial Secretary, Ministry of Finance and Planning, Jamaica Board Membership: 1972-73 Jamaica Movement for the Advancement of Literacy 1981- Project Analysis and Monitoring Company, Jamaica Professional Membership: 1975 Charter Member Phi Delta Kappa, Andrews University chapter, Berrien Springs, Michigan Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.