Using Examinations to Improve...
98
WORl D BA.NK TECHNICAL PAPER NUMBER 165 AFRKI. A TECHNICAL DEPART MEN,T SERES Using Examinations to Improve Education A Study in ]Fourteen African Countries Thomas Kellaghan and Vincent Greaney A L Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized
Transcript of Using Examinations to Improve...
WORl D BA.NK TECHNICAL PAPER NUMBER 165
AFRKI. A TECHNICAL DEPART MEN,T SERES
Using Examinations to Improve Education
A Study in ]Fourteen African Countries
Thomas Kellaghan and Vincent Greaney
A L
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
RECENT WORLD BANK TECHNICAL PAPERS
No. 103 Banerjee, Shrubs in Tropical Forest Ecosystems: Examples from India
No. 104 Schware, The World Software Industry and Software Engineering: Opportunities and Constraintsfor Newly Industrialized Economies
No. 105 Pasha and McGarry, Rural Water Supply and Sanitation in Pakistan: Lessons from Experience
No. 106 Pinto and Besant-Jones, Demand and Netback Values for Gas in Electricity
No. 107 Electric Power Research Institute and EMENA, The Current State of Atmospheric Fluidized-BedCombustion Technology
No. 108 Falloux, Land Information and Remote Sensing for Renewable Resource Management in Sub-SaharanAfrica: A Demand-Driven Approach (also in French, 108F)
No. 109 Carr, Technologyfor Small-Scale Farmers in Sub-Saharan Africa: Experience with Food Crop Productionin Five Major Ecological Zones
No. 110 Dixon, Talbot, and Le Moigne, Dams and the Environment: Considerations in World Bank Projects
No. 111 Jeffcoate and Pond, Large Water Meters: Guidelines for Selection, Testing, and Maintenance
No. 112 Cook and Grut, Agroforestry in Sub-Saharan Africa: A Farmer's Perspective
No. 113 Vergara and Babelon, The Petrochemical Industry in Developing Asia: A Review of the CurrentSituation and Prospects for Development in the 1990s
No. 114 McGuire and Popkins, Helping Women Improve Nutrition in the Developing World: Beating the ZeroSum Game
No. 115 Le Moigne, Plusquellec, and Barghouti, Dam Safety and the Environment
No. 116 Nelson, Dryland Management: The "Desertification" Problem
No. 117 Barghouti, Timmer, and Siegel, Rural Diversification: Lessons from East Asia
No. 118 Pritchard, Lending by the World Bankfor Agricultural Research: A Review of the Years 1981through 1987
No. 119 Asia Region Technical Department, Flood Control in Bangladesh: A Plan for Action
No. 120 Plusquellec, The Gezira Irrigation Scheme in Sudan: Objectives, Design, and Perfornance
No. 121 Listorti, Environmental Health Components for Water Supply, Sanitation, and Urban Projects
No. 122 Dessing, Support for Microenterprises: Lessons for Sub-Saharan Africa
No. 123 Barghouti and Le Moigne, Irrigation in Sub-Saharan Africa: The Development of Publicand Private Systems
No. 124 Zymelman, Science, Education, and Development in Sub-Saharan Africa
No. 125 van de Walle and Foster, Fertility Decline in Africa: Assessment and Prospects
No. 126 Davis, MacKnight, IMO Staff, and Others, Environmental Considerations for Port and HarborDevelopments
No. 127 Doolette and Magrath, editors, Watershed Development in Asia: Strategies and Technologies
No. 128 Gastellu-Etchegorry, editor, Satellite Remote Sensing for Agricultural Projects
No. 129 Berkoff, Irrigation Management on the Indo-Gangetic Plain
No. 130 Agnes Kiss, editor, Living with Wildlife: Wildlife Resource Management with Local Participationin Africa
No. 131 Nair, The Prospects for Agroforestry in the Tropics
No. 132 Murphy, Casley, and Curry, Farmers' Estimations as a Source of Production Data: MethodologicalGuidelines for Cereals in Africa
No. 133 Agriculture and Rural Development Department, ACIAR, AIDAB, and ISNAR, AgriculturalBiotechnology: The Next "Green Revolution"?
(List continues on the inside back cover)
v<o)AFRICA TECHNICAL DEPARTMENT SERIES
Technical Paper Series
No. 122 Dessing, Support for Microenterprises: Lessons for Sub-Saharan Africa
No. 130 Kiss, editor, Living with Wildlife: Wildlife Resource Management with Local Participation in Africa
No. 132 Murphy, Casley, and Curry, Farmers' Estimations as a Source of Production Data: MethodologicalGuidelines for Cereals in Africa
No. 135 Walshe, Grindle, Nell, and Bachmann, Dairy Development in Sub-Saharan Africa: A Study of Issuesand Options
No. 141 Riverson, Gaviria, and Thriscutt, Rural Roads in Sub-Saharan Africa: Lessons from World Bank Experience
No. 142 Kiss and Meerman, Integrated Pest Management and African Agriculture
No. 143 Grut, Gray, and Egli, Forest Pricing and Concession Policies: Managing the High Forests of Westand Central Africa
No. 157 Critchley, Reij, and Seznec, Water Harvestingfor Plant Production, vol. I: Case Studiesand Conclusions for Sub-Saharan Africa
No. 161 Riverson and Carapetis, Intennediate Means of Transport in Sub-Saharan Africa: Its Potentialfor Improving Rural Travel and Transport
Discussion Paper Series
No. 82 Psacharopoulos, Why Educational Policies Can Fail: An Overview of Selected African Experiences
No. 83 Craig, Comparative African Experiences in Implementing Educational Policies
No. 84 Kiros, Implementing Educational Policies in Ethiopia
No. 85 Eshiwani, Implementing Educational Policies in Kenya
No. 86 Galabawa, Implementing Educational Policies in Tanzania
No. 87 Thelejani, Implementing Educational Policies in Lesotho
No. 88 Magalula, Implementing Educational Policies in Swaziland
No. 89 Odaet, Implementing Educational Policies in Uganda
No. 90 Achola, Implementing Educational Policies in Zambia
No. 91 Maravanyika, Implementing Educational Policies in Zimbabwe
No. 101 Russell, Jacobsen, and Stanley, International Migration and Development in Sub-Saharan Africa,vol. I: Overview
No. 102 Rlussell, Jacobsen, and Stanley, International Migration and Development in Sub-Saharan Africa,vol. H: Country Analyses
No. 147 Jaeger, The Effects of Economic Policies on African Agriculture: From Past Harm to Future Hope
WORLD BANK TECHNICAL PAPER NUMBER 165
AFRICA TECHNICAL DEPARTMENT SERIES
Using Examinations to Improve Education
A Study in Fourteen African Countries
Thomas Kellaghan and Vincent Greaney
The World BankWashington, D.C.
Copyright © 1992The Intemational Bank for Reconstructionand Development/THE WORLD BANK1818 H Street, N.W.Washington, D.C. 20433, U.S.A.
All rights reservedManufactured in the United States of AmericaFirst printing February 1992
Technical Papers are published to communicate the results of the Bank's work to the developmentcommunity with the least possible delay. The typescript of this paper therefore has not been prepared inaccordance with the procedures appropriate to formal printed texts, and the World Bank accepts noresponsibility for errors.
The findings, interpretations, and condusions expressed in this paper are entirely those of the author(s)and should not be attributed in any manner to the World Bank, to its affiliated organizations, or tomembers of its Board of Executive Directors or the countries they represent. The World Bank does notguarantee the accuracy of the data included in this publication and accepts no responsibility whatsoeverfor any consequence of their use. Any maps that accompany the text have been prepared solely for theconvenience of readers; the designations and presentation of material in them do not imply the expressionof any opinion whatsoever on the part of the World Bank, its affiliates, or its Board or member countriesconcerning the legal status of any country, territory, city, or area or of the authorities thereof orconcerning the delimitation of its boundaries or its national affiliation.
The material in this publication is copyrighted. Requests for permission to reproduce portions of it shouldbe sent to Office of the Publisher at the address shown in the copyright notice above. The World Bankencourages dissemination of its work and will normally give permission promptly and, when thereproduction is for noncommercial purposes, without asking a fee. Permission to photocopy portions fordassroom use is not required, though notification of such use having been made will be appreciated.
The complete backlist of publications from the World Bank is shown in the annual Index of Publications,which contains an alphabetical title list (with full ordering information) and indexes of subjects, authors,and countries and regions. The latest edition is available free of charge from the Distribution Unit, Officeof the Publisher, Department F, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., orfrom Publications, The World Bank, 66, avenue d'I6na, 75116 Paris, France.
ISSN: 0253-7494
Thomas Kellaghan is director of the Educational Research Centre, St. Patrick's College, Dublin.Vincent Greaney is an education specialist at the World Bank.
Library of Congress Cataloging-in-Publication Data
Kellaghan, Thomas.Using examinations to improve education: a study in fourteen
African countries / Thomas Kellaghan and Vincent Greaney.p. cm. - (World Bank technical paper, ISSN 0253-7494; no.
165. Africa Technical Department series)Includes bibliographical references (p. ).ISBN 0-8213-2052-11. Examinations-Africa, Sub-Saharan. 2. Educational evaluation-
-Africa, Sub-Saharan. I. Greaney, Vincent. II. Title.Im. Series: World Bank technical paper; no. 165. IV. Series:World Bank technical paper. Africa Technical Department series.LB3058.A357K45 1992371.2.'6'0967-dc2O 92-164
CIP
Foreword
Public examinations tend to exert enor- country are outlined in the report. Issuesmous influence on the nature of learning and discussed include low pass rates, the back-teachLing; they tend to dictate not only what wash effect that examinations have on teach-is taught but also how it is taught. In devel- ing and on grade repetition, the roles ofcping countries, while most examinations school-based and practical assessment, theserve a number of functions, including cer- implications of national language policies,tification and accountability, their main func- the use of quotas and compensation proce-tion is to select students for the next highest dures, the publication of national schoollevel of the educational system. Their im- rankings based on examination performance,pact is most pronounced due to the short- and the low level of technical support avail-age of places, particularly at the secondary able to examination authorities. Guidelinesand tertiary levels of formal schooling. are offered for improving the quality of ex-
This study presents for the first time a aminations and for using examinations todetailed description of the types, functions, improve education.performance levels, governance, administra- Many current international issues in thetion, and funding of public examinations in controversial topic of examinations anda range of African countries with different assessment are addressed. Thus, the studyeducational traditions. The national public- should be of interest to ministry of educa-examination systems of fourteen Sub- tion officials, national examination bodies,Saharan countries are reviewed. Six of the development agencies, and national and in-countries are Anglophone (Kenya, Lesotho, ternational educational organizations. ByMauritius, Swaziland, Uganda, and Zambia) offering concrete proposals for improvingand six Francophone (Chad, Guinea, Mada- the quality of public examinations, and bygascar, Mauritania, Rwanda, and Togo). The focussing on the close interrelationshiptwo remaining countries are Cape Verde and between formal assessment, teaching andEthiopia. Public examinations are offered learning, it helps pinpoint the way to rais-in virtually all of the countries at the end of ing the level and the quality of education ofprimary, lower-secondary, and upper- pupils in Sub-Saharan Africa.secondary school.
Procedures for funding examinations; forconstructing, administering, and scoringpapers; and for reporting results in each
Ismail SerageldinDirectorTechnical DepartmentAfrica Region
v
Preface
The studies synthesized in this report to deepen and extend the studies. Follow-originated from the emphasis on improve- ing consultation, fourteen countries agreedment in educational quality in the World to participate in a set of studies under theBank's policy paper, Education in Sub-Sa- terms of reference. Public examination sys-haran Africa, published in January 1988. tems in six Anglophone countries (Kenya,This emphasis was highlighted by twelve Lesotho, Mauritius, Swaziland, Uganda, andAfrican Ministers of Education at the first Zambia), six Francophone countries (Chad,plenary meeting of the Donors to African Guinea, Madagascar, Mauritania, Rwanda,Education (DAE), established in the same and Togo) and two other countries (Capemonth to support priority educational de- Verde and Ethiopia) were investigatedvelopments. When the DAE Task Force met The studies were started in early 1989in June 1988, a Working Group on School under AFTED management, with Bank andExaminations (WGSE) was also established, Irish Government funding. Ireland agreedin recognition of the important role exami- to act as lead donor, operating through thenations can play in quality improvement. Irish Aid Agency, Higher Education for De-
The World Bank Education and Train- velopment Cooperation (HEDCO). An ini-ing Division of the Africa Technical Depart- tial model study was carried out in Swazi-ment (AFTED), in assuming responsibility land. Studies of examination systems in thefor developing appropriate activities to fur- six Anglophone countries and in Ethiopiather quality improvement, then prepared were carried out under the direction of theterms of reference to undertake studies on ERC, Dublin; examination systems in Fran-examinations in primary and secondary edu- cophone countries and in Cape Verde werecation in Sub-Saharan Africa. The main ob- studied by Management Planning and Re-jectives of the studies may be summarized search Consultants (MPRC), Bahrain. Anas follows: important feature of these studies was the(1) to improve educational quality in a cost- active participation of counterparts nomi-effective way through adjustment of inputs nated by Ministries of Education (MOE). Inrelative to examination systems addition, lead donor Ireland hosted periodic(2) to help develop institutional capacity in meetings of the WGSE to advance the imple-Sub-Saharan countries accordingly. mentation process. Progress reports were
Five studies (funded by the Bank's Eco- presented at DAE Task Force meetings. Thenomic Development Institute (EDI) and the findings of this synthesis report were dis-Irish Government), completed in Fall 1988 cussed at a November 1990 World Bank-by the Educational Research Centre (ERC), sponsored seminar in Killiney, Dublin, at-Dublin, were discussed at a seminar on Us- tended by examination officials from eaching Examinations and Standardized Testing of the fourteen countries and representativesto Improve Educational Quality in Lusaka, of donor agencies.Zambia in November 1988. At this semi- This synthesis report is based on: (1) annar, AFTED presented its terms of reference analysis of existing examination systems;
vii
(2) diagnosis of qualitative problems; and Preparation of the series of studies on(3) assessment of existing institutional ca- Using Examinations to Improve Educationpacity. The report draws also on research was managed by James McCabe, Principalevidence from work undertaken in Western Education Planner, Education and Trainingas well as African countries to form an as- Division, Africa Technical Department of thesessment of examination practice in Africa World Bank.and to suggest guidelines for the construc- Individual country studies were pre-tion and administration of examinations that pared in collaboration with Ministry of Edu-improve the quality of education. cation authorities in each country, by the
following principal authors:
1. Ethiopia: R. Reilly, M. O'Donoghue, V. Greaney2. Kenya: S. McGuinness, M. O'Donoghue, A. Yussufu, M. Kithuke3. Lesotho: J. McCabe, E. MacAogain, T. Kellaghan4. Mauritius: N. Baumgart, M. O'Donoghue5. Swaziland: T. Kellaghan, J. McCabe, S. Sukati6. Uganda: J. Sheehan, S. McGuinness7. Zambia: T. Kellaghan, M. Martin, J. Sheehan8. Chad: A. Brimer, J. Hosman, D. Ouman Hamid9. Guinea: M Chaibderraine, I. Sankhon
1 0. Madagascar: N. Constantine, J. Hosman, Mr. Rakoto, et al.11. Mauritania: M Chaibderraine, M. El Hafed Ould12. Rwanda: A. Brimer, J. Hosman, D. Kayinamura, E. Munyantwali13. Togo: N Constantine, A. Sewa14. Cape Verde: F. Oliveira, 0. Carvalho
The authors are indebted to Stephen MacAogain, and Ingvar Werdelin for com-Heyneman for his input into the de- ments on an earlier draft; and to Thomassign stage of the studies; to Marlaine Poole, Alicia Hetzner, and Michael MatovinaLockheed, George Madaus, Eoghan for preparing the report for publication.
viii
Acronyms for Public Examinationsand Examination Authorities
BEPC Brevet d'etudes du premier cycle. External examination taken at end of first cyclesecondary (four years) in Togo, in Guinea, and in Chad.
CEE Elementary certificate, taken at the end of primary school, Mauritania.CEP Certificat defin d'etudes primaires. Primary-school leaving examination, Rwanda.CEPD Certificat de fin d'etudes de l'enseignement du premier degre External certificate
examination at end of primary; also admission examination to secondary school,Togo.
CEPE Certificat d'etudes primaires elementaires. External examination taken at end ofprimary school in Madagascar, Chad.
CEPEA Certificat d'etudes primaires 6Mementaires. External examination taken at end of
primary school in Chad, for Arabic-speaking students.CFE/FP1 Certificat defin d'etudes de laformation professionnelle du niveau 1. External
examination taken at end of first two-year vocational training period (Level I) inMadagascar.
CFE/FP2 Certificat defin d'etudes de laformation professionnelle du niveau 2. Externalexamination taken at end of second two-year vocational training period (LevelII) in Madagascar.
CFEPCES Certificat de fin d 'etudes du premier cycle de I 'enseignement secondaire.External examination taken at end of first cycle of secondary school in Madagascar.
CPE Certificate of Primary Education, Mauritius.DEC Director of Examinations and Concours, Togo.ESLCE Ethiopian School Leaving Certificate Examination.HSC Higher School Certificate, Mauritius.KCPE Kenya Certificate of Primary Education, taken at the end of the eight-year
primary cycle.KCSE Kenya Certificate of Secondary Education taken at the end of the four-year
secondary cycle (introduced in 1989).KNEC Kenya National Examinations Council.MES Mauritius Examination Syndicate.PC Primary Certificate.PLE Primary Leaving Certificate.SC School Certificate, Mauritius.SEC Service of Examinations and Concours, Chad.UACE Uganda Advanced Certificate of Education.UCE Uganda Certificate of Education.UCLES University of Cambridge Local Examinations Syndicate.UNEB Uganda National Examinations Board.
ix
Contents
1. SUMMARY 1
2. THE NEED FOR REFORM IN AFRICAN EDUCATION 5Growth in Education 5Concern with Quality 6Strategies to Improve Education 7Examinations and Reform 8
Effects of Examinations on Curricula 8Quality of Examinations 9Role of Examinations in Reform 11
3. THE EDUCATIONAL SYSTEMS OF FOURTEEN COUNTRIES 15Countries in the Study 15Current Size and Participation Rates of Educational Systems 16Concern with Quality of Provision 16Scarcity of Places 18Repetition 18Dropout 19Teacher Quality 20
4. PUBLIC EXAMINATIONS 23Tradition 23Titles and Descriptions of Examinations 23Size of Examination Enterprise 24Setting of Examinations 24Use of Aptitude Tests 24School-based Components of Examinations 26Scoring 27Technology 28Functions of Examinations 28
Selection 28Certification 29Accountability 30
Performance 30Examination Repetition 32Governance and Administration 33Localization of Examinations 33Funding 34
Fees 34
Security 35Feedback 36
Comparisons among Schools/Districts 36Analysis of Results 37Timeliness 37
Challenges to Examination Results 37Outside-School Tuition 38
5. ISSUES IN EXAMINATIONS AND ASSESSMENT 39Expanding Enrollments 39Resources for Examinations 40Effects of Examinations on Curriculum and Teaching 40Reducing the Burden of Examinations and
Introducing School-based Assessment 42Use of Multiple-Choice Tests and Their
Associated Technology 44Validity 45Accountability 47Language of Examination 48Comparability of Performance 48Quotas 49
6. USING EXAMINATIONS TO IMPROVE EDUCATION 51Action Plans 52
Administration 52Technical and Training Requirements 55Equipment 56Construction/Rehabilitation 56Cross-National Cooperation 56
Improving Assessment in Public Examinations 57Anticipating Problems and Undesirable Effects 60Conditions to Improve the Administration of Examinations 62
Examination Authority 62National Cooperation 63International Cooperation 63Number and Format of Public Examinations 64Formal Training 64
7. CONCLUSION 658. REFERENCES 679. APPENDICES 73
National Education Systems 73Titles and Descriptions of Examinations 74Kenya Certificate of Primary Education: 80Sample Essay and Marker's Comments 81
1 Summary
This Synthesis Report summarizes data Students in virtually all of the systemsand information from a set of studies un- take public examinations at the end of pri-dertaken in: mary, first cycle secondary and second cycle
(1) Kenya, Lesotho, Mauritius, Swazi- secondary school. Candidates for primary-land, Uganda and Zambia; school leaving certificate examinations in(2) Chad, Guinea, Madagascar, Mau- most countries far exceed those taking sec-ritania, Rwanda, and Togo, and ondary school examinations. Countries in(3) Cape Verde and Ethiopia. Francophone areas tend to require their stu-
It outlines the findings, issues and rec- dents to take more formal examinations thanommended improvements from the four- those in Anglophone areas; most often theyteen individual studies undertaken on pri- also offer term, end-of-year and concoursmary and secondary school examinations. (competitive selection) examinations.
Considerable reliance has been placed Setting of examinations seems a ratheron public examinations in African educa- haphazard exercise, with some notable ex-tion as a means of ensuring that teachers ceptions. In some instances, a final selec-and students cover a common curriculum, tion is simply made from a pool of ques-and accordingly as a particularly effective tions submitted by teachers. Most of theinstrument for raising academic standards. examinations reviewed show serious weak-However, it has also been argued that while nesses. Very little use is made of school-public examinations may help raise aca- based assessment.demic standards, they may also give rise to Most scoring (or correction) of public ex-problems in the educational system. The aminations is carried out by teachers, thoughfindings of the fourteen studies confirm this a relatively small number of systems use op-point. tical scanning and/or computer equipment.
Two countries (Ethiopia and Swaziland)Findings have had to abandon optical scanning.
The important issue of marker reliabilityPublic examinations have three main does not appear to have been addressed in
functions: certification, selection and ac- most countries, though it should be, givencountability. While most examinations the seriousness of the decisions made on theserve all three to some extent, selection is basis of examination performance.undoubtedly the main function. The para- Pass rates tend to be low. In some in-mount importance of selection is reflected stances the pass rate is determined by thein the large numbers each year who are not number of places available at the next high-promoted to the next highest level of the est level of schooling. Data on pass rateseducational system. There is also a grow- should be interpreted with caution. For ex-ing tendency to use a single examination ample, quota systems are sometimes usedfor purposes of both selection and certifica- to attain national objectives such as equalitytion of level of attainment. of male and female pass rates. Pass marks
from region to region within countries may
1
vary in an effort to ensure regional equality Issuesof opportunity.
Public examinations undoubtedly exert To date, not enough attention appears toenormous pressure on activities in schools. have been focused on the educational con-Teachers tend to gear teaching to the tests sequences of existing examination systems.to be taken and to ignore material not fea- However, the likely impact of changingtured in such tests, even if it is mandated in present systems needs to be studied care-the official curriculum. fully, given the experience with public ex-
Most of the questions in public examina- aminations in non-African settings.tions tend to measure students' ability to It must be appreciated that however ex-recall facts. Relatively little attention is given cellent a test is in terms of validity, reliabil-to higher-order cognitive skills such as those ity, and efficiency, the pressure on studentsinvolved in synthesis or evaluation. Exami- to perform will remain substantially un-nation content tends also to be academic in changed while lack of places at the next high-nature; life outside of school seldom features est level persists. Where such bottlenecks ex-in examination questions. In most countries, ist, it is naive to think that teachers will notparticularly those of Francophone tradition, attempt to gear teaching to testing, or thatexaminations are of the written essay or oral students will not seek outside help to enabletype. Other forms of assessment (multiple- them to advance to the next level.choice, project, practical, or aural) are less In this context also, it must be empha-frequently used. sized that the high repetition rate attribut-
Whereas Ministries of Education admin- able to "failure" may represent a seriousister public examinations in Francophone waste of scarce educational resources.countries, independent or semi-independent The need for definition and implemen-examination boards generally fulfil this func- tation of language policies by governmentstion in Anglophone countries. In a small is urgent, and in that regard, decisions arenumber of countries, universities have re- required on the most appropriate languagessponsibility for the secondary-school leav- to be used in public examinations. In mosting examination. countries the present emphasis on French
For examinations run by Ministries of and English has probably contributed to lowEducation, fees are generally not charged at pass rates.the primary level. However, Anglophone Clearly, practical subjects and school-countries are more likely to charge fees. based assessments will tend to receive littleMost systems charge fees for secondary emphasis until they are incorporated intoschool leaving examinations; fees can be very public examinations. These types of assess-high in some cases. ment, however, are time-consuming, expen-
Concern over security is a conspicuous sive, and difficult to moderate.feature of most systems. Legal provision for Efforts to localize (or nationalize) exami-breaches of security vary. nations, following long experience with sys-
Examination results are seldom used to tems such as those offered by the Univer-provide useful feedback to schools, admin- sity of Cambridge Local Examination Syn-istrators or curriculum bodies. Thus, a good dicate, run the risk of undermining publicopportunity to effect change is little ex- confidence in the examination system.ploited.
2
likely impact of such tests on the vast ma- should be taken of factors other than teach-jority of students who may not proceed to ing effort.the next highest level must be considered. 6. The number of public examinations
Publication of school results and national should be reduced to help diminish repeti-rankings are designed to provide "incentive tion and dropout rates and the inevitableinformation," a form of public accountabil- sense of failure experienced by students.ity expected to motivate schools. Cogni- 7. The amount of time teachers spend onzance should also be taken, however, of the testing and preparing for public examina-many factors (e.g., socio-economic) other tions should be lessened to provide morethan teaching effort and quality elements time for teaching.that affect examination performance. 8. Detailed, timely feedback should be
Efforts to modernize the processing of provided to schools on levels of pupil per-examination data should take into account formance and areas of difficulty in publicthe level of technical support available and examinations.particularly the extent to which foreign cur- 9. Predictive validity studies of public ex-rency is likely to be available for procure- aminations should be conducted.ment of required equipment, materials, and 10. The professional competence of exami-expertise. nation authorities needs to be developed, es-
pecially in test construction.Recommended Improvements 11. Each examination board should have a
research capacity.It is clear that a public-examination sys- 12. Examination authorities should work
tem can exert a highly positive influence not dosely with curriculum organizations andonly on the nature of assessment, but much with educational administrators.more importantly on the nature of teaching 13. Regional professional networks shouldand learning. While examinations cannot be developed to initiate exchange programschange the educational system, they can be and share common interests and concerns.made to reflect the objectives of curriculum An example is the Association of Heads ofdevelopers and educational planners. The Institutions Responsible for Examinationsfollowing recommendations are offered: and Education Assessment in East and
Southern Africa.1. Examinations should reflect the full cur- 14. A post-graduate degree course shouldriculum, not merely a limited aspect of it. be established in an African country for ex-2. Higher-order cognitive skills should be amination authority personnel.assessed to ensure they are taught.3. Skills to be tested should not be lim- Finally, while implementation of theited to academic areas but should also be above can effect improvement, it should berelevant to out-of-school tasks. recognized that other factors such as avail-4. A variety of examination formats ability of places, language of instruction andshould be used, including written, oral, au- assessment, amount and quality of literacyral, and practical. materials and quality of teaching, together5. In evaluating published examination with examinations, will play critical roles inresults and national rankings, account improving the quality of education.
3
2 The Need for Reform in African Education
After a period of massive expansion, edu- tween 1960 and 1983, the number of pri-cation in many African countries has, dur- mary-school pupils in Sub-Saharan Africaing the last decade, encountered a series of increased by well over 400 percent, fromproblems. While demand remains high, 11.85 to 51.35 million. Gross enrollment ra-progress towards universal provision at the tio over the same period rose from 36 to 75primary level and expansion at the second- percent (World Bank 1988). At secondaryary level has slowed, mainly because of a level, the increase, at 1,400 percent, was evenshortage of money. More limited financial greater. Student numbers rose from .79 toresources than had been available in the past 11.12 million and gross enrollment ratio fromare being spread more thinly over increas- 3 to 20 percent (World Bank 1988).ing numbers of students (Fuller 1986; WorldBank 1988). Given this situation, the choice Despite massive expansion, the discrep-for policymakers, for at least the remainder ancy between provision at primary and sec-of this century, would appear to lie either in ondary levels remained great. The percent-increased efficiency in the use of existing age of students in Sub-Saharan countriesresources or in the acceptance of declining who transferred from the last grade of pri-standards of access, equity, and scholastic mary school to the first grade of secondaryachievement (Windham 1986). general education in 1983 was 43 (World
Bank 1988). This figure has a particularIn this chapter, we outline some current relevance to examinations because they are
concerns about education in Africa. Pro- used to control the flow of students at thisposals to address these concerns relate to a juncture, as well as at other points, in thevariety of actions-the provision of books educational system. The success or failureand equipment, teacher training, enlisting of a student at any of the important selec-the support of children's homes, and im- tion points in the system can have very se-proving the nutritional status of children. rious consequences for his or her educationalOur concern is with the possible role of and occupational future. It is precisely be-public examinations, a well-established fea- cause of their role as gatekeepers in educa-ture of African educational systems, in im- tional systems, in that the number of placesproving educational achievement. diminishes as one ascends the educational
hierarchy, that examinations have acquiredGrowth in Education the importance they possess in African coun-
tries.Africa has the lowest primary, second-
ary, and university enrollment rates of any Educational growth in Africa has slowed,wforld region (Nafziger 1988). This is so and in some countries has stagnated or ac-despite the fact that educational systems tually declined, in recent years. This hasthroughout the continent expanded rapidly been attributed largely to poor economicin the 1960s and 1970s. For example, be- conditions that were aggravated, in the first
5
instance, by the oil crisis in the 1970s and, for economic reasons (see, for example,more recently, by increasing levels of debt- Swaziland, Ministry of Education 1986; Tanrepayment obligations (Windham 1986). To- 1985; Zambia, Ministry of Education 1977).tal public expenditure on education in Af- Attempts to meet these commitments canrica declined from $10 billion in 1980 to $8.9 be expected to contribute to the growth inbillion in 1983, while per-pupil expenditure, numbers attending school.both in absolute terms and as a proportionof gross national product, declined mark- Concern with Qualityedly at both primary and secondary levels(World Bank 1988). Side by side with the problems of de-
creased financial resources and stagnatingThere are a number of reasons why the enrollments, commentators in a number of
educational systems of Africa are particu- African countries, as in countries elsewherelarly vulnerable to economic circumstances throughout the world (see Heyneman andand the conditions of government finances. White 1986), have expressed concern aboutFirst, the role of the public sector in financ- a decline in quality of the education beinging education is a major one. In 1983, for offered in schools (see, for example, Evalu-example, 85.4 percent of primary students ation Research of the General Educationand 85.1 percent of secondary students were System in Ethiopia - ERGESE, 1986; Kellyin public schools. The proportion for Fran- 1991; Zambia, Ministry of Education 1977).cophone countries was larger than for An- Although many countries now collect edu-glophone countries (World Bank 1988). cational statistics, for example on participa-Second, state subsidies are large, as fees for tion rates, on a fairly routine basis, evidencepublic education recover only 5.7 percent of relating to the quality of provision or ofthe cost of primary education, 11.4 percent output is more difficult to come by. Thereof the cost of secondary education, and 1.9 is some evidence of decline in the quality ofpercent of the cost of higher education. provision. Supplies of key inputs, especiallyThird, because of the low level of gross books and other learning materials, are re-national product (GNP) in many African ported to be critically low in many coun-countries, the cost of public education for tries (World Bank 1988). Concern has alsoone student as a percentage of per capita been expressed about decline in quality ofGNP is higher in Africa than in any other output (as measured, for example, by stu-region (Mingat and Psacharopoulos 1985; dent achievement). Such concern is basedNafziger 1988; Unesco 1990, Table 19). for the most part on impressionistic evi-
dence, evidence, of course, that should notAlthough resources are limited, and even be ignored. The available empirical evi-
diminishing, school-age populations are in- dence on a decline in standards, however,creasing rapidly. During the past twenty is far from satisfactory. While the Interna-years, over 50 million new pupils enrolled tional Association of the Evaluation ofin school throughout the continent, and it is Eduactional Achievement (IEA) studies ofexpected that the number that will become achievement in mathematics in Nigeria andeligible in the next twenty years will be 110 Swaziland have been cited as evidence ofmillion (Windham 1986). In addition, the declining standards (World Bank 1988, p.demand for educated manpower to support 33), these studies do not tell us anythingeconomic growth is likely to grow, putting about change as they were carried out onpressure on governments to expand educa- only one occasion.tional facilities. Moreover, many countriesmade commitments following independence Examination performance over time is ato provide education for social as well as possible source of evidence regarding chang-
6
ing standards (see Kahn 1990). However, Hallak 1990; Heyneman 1985,1987; Mingatthe characteristics of students taking an ex- and Psacharopoulos 1985). The World BankamLnation as well as the standard required (1988) has shown a particular interest in thisto aLchieve a particular grade in an examina- problem and has identified three major ar-tion, particularly if a predetermined propor- eas to be considered in the formulation oftion of students is assigned to each grade, policy:may vary from year to year. Thus, knowl-edge of the numbers achieving particular (1) Adjustment to current demographic andgrades over time does not permit unambigu- fiscal realities, that will require: (a) diversi-ous inferences about changes in standards. fication of sources of finance (in the form ofIt is worth noting, however, that in Swazi- increased cost sharing in public educationland, pass rates increased on the Cambridge and encouragement of nongovernmentalOverseas School Certificate (that was con- suppliers of educational services); and (b)trolled by the Cambridge Syndicate) from unit-cost containment, that will be of greater36 percent in 1980 to 54 percent in 1988. importance than cost sharing in countries inThere was an even greater increase in the that the scope for cost sharing is negligiblepercentage of first and second-class passes. or non-existent. One aspect of the organiza-It is of interest, in the context of the present tion of education that could contribute toreport, that the improvements have been at- unit-cost containment would be a reductiontributed to strategies (that were implemented in the amount of grade repetition.between 1984 and 1987) designed to increasethe competence of inspectors and teachers (2) Revitalization of the existing educationalin the assessment of students (Lulsegged infrastructure, that will include: (a) a re-1988). It is also of interest that the nature of newed commitment to academic standardsstudent assessment in Swaziland, even be- (principally by strengthening examinationfore the efforts to improve it, may have dif- systems); (b) restoration of an efficient mixfered from that in other countries. An analy- of inputs in education (especially increasingsis of IEA data, that were collected in Nige- the amount of textbooks and other learningria and Swaziland in 1980-81, indicates that materials); and (c) greater investment in thetime spent by teachers in monitoring and operation and maintenance of physical plant.evaluating student performance was posi-tively associated with achievement in Swazi- (3) Selective expansion, that will be viableland but not in Nigeria (Lockheed and only after measures of adjustment and re-Kormenan 1989). The Swazi data, however vitalization have begun to take hold andone interprets them, do seem to contradict will involve: (a) renewed progress towardsthe general perception of declining standards the long-term goal of universal primaryof achievement in African schools, under- education; (b) the development of alterna-lining the need for more systematic empiri- tive ways of delivering educational servicescal evidence relating to standards of achieve- (including distance education); (c) trainingment over time. for those who have entered the work force
in job-related skills; and (d) research andStrategies to Imrprove Education postgraduate education.
Because of the perceived need in many In a document adopted by the Worldcountries not jutst to maintain standards of Conference on Education for All, that wasaccess, equity, and academic achievement convened jointly by the executive heads ofbut to improve them, a number of strategies the United Nation's Children's Fundhave been considered for educational reform (UNICEF), United Nations Development(see Fuller 1986; Fuller and Heyneman 1989; Programme (UNDP), Unesco, and the World
7
Bank; and was held in Jomtien, Thailand, were written verbal and reading comprehen-from March 5 to 9, 1990, the emphasis on sion tests. This did not seem to make muchbasic education and the achievements of stu- sense given the tasks that gunners were sup-dents who receive that education is again posed to do, such as maintaining, adjusting,evident. Article 3 of the World Declaration and repairing guns. On examining whaton Education for All (1990) states that "Basic went on in training, the researcher foundeducation should be provided to all children, that classes consisted of lecturing and dem-youth and adults" (p. 4). Recognizing that onstrating how to use equipment, ratherthe provision of such education is only than having students do the actual tasksmeaningful if people actually acquire useful themselves. He then designed performanceknowledge, reasoning ability, skills, and tests in that students were required to dem-values, Article 4 of the Declaration states that onstrate competence by, for example, remov-the focus of basic education must be "on ing and replacing the extractor plunger on aactual learning acquisition and outcome, 5 /38-inch anti-aircraft gun. Few of the stu-rather than exclusively upon enrolment, dents could perform the tasks.continued participation in organized pro-grams and completion of certification re- New students soon found out what thequirements" (p. 5). new performance tests were like and began
practicing the assembly and disassembly ofExaminations and Reform guns. The instructors also moved from lec-
turing to more practical work with guns andThe idea that examinations may have an gun mounts. At the end of the course, stu-
important role to play in effecting a reform dents performed very much better on thein education in developing countries arises performance tests than students who hadfrom the belief that they exercise a strong experienced the older instructional method.influence on what is taught in schools and Further, the predictive validity of the verbalcan be used as instruments of accountabil- and reading tests dropped while the valid-ity. It can also be argued that they provide ity of the mechanical aptitude and mechani-a relatively simple means of controlling a cal knowledge tests improved.system in that resources for other means ofcontrol, such as school inspection and The interesting point of this anecdote isteacher training, are limited and in that stu- that no attempt was made to change the cur-dents attend private schools and non-formal riculum or teacher behavior. The dramaticeducational establishments as well as pub- changes in curriculum and achievement camelic schools. about solely through a change in testing.
Effects of Examinations on Curricula In the day-to-day workings of schools,the effects of assessment on curricula and
What evidence do we have that exami- on student achievement are not so dramatic.nations affect curricula and teaching in However, there can be little doubt that ex-schools? Perhaps the most striking example aminations, to which high stakes are at-of how an assessment procedure can affect tached, exert considerable influence on whatthe content and skills covered in a curricu- goes on in schools (Fredericksen 1984;lum is to be found in an anecdote of a per- Madaus and Kellaghan, in press). Thoseson who carried out research on the selec- involved in an educational system that hastion and training of U.S. naval personnel such examinations will attest to the fact thatduring the second World War (Fredericksen the topics that teachers and students attend1984). This person found that the best tests to in class and in study are the topics thatfor predicting grades given by instructors are likely to appear on examination papers.
8
One commentator has concluded that since many examinations contain very little refer-examinations represent "the ultimate goal ence to the everyday life of students outsideof the educational career, they define what the school, dealing with scholastic topics andare the important aspects of a school cur- applications for the most part, rather than,riculum and they dictate to a large degree for example, trying to find out if a studenttheA quality of the school experience for both can use money in the market place. Fifth,teacher and student alike" (Little 1982). the quality of actual items used in tests is
often poor (see Cambridge Educational Con-If the examinations are good, this might sultants 1988; ERGESE 1986; Kelly 1991:
be a satisfactory situation. If the objectives Lesotho 1982; Little 1982; Myeni 1985;and skills to be measured are carefully cho- Oxenham 1983).sen and if the tests truly measure them, thenthe goals of instruction will become explicit If schools gear their teaching to suchand well-defined targets for teachers and examinations, then they are unlikely to bestudents on that they can focus their efforts. very successful in developing in theirFurthermore, the examinations will provide students the kind of knowledge and skillsstudents and teachers with standards of ex- that most people would regard aspected achievement. Given this situation, desirable-skills of observation, problemthere should be no reason why students identification, problem-solving andshould not work for marks, and good rea- reasoning, and particularly knowledge andsons why they should ( Fredericksen 1984). skills that can be applied in the day-to-dayIn practice, however, examinations may lack life of the many students who will have only"construct validity" or, for other reasons, minimal exposure to formal education (seemay not meet the high standards that would National Commission on Testing and Publicjustify teachers and students devoting their Policy - NCTPP, 1990; Nigam 1982). Rather,efforts to performing well on them. the effect of such examinations, as was
observed in Ghana, is likely to be "to supportQuality of Examinations and encourage rote-memorization, routine
drilling, bookishness" (Brooke andEvidence from many countries through- Oxenham 1984, p. 158). Indeed, the view
out the world over the past century sug- was expressed in a government report ingests that public examinations suffer from a Lesotho (1982) that many problems withvariety of defects (Madaus and Kellaghan, curriculum and instruction seem to stemin press). In African countries, defects of fromexaminations have been pointed out on nu- ... .the inordinate emphasis givenmerous occasions, in both official and unof- to the preparation for terminalficial reports. First, most examinations, at examinations which undermines theboth primary and secondary level, are lim- attainment of certain objectives thatited to pencil-and-paper tests and so ignore are critical to the country's economica variety of skills that cannot be measured development...in this way. Second, examinations empha- The JC [Junior Certificate]size the achievement of scholastic skills (par- examination heavily emphasizes theticularly language and mathematics at the accumulation of factual knowledgeend of primary schooling) paying very little and neglects general reasoning skillsattention to more practical skills. Third, in and problem-solving activities" (p.most examination questions, the student is 94).required to recall or recognize factual knowl-edge, rather than to synthesize material or In Ghana, Brooke and Oxenham (1984)apply principles to new situations. Fourth, have noted that teachers tend to neglect non-
9
examination subjects even though the offi- labus, only six of 39 authors named werecial timetable might require that allocation African, and only three of these were Westof time be adhered to strictly. Teachers even African. As well as studying Shakespeare,pick topics within subjects according to students had the option to study the poetrywhether they judge them as likely or not to of Chaucer, John Donne, and Georgeappear in examinations. The same phenom- Herbert. The Senegalese Baccalaureat of-enon has been observed in Uganda, where fered examinations in ten languages, noneneglect of the teaching of practical skills, es- of them African (Bray, Clarke, and Stephenspecially at the primary level, and the conse- 1986).quences of this for students on leavingschool have been noted (Uganda, Ministry While examinations at the primary levelof Education 1989). In such situations, the are controlled locally in all countries, for-effect of examinations can be inhibiting, serv- eign influences are to be found in curriculaing to distort or prevent learning rather than and examinations at this level also. In ato promote or facilitate it (Heyneman and study of the national primary-certificate ex-Ransom 1990; Little 1990). aminations of ten Anglophone countries,
Hawes (1979) found that, in 1978, five usedThe use of examinations that were set only the English language, a further four
and marked outside of African countries used English in all papers except in the oneprobably contributed to the perpetuation of examining the local language, and only onethe situation in that examinations took little used a local language as the main examin-cognizance of the conditions in that students ing language. Reliance on English in theselived and were likely to live in the future cases invariably reflects the fact that there(Kellaghan 1991). As late as 1981, the school- are several, in some cases a great many, localleaving certificate examinations in seven languages, none of that is spoken through-African countries were set and marked by out the country. In linguistically heteroge-the University of Cambridge Local Exami- neous situations, learning and being exam-nations Syndicate (Bray, Clarke, and ined in a metropolitan language is perceivedStephens 1986), a situation that has now to favor no particular group. Furthermore,changed. Since preparation for examinations the use of metropolitan languages, such aswas important in schools, we might expect English and French, is seen as conferringthat the examinations contributed signifi- benefits, particularly on those who are likelycantly towards the maintenance of a west- to proceed to third-level education (Eisemonern academic education. The examinations 1990). However, the advisability of teach-also, of course, contributed to the mainte- ing and examining through English ornance of standards and acceptance of certi- French students who have poor proficiencyfication at the international level, important in the language and who rarely if ever useconsiderations in developing countries. the language outside school is something
that has been questioned many times. Stu-The setting up of local examination dents with limited knowledge of a language
boards did not result in a sudden break with will inevitably be handicapped in the ac-the traditions of the colonial system in the quisition of knowledge and skills presentedcontent of curricula or examinations or in in the language as well as in their ability tomodes of examining. For example, the West demonstrate in examinations the knowledgeAfrican Examinations Council was set up in and skills they have acquired (Eisemon1951 to serve the Gambia, the Gold Coast, 1990).Sierra Leone, Nigeria, and later Liberia, as alocal independent body. However, in its A further problem with many external-1980 Advanced-level English Literature syl- examination systems in Africa is that their
10
pr-mary use is to control the flow of stu- tributing to the problems, while others see adents through the educational system. This role for examinations as part of a possibleuse is understandable, and the selection solution. Those who perceive examinationsfunction of examinations will no doubt con- as part of the problem cite the evidencetinue to be important as long as there is a considered above concerning the inadequacyshortage of places at higher levels of educa- of examinations. If examinations serve totional systems. However, there are a num- distort or prevent desirable learning, it mightber of dangers inherent in focussing on the not seem unreasonable to conclude thatuse of examinations for selection. First, the public examinations should be abolished.examinations will tend to be geared to the Together with this view, it is usually pro-needs of pupils who are doing relatively well posed that a system of school-based assess-in the systerm while the needs of lower ment should be installed in place of exter-achieving students may not be adequately nal examinations.met. And, second, it is likely that the ex-amriinations will focus on academic topics. An alternative view is that since there isIn the early 1970s, the ILO (1972), in a com- a lack of the resources that would be neededrment that could have been applied to the to introduce alternative assessment proce-educational systems of many African coun- dures, for example school-based assessment,tries, noted that, in Kenya, most primary cur- to the educational systems of most Africanricula and examinations ignored the needs countries and since the need for selectionof terminating students. will continue in the foreseeable future, it
would appear to be unrealistic to talk of dis-To help avoid such undesirable effects pensing with present examination systems.
in constructing examinations, it is important This view also recognizes that examinationto keep clearly in mind the need for certify- systems in Africa are perceived by many toing the achievements of all students, as well be relatively fair and impartial and that theyas the need for selection. Further, cognizance also serve to legitimate the allocation ofshould be taken of the motivational effect of scarce educational benefits. Before an alter-examinations on students and the influence native system could be introduced, it wouldof examinations on what is taught and em- be necessary to demonstrate that it couldphasized in schools. Examinations that are fulfil this task equally well. It is also ar-designed for only the top 20, 30, or even 50 gued that examinations, if properly de-percent of students and the curricula that signed, could have a beneficial effect on theprepare students for such examinations are quality of education in schools. Because oflikely to be seen as irrelevant to lower- the high stakes associated with examinationsachieving students. in terms of student opportunities and teacher
accountability, changes in examinationsRole of Examinations in Reform would most likely be reflected in changes in
educational practice in schools. If theGiven the importance of public exami- changes involve improving the quality and
nations in educational systems, to an extent scope of examinations, these in turn shouldthat a school's success may be "judged result in improving the educational experi-strictly by the performance of its students in ences of students in schools (Heynemanthe examinations" (Fafunwa 1974, p. 193), it 1987; Heyneman and White 1986;is not surprising that the role of examina- McNamara 1982). In Little's (1984) words,tions has received particular attention in the examination improvements could help turncontext of the problems facing education in the educational system into one "which en-Africa today. For some observers, external courages, rather than stultifies, desirable out-public examninations are perceived as con- comes" (p. 228). Again, Little (1982) argues
11
that "the quality of the examination system and over time might even have importantitself can have a considerable impact on the effects on the economic performance of aquality of skill formation encouraged by the nation (Heyneman 1987). However, such aeducation system, that skills in turn could system could have serious and damaginghave a considerable impact on the inputs to effects on the educational experiences ofthe labor market" (p. 177). many students if it ignores the fact that for
many students-in fact the majority-learn-The possible role of examinations in edu- ing has to have utility beyond that of quali-
cational reform has been considered in a fying individuals for the next level of edu-number of policy documents. In the World cation (World Bank 1988). Thus, a procedureBank (1988) strategies outlined above, the that most efficiently selects students may beneed to strengthen examination systems in inadequate for certification purposes. Simi-the interest of raising academic standards is larly, a procedure that is adequate for certi-specifically mentioned in the context of re- fication is unlikely to be the most appropri-vitalizing the existing educational infrastruc- ate one for monitoring the quality of perfor-ture. Further, examination and assessment mance of a school or of the educational sys-systems could be rationalized and made tem in general.more efficient; this seems important at a timewhen numbers taking examinations are In considering examination reform, it isshowing large increases in many countries. important to bear in mind the different func-Again, assessment procedures could be used tions of examinations and the many possiblein a formative way to guide instructional effects of examination systems on schools,and learning processes in schools to reduce teaching, and learning. It is also importantdropout rates and grade repetition. In their to realize that in selecting the specific inno-important role in educational selection, ex- vations most likely to fulfill a country's re-aminations could contribute to greater effi- quirements, insofar as is possible, effectsciency in the educational system by identi- should be anticipated and judged as desir-fying students most likely to benefit from able or undesirable in the context of thefurther education. Finally, in future expan- general goals and aims of each country'ssion of educational systems, examination own educational system (Eckstein and Noahand assessment procedures could be as- 1988; Heyneman 1987). We shall return tosigned an important role in the provision of these issues in Chapter 5.feedback information to schools on studentachievement levels. They could also have Efforts to reform curricula and examina-an important role in ensuring comparability tions are already in evidence in a number ofof standards between school-based and non- countries. For example, a program of cur-school-based candidates, if non-school-based riculum reform has been carried out in Le-ways of delivering educational services are sotho. Syllabi in the core subjects (Sesotho,developed further. English, Science, Mathematics, Social Stud-
ies) have been designed and disseminated.The fact that systems of examinations and Syllabi comprise units, specific objectives,
assessment can be used for a variety of pur- suggested activities, skills and concepts toposes should not be taken to imply that a be learned by students, and resource mate-single system of examinations or assessment rials. Side by side with these activities, skillscan readily serve all purposes equally well. checklists for Sesotho, English, and Math-For example, an examination system that ef- ematics (Standards 1-3) and sample testficiently selects the pupils most likely to ben- question booklets in English and Mathemat-efit from further education might contrib- ics (Standards 4-6) have been constructed toute to the identification of a technical elite reflect the new curricula.
12
Initiatives in Kenya have been more ob- certificate examination although the infor-viously directed towards the use of exami- mation did not receive wide publicity untilnations to improve the quality of learning after the 1978 examination. The first full-in schools. In these initiatives, that have length newsletter was based on the 1978 ex-been reported by Somerset (1987, 1988), the amination results.content of the public examinations at the endof primary schooling was changed and a sys- Only limited information is available re-tern to provide feedback information to the garding the impact of these procedures onpublic and to schools was introduced. The school practice or student achievement. Atreform in the content of examinations in- the time he prepared his report, Somersetvolved the inclusion of a much broader spec- (1987) indicated that relevant data on teach-trum of cognitive skills than had previously ing methods and content had not been col-been included in examinations, skills de- lected. He did, however, provide informa-signed to measure comprehension and ap- tion on changes in the relative mean scoresplication, as well as skills that could be ap- of districts throughout the country betweenplied in a wide range of contexts, in and out 1976 and 1981. He attributed changes inof school. This was done in recognition of these scores between 1976 and 1979 mainlythe fact that examinations should not be con- to the impact of the incentive feedback sys-fined to the measurement of students' abil- tem. If that is correct, then the impact wasity to memorize factual information but negative since mean performances betweenshould also promote the teaching and learn- districts, that it had been hoped would nar-ing of competencies that would be useful row, substantially widened during the pe-not only to those who stay in school but riod. In the following two years, duringalso to the majority who would leave after which guidance information was available,the examination. Little (1984) has described the trend was reversed and some districtsthe changes between 1971 and 1979 in the in which mean scores had been relativelydistribution of items designed to test knowl- low improved their positions. It may beedge, comprehension, and application in the that guidance information contributed to thisCertificate of Primary Education in Kenya improvement, but any firm conclusionsas "dramatic," about the effects of any of the changes in
the examinations system are not possible onIn the feedback system in Kenya, lists of the basis of the available information.
schools were published that reported theoverall mean of the performance of students Since students' test performances on thein each school on the examination. Mean primary-certificate examination each yearstandard scores for each district in the coun- were converted to standard scores derivedtry were also published. This was described from a distribution with a mean of 50 and aas "incentive information." "Guidance in- standard deviation of 15, inferences cannotformation," based on an analysis of the per- be drawn about possible trends in studentformance of students nationally on indi- achievement over time. Yussufu (1989), how-vidual questions, was also provided in a ever, has stated that the performance of can-newsletter that was sent to schools. The didates on the Kenya Certificate of Primarynewsletter explained changes in the content Education Examination (KCPE), first intro-and skills covered in examinations, identi- duced in 1985, "'has, in general terms, shownfied topics and skills causing problems, and a steady consistent improvement." This con-suggested ways of teaching these topics and clusion is presumably based on changes inskills. The incentive information was first students' raw scores. Information about thepublished following the 1976 primary-school nature or extent of the changes, however, is
not provided.
13
Neither is information available on other will not necessarily make schools anypossible effects of the examination and re- more effective especially insofar asporting systems, though we might expect fostering outcomes like better healtheffects on, for example, retention, repetition, are concerned.and dropout rates in schools.
How precisely examinations might con-Eisemon, Patel, and Abagi (1987) observe tribute to the improvement of education in
that the introduction of new question items ways that everyone regards as desirable ison the primary-school leaving examination not clear. Is it, for example, necessary toin Kenya has not, in their opinion, changed build tests in accordance with cognitive theo-primary-school instruction "in ways in ries of the measurement of achievement, aswhich greater emphasis on problem solv- Eisemon and others suggest? While such aning, reasoning and explanation can be dis- approach might be helpful, though its limi-cerned." They stress the need to build ex- tations must also be recognized in light ofaminations in accordance with cognitive our poor understanding of the process oftheories of the measurement of achievement achievement measurement, it would hardlyif the examinations are to have a beneficial seem to be necessary. Insofar as we know,effect on school practice. They studied the no such theories guided the work of theimpact of examinations that had been con- person involved in the measurement of thestructed in accordance with such a theory performance of the young navy personnelon instruction and learning in a Nairobi pri- in Fredericksen's study (1984), describedmary school. The study is limited in its above. Even if construction of a test alongscope in a number of respects: it is confined certain principles is more likely to assist into one school; it deals only with health edu- student learning than a test that does notcation; and it was carried out under experi- follow those principles, we know very littlemental conditions. Thus the findings might about the mechanism that can translate ex-not apply in the normal circumstances un- perience of examinations and knowledge ofder which examinations are administered. results (that might be in the heads of stu-In some instances examination items assess dents, teachers, parents, or administrators)pupils' abilities to integrate existing and new into higher student academic achievement.knowledge. These items are explicitly related So long as that is so, we should be circum-to the competent performance of target be- spect in our recommendations about the usehaviors. Because of this deliberate structur- of examinations in the interest of raisinging of items, teachers' explanatory behav- student achievement. Actions based on de-iors and their emphasis on procedures to cisions taken on an a priori basis, and with-foster pupil understanding will improve. out the benefit of conceptual analysis andOn the basis of this study and of their con- empirical evidence, run the risk of being atsideration of reform efforts in Kenyan edu- best ineffective and at worst damaging. Be-cation, Eisemon and others (1987) conclude cause of this, we emphasize the need to con-that: sider possible problems and undesirable ef-
Psychometric manipulation of exami- fects that can arise when examinations arenation items in the absence of more used to direct the activities of teachers andresearch into instructional and cogni- students. We consider these problems intive processes may produce better Chapter 5 after we set out guidelines relat-tests and different teaching, but this ing to the improvement of examinations.
14
3 The Educational Systems of Fourteen Countries
Countries in the Study
In this report, we are concerned with the Table 2.1examination systems of fourteen African POPULATION, AREA, AND PER CAPITA GNP
countries. While many of these countries OF STUDY COUNTRIES, 1984experienced a variety of colonizers (Arab,Belgian, British, French, German, Italian, and Countries/ Population Area Per capitaPortuguese), they are sometimes categorized linguistic (millions) (thousands of GNP ($)in terms of the European language that in status square kms)
addition to local languages, is used in thecountry today. Using this criterion, six of Anglophonethe countries can be described as Anglo-phone (Kenya, Lesotho, Mauritius, Swazi- Kenya 19.5 583 310
land, Uganda, and Zambia); six as Franco- Lesotho 1.5 30 530phone (Chad, Guinea, Madagascar, Mauri- Mauritius 1.0 2 1,090
tania, Rwanda, and Togo); while two fall Sinto neither of these categories (Cape Verde, Swaziland 0.7 17 790Ethiopia). Uganda 15.0 236 230
Zambia 6.4 753 470At the outset, it has to be recognized that
great variation exists among (and oftenwithin) the countries that were studied for Francophone
this report. They range in population from Chad 4.9 1,284
less than a million (Cape Verde, Swaziland) Guinea 5.9 246 330
to over 40 million (Ethiopia). They also varyconsiderably in size, from less than 2,000 Madagascar 9.9 587 260
square kilometers (Mauritius) to 1,222,000 Mauritania 1.7 1,031 450square kilometers (Ethiopia). Eleven of the Rwanda 5.8 26 280
countries fall into the World Bank (1990)low-income category and three into the Togo 2.9 57 250middle-income category. Economic differ-ences are reflected in annual per capita GNP Otherthat ranges from $110 (in Ethiopia) to $790 CapeVerde 0.3 4(in Swaziland) and $1,090 (in Mauritius) Ca Vd.3
(Table 2.1). Even these figures may under- Ethiopia 42.2 1,222 110
estimate the extent of poverty in countries(Swaziland, Ministry of Education 1985). Source: WorldBank(1988),exceptforCapeVerde,forwhich
Since they are based on the monetary as- Unesco (1988) data for 1985 are used.sessment of activities related to the modernsector of the economy (that in many casesrelies heavily on foreign capital), they throw
15
little light on the standard of living of the countries exceeded 100. Figures over 100majority of the population who in all the percent can be attributed to the relativelycountries live in rural areas and are engaged high number of over-age pupils in primaryin the traditional sector of the economy, that school. The percentage of primary-schoolis mainly agriculture. students who were female ranged from a
low of 28 percent (Chad) to a high of 56Some of the countries are ethnically and percent (Lesotho).
culturally fairly homogenous; others arevery diverse. In some countries, for ex- Relatively small numbers of students ad-ample, a single language is spoken by most vance to secondary school. In all but threeof the population (Cape Verde, Lesotho, countries, the gross enrollment ratio at theMadagascar, Swaziland). In others, and this secondary level was less than 25 percent inis the more common situation, a diversity of 1986. Ratios at the secondary level exceededlanguages and dialects is in use (see World 20 percent in only five of our study coun-Bank 1988). tries (Lesotho, Madagascar, Mauritius,
Swaziland, Togo). Female participation wasDifferences among countries, that we will lower at the secondary than at the primary
again advert to when considering the edu- level. In nine of the fourteen countries, fe-cational and examination systems of the males constituted less than 40 percent of to-countries, are outlined here to draw atten- tal enrollment (Table 2.2). Lesotho is thetion to the fact that any general conclusions only country in which the number of femalereached in this report will have to be secondary-school students was greater thanadapted and interpreted within the context the number of males. The ratio of primaryof the particular circumstances of each indi- to secondary pupils was particularly largevidual country. in a number of countries, most notably
Rwanda (43:1), Uganda (12:1), Kenya (11:1),Current Size and Participation Rates of Zambia (9:1), and Lesotho (8:1). Clearly, forEducational Systems the vast majority of pupils, formal educa-
tion can be equated with primary school-Twelve of the fourteen countries in- ing.
cluded in our survey have either twelve orthirteen school grades altogether; one has Concern with Quality of Provisioneleven and one has fourteen. All the sys-tems are divided into primary and second- In recent decades, in each of the four-ary sectors. At the primary level, one coun- teen countries reviewed, the challenge fortry has five grades, seven have six grades, Ministries of Education was to providefour have seven grades, and two have eight schools and teachers for a dramatically in-grades. At the secondary level, one country creasing population. To date, the emphasishas four grades, four have five grades, five appears to have been on quantity rather thanhave six grades, and four have seven grades on quality, with the result that concern has(Appendix 1). been expressed with many aspects of the
quality of educational provision. In manyKenya has the largest number of pupils countries, teacher morale, for instance, ap-
in primary schools followed by Ethiopia, pears to be low. In Ethiopia, one studyUganda, Madagascar, and Zambia (Table reported that 40 percent of primary and 762.2). In 1986, the gross primary enrollment percent of secondary teachers would, ifratio of the countries varied from 29 percent given the opportunity, abandon the teach-(Guinea) to 115 percent (Lesotho) (Table 2.2). ing profession (ERGESE 1986). In LesothoThe gross enrollment ratio for six of the in 1988, 805 teachers and almost 55,000 pu-
16
pils had no shelter. In four of the fourteen In its policy study, Education in Sub-countries, class size has increased between Saharan Africa, the World Bank (1988) con-1983 and 1986. The pupil-teacher ratio in cluded that "... the safest investment in edu-Mauritania has gone from 24:1 (1970-71) to cational quality in most countries is to make5X.:1 (1986-87). Class sizes as large as 120 sure that there are enough books and sup-have been reported for Togo. In many in- plies" (p. 4). Other studies have shown that,stances, attendance rates are poor. Children in the case of developing countries, factorsmay be absent to help their parents either in such as school facilities, textbook availabil-the home or with planting or harvesting. ity, and teacher training account for a large
Table 2.2 portion of the variation in student achieve-ENMROLLMENT AND PARTICIPATION RATES ment (Eisemon 1988; Fuller 1987; Heyneman
and Loxley 1983).
Primary Secondary Resources that are lacking include suchbasic items as desks and chairs as in Togo.
Countries/ Gross Gross Many of the national reports commented onLngiuistic Number enrollment Number enrollment the lack of basic materials, most notably text-
atus (DODS) %femnale ratio (0COs) %femnale ratio books. In Madagascar, up to ten students
share a textbook. Over a sixteen-year pe-Anglophone riod, education in Guinea was conducted
Kenya 5124* 49* 96* 540* 40* 20 practically without any textbooks. Many
Lesotho 330* 56 115 39 60* 24 textbooks used in Cape Verde are producedMauritius 145 49 107 71 47 51 in Portugal, primarily for the home market,
Mand do not provide an adequate coverageSwaziland 142 50 105 32 50 42 of the official local syllabus. In Kenya andUganda 2204 45 68 187 33 9 Madagascar, as well as in other countries,
Zambia 1366 47 96 145 37 18 basic science equipment is either in shortsupply or, in some instances, is nonexistentin secondary schools.
Francophone
Chad 341 28 43 44 16* 6 Resource availability has implicationsGuinea 270 31 29 77* 25* 15* both for the format and functioning of pub-
lic examinations. In particular, practical ex-Madagascar 1492* 48 124 357* 44 42 aminations in subjects such as woodwork,
Mauritania 157* 40* 51 37* 30* 16 metalwork, and home economics requireRwanda 904 48 65 21 34 3 basic materials, not alone at examination
time but also during the school year. LackTogo 511 38 102 78* 24 21 of appropriate resources is one of the main
impediments to the introduction of more
Other appropriate forms of assessment (essay,Cape Verde 66* 49 108 7* 30* 14 practical, oral, and aural) in Ethiopia. Prob-
lems have been experienced in Chad andEthiopia 2884* 39 38 843* 39* 14 Swaziland, as well as in other countries, in
the acquisition of equipment for practical
Source: World Bank (1990). subjects.
* Updated in national report.
17
Scarcity of Places educational systems that cannot accommo-date all children of school-going age, repre-
The percentage of students progressing sent a waste of scarce resources. While somefrom the final grade of primary to the first repetition is to be expected for reasons suchgrade of general secondary education is in- as illness or low achievement, the relativelyfluenced greatly by the number of available high levels reported for the countries in thissecondary school places. Students complet- study (Table 2.4) is cause for concern. Rep-ing primary school compete against each etition ranges from 2 percent of total enroll-other for the relatively small number of ment in Zambia to 35 percent in Togo. Onplaces at the higher level. At one extreme,80 percent of Ethiopian primary-school Table 2.3graduates transfer to secondary school while, PERCENTAGES OF STUDENTSat the other, as few as 4 percent of Rwandan PROGRESSING FROMstudents with the same attainment level do THE LAST GRADE OF PRIMARYso. However, in the former case, it should TO FIRST GRADE OFbe noted that the gross primary enrollmentratio is very low (38 percent). For eleven of Countries/the thirteen countries for which transfer data linguisticwere available (Table 2.3), less than 50 per- status Percentagecent of students advanced from the lastgrade of primary to the first grade of sec- Anglophoneondary school. In Uganda, for example, Kenya 28*approximately one-third of the students Lesotho 26*completing the primary-leaving examination Mauritius 43can expect to be accommodated at second- Swaziland 67ary level. In Kenya, of the 341,000 who sat Ugand NAfor the primary-leaving examination (KCPE) Zambia 23+in 1987, 51 percent were successful in ob-taining places at the next level. In Malawi Francophone(examined in an earlier study), in 1986-87, Chad 19the number of places in Form 1 of the sec- Guinea 27ondary system only amounted to 7.6 per- Madagascar 28cent of the total Grade 8 enrollment Mauritania 29(Kellaghan and Greaney 1989). The bottle- Rwanda 12**neck between primary and secondary school Togo 34gives rise to great pressure on students toperform well on the primary school-leaving Otherexamination, that normally serves as the Cape Verde 45selection test for secondary school. Ethiopia 80
RepetitionSource: World Bank (1990).
Traditionally, grade repetition has been * 1970.
used as an indicator of educational ineffi- Data for an adjacent year.ciency. High repetition rates, especially in ** 1989-90.
18
Table 2.4 Dropout
REPEATERS AS A PERCENTAGE OF Dropping out of school represents aTOTAL ENROLLMENT, 1986 waste of human and financial resources
Countries/ unless students acquire basic skills duringlinguistic their time at school. There is a considerablestatus Primary Secondary body of evidence to indicate that large per-
centages of students fail to reach the final
Anglophone grade of primary school in Africa. Some ofK;enya 13 NA this evidence has taken the form of report-Lesotho 23 NA ing the proportion of a cohort that reachesMiauritius 7 14 the final grade.Swaziland 15 8 Available data indicate that in almost half
Zganda 14 NA of the countries in our study, fewer than 50
percent of those entering primary school per-
Francophone sist to the final grade of the primary schoolChad 24 ll (Table 2.5). In Lesotho, it has been estimatedGuinea 25 38 that about one-quarter of those entering theMadagascar 29* 20** lowest level complete primary school (Le-Mauritania (1987-88) 19+ First cycle 11+ sotho, Ministry of Education 1987, p. 13).
Second cycle 16+ Recent Ethiopian data highlight the ex-Rwanda (1987-88) II+ 6+
tent of early dropping out of school. In 1987,ogo 35 32 the enrollment figure in Grade 2 (N: 440,515)
Other was less than half that in Grade 1 (N:Cape Verde (1988) 27 15+ 942,541). In Grade 3 (N: 350,197), there wasEthiopia 10 11 a further drop of almost 100,000 pupils.
While grade repetition would account forsome of these differences, early dropout ap-
- 1970. pears to be the main reason for the sharp
** Adjacent year. decline in enrollment figures.
+ National Examinations Report. In Lesotho and Swaziland, the highest
the whole, repetition at secondary level, dropout rates at the primary level were re-though still substantial, is lower than at pri- corded at the early and final grade levels.mary level. Data for Sub-Saharan Africancountries in general show that the rate of Evidence from a number of countries sug-repetition for Francophone countries is ap- gests that the dropout rate is higher at theproximately four times that for Anglophone secondary than at the primary level. In Togo,countries. Among the countries in our for instance, in 1986, the dropout rate at pri-study, the highest repetition rates were re- mary level was 10.2 percent; it was almostcorded for four Francophone countries: twice as high (19.4 percent) over the first four-Togo, Guinea, Chad, and Madagascar. year period of secondary schooling.
19
Table 2.5 reason for not continuing in school was theDROPOUT: PROPORTION OF non-availability of a higher grade in the area
COHORT (N: 1000) in that they lived (Swaziland, Ministry ofREACHING THE FINAL Education 1986). We cannot, of course, as-
PRIMARY-SCHOOL GRADE, 1986 sume that reasons for dropout would be thesame in other counties.
Countries/ Dropoutlinguistic Index Teacher Qualitystatus
Most of the educational systems in ourstudy have large numbers of untrained
Kgenya NA teachers. For example, the majority of pri-
Lesotho 517+ mary teachers in Cape Verde have no for-Mauritius 959+ mal qualification. Frequently, those who failSwaziland 615 to gain admission to teacher training on theUganda NA basis of performance on a series of achieve-Zambia 801+ ment tests are offered teaching positions in
areas where there is a shortage. In Lesotho,
Francophone 19 percent of primary and 16 percent of sec-Chad 172 ondary teachers have no formal teachingGuinea 437+ qualification. In Ethiopia, approximately oneGuineadaga NA in four primary and secondary teachers have
Mauritania 923+ had no professional training; in fact, a totalRwanda 486+ of 11 percent had less than twelve years ofTogo 592+ formal school experience. At the secondary
level, one in five 7th and 8th Grade teachers
Other had a college qualification, while fewer thantape Verde 340+ one in two of those teaching the final two
Ethiopia 496, years of secondary school had a universityEthiopia 496+ degree (ERGESE 1986). Madagascar authori-
ties, faced with the difficulty of recruitingSource: World Bank 1990. teachers for rural areas, have filled vacan-+Data for adjacent year. cies with men and women doing their na-
tional service. While the percentage of quali-fied teachers has improved dramatically in
In a Swazi study of reasons for dropping recent years in a number of countries (mostout at the primary level, students who notably Swaziland), the Kenya teaching forcedropped out and their parents cited finan- had a lower percentage of trained primarycial difficulty as the most important cause. teachers in 1988 than in 1963. Like most otherTeachers, on the other hand, attributed drop- countries, Kenya has experienced a shortageout to poor school performance and lack of of teachers of specialized subjects.interest in studies, though they accepted thatfinancial difficulty played a role. With one While entry qualifications to teacher-exception, similar reasons were advanced for training colleges vary considerably, theydropout at the secondary level; the excep- generally tend to be low. Primary pre-ser-tion was the daim of dropouts that the main vice teachers in Mauritius embark on a two-
20
year teacher-education program after pass- amounts to around $26 per month. It ising the senior cycle examinations. In Mau- evident that in this instance a teacher's sal-ritania, candidates for the one-year teacher- ary cannot support a family. In Chad,training program should have passed the unsalaried teaching assistants are often sup-Baccalaureat; those without this academic ported by the villages. In Uganda, the veryqualification may still opt for teacher train- low level of state salaries necessitates theing by undergoing a three-year course. In raising of most of a teacher's income throughMadagascar, on the other hand, entrants to direct parents' contributions and local fund-teacher training for primary school are re- raising efforts.quired to have passed only the junior-cycleexamination (CFEPCES) of the secondary It may be important to distinguishschool. They are then required to undergo teacher quality from teaching quality. Thea two-year training period and a further year findings of research relating to indices ofof practical training. teacher quality, such as teacher education
and teacher experience, are not consistent.Low salary levels are perceived to con- In some studies they have been found to be
tribute to the low status of teachers though related to student achievement, in others not.teachers' salaries in Africa are two to three In general, however, it seems that formaltimes higher relative to income per capita teacher education is less effective in Africathan in Asia or Latin America (Mingat and than in other developing-country regions.Psacharopoulos 1985). The median salary Further, the effects of teaching experiencearound 1986 for teachers in countries in the have generally been found to be less posi-study for that we had data was close to tive than the effects of formal teacher edu-$1,400 for primary teachers and $2,200 for cation. However, in analyses of IEA datasecondary teachers. But there were collected in Swaziland, while teacher qual-exceptions to this. In Madagascar, teacher ity (education and experience) did not cor-salaries generally range from $45 to $63 per relate with student achievement (when stu-month. The value of this salary is brought dent background and school characteristicsinto sharp focus when one considers that were controlled), the effects of teaching qual-rice costs an average family about $28 per ity (expressed in terms of teaching processmonth, while rent for a simple apartment variables) were manifest (Lockheed and
Komenan 1989).
21
4 Public ]Examinations
In this chapter, we provide a description fered a range of other public examinationsof the examination systems of the fourteen in the areas of vocational education andcountries included in our study, based on teacher training. The numbers of candidatesthe reports prepared for the individual coun- taking these examinations are small com-tries. pared to the numbers taking public exami-
nations at the primary and secondary lev-Tradition els. Vocational and teacher-training exami-
nations were considered to be beyond theExamination systems in African coun- scope of the present study. In general, pub-
tries, not surprisingly, have been strongly lic examinations are offered at the end ofinfluenced by traditions in the European primary schooling, midway through thecountries that colonized them. Following secondary school, and at the end of second-traditions that are both British and French, ary schooling. Students in Francophoneexaminations are formal, terminal, and sub- countries tend to be subjected to more ex-ject-based. Countries in Francophone Af- aminations than those in Anglophone coun-rica use the French Baccalaur6at system at tries. In particular, additional examinationsthe end of secondary school, while Anglo- are administered during the primary schoolphone countries have had close ties with years in Francophone countries. In a num-British examining boards. In fact, the final ber of Francophone countries, also, a com-school-leaving examination (GCE) in two of petitive examination, termed the concours, isthese countries continues to be set and used to select pupils for the next highestmarked in Britain. Anglophone countries educational level.have also tended to use features of Britishexaminations such as moderators. In recent Written essay examinations are offeredyears, national examination systems have be- in most countries. In two countries, how-come more independent and have tended ever (Ethiopia and Zambia), the multiple-to develop their own examinations. The choice format predominates. The broad ar-problem of processing very substantial num- ray of examination formats offered also in-bers of examinations has led to an increas- cludes practicals, orals, and aurals. Anglo-ing; reliance on multiple-choice tests, favored phone countries in general tend to use a va-by educational systems in the United States. riety of examination types. In each of these
countries practical tests are used in publicTitles and Descriptions of Examinations examinations. Virtually all use essay and
multiple-choice items in at least one publicA total of fifty-two public examinations examination. Four Anglophone countries in-
offered at primary and secondary (general) dude aural components in their public ex-levels of education was identified. In addi- aminations. Other formats (oral, projects,tion to these, many examination bodies of- short-answer) are used less frequently. Fran-
23
cophone countries tend to opt for the writ- Second, in Francophone countries, inten and oral formats. general, a less formal approach is used that
allows for much more direct input fromFor each of the fourteen countries in our teachers. In Guinea, for instance, secondary-
study, details of examination titles, types, school teachers are asked to submit topicsfunctions, and the stage of education at for examination. A commission adds its ownwhich they are administered are given in topics from a data bank. The Minister thenAppendix 2. The name of the agency that makes the final choice of questions. Teach-has responsibility for each public examina- ers who set the examinations may havetion is also given. unrealistic goals and expectations. In
Rwanda and Mauritania, the Minister isSize of Examination Enterprise closely involved in the examination-setting
process; concern over test security is the pri-The primary school-leaving certificate ex- mary reason for the adoption of this proce-
amination is clearly the most significant in dure. The ability of teachers to judge thenumerical terms (Table 3.1). Kenya appropriateness of items or to estimate the(N:354,802), Ethiopia (N:281,734), Zambia time required to answer individual papers(N:197,000), and Madagascar (N:192,405) has attracted critical comment. In Rwanda,have the largest numbers of candidates for for example, an examination for which athis examination. In all but two (Ethiopia three-hour time period was set aside wasand Swaziland) of the fourteen countries the completed in a half-hour by the candidates.number of secondary junior-cycle candidatesin the system is less than half the number of Third, in some countries (Lesotho andprimary candidates. Over 100,000 take the Zambia) formal courses for examiners aresecondary school-leaving examination in organized by examination councils. In 1987,Ethiopia (N:175,277) and Kenya (N:131,000). in Lesotho, Primary Certificate multiple-In the case of all but these two countries, choice items were written by teachers andthe senior-cycle school-leaving certificate subject specialists in specially organizedcandidates total fewer than one-quarter of workshops. Final examination items werethe number of primary candidates. In ap- selected on the basis of their degree of dis-proximately half of the countries for which crimination and difficulty, and on the judg-data were available, fewer than 5,000 candi- ment of subject specialists.dates took the senior-cycle terminal exami-nation. In a small number of countries, the writ-
ing of items for the terminal secondary-Setting of Examinations school examination is entrusted to univer-
sity personnel. In Chad and Ethiopia, itemsThree distinctively different approaches are written by staff members from the Uni-
are used for writing items and setting ex- versity of Chad and the University of Addisamination questions in the countries re- Ababa respectively.viewed.
Use of Aptitude TestsFirst, in a number of countries with de-
veloped national examination centres, such Scholastic aptitude tests have not provedas Kenya, Swaziland and Mauritius, much popular in the context of public examina-of the work is conducted by staff members tions in Africa. This is perhaps because theyworking closely with others, usually teach- have not been very popular in Europe ei-ers. In these countries, multiple-choice is ther, as compared, for example, with thethe preferred examination format. United States. There are a number of other
24
Table 3.1SIZE OF EXAMINATION ENTERPRISE
Countries/linguisticstatus Primary Secondary
AnglophoneKCPE KCE KSE
Kenya (1988) 354,802 (Discontinued, 131,000,1987)
PSLCE JCE COSCLesotho (1987) 23,998 6,642 2,790
CPE SC HSCMauritius (1989) 33,083 11,141 3,977
PC JCE COSCSwaziland (1987) 13,943 7,500 (1988) 2,561
PLE UCE UACEUganda (1989) 134,669 42,507 13,248
G7 G9 SCZambia (1987) 197,000 59,933 15,343
Francophone
CEPE/CEPEA Concours BEPC BacChad (1988) 17,840 20,550 7,985 NA
7th yearEntrance BEPC Bac
Guinea (1988) 24,668 12,148 (i) 9290(ii) 4209
CEPE CFEPCES BacMadagascar (1988) 192,405 84,635 47,624
1st yearCEE Secondary BEPC Bac
Mauritania (1988) 23,680 23,680 3,463 7,361
Rwanda CEP Concours Examens de reclassement57,089 2-2,500
NA End of 3rd NAyear 3,500
CEPD BEPC BacTogo (1987) 46,364 18,190 (i) 5163
(ii) 5610
OtherEnd of Basic First Cycle Second Cycle
Elementary Complementary General ComplementaryCape Verde
NA NA NA NA
G6 G8 ESLCEEthiopia (1.988) 281,734 237,011 175,277
25
possible reasons for their lack of popularity. reasoning tests are administered in conjunc-First, curricula in schools are more closely tion with four subject achievement tests inprescribed in Africa (as they are in Europe) the primary school-certificate examination.than in the United States. Second, in a num- In a small-scale study in the 1960s, it wasber of studies, scores on scholastic aptitude found that ability tests (measuring verbaltests have been found to predict performance reasoning, nonverbal reasoning, and numeri-in third-level institutions less well than sub- cal skills), that were administered in addi-ject-based examinations (see O'Rourke, tion to, but not as part of, the secondary-Martin, and Hurley 1989). Third, the use of school selection examination of achievement,an aptitude test might lead schools to em- were better predictors than the selection ex-phasize preparation for this kind of test at amination of performance on the Juniorthe expense of covering material designed Certificate national examination taken twoto achieve curricular objectives. years later (Irvine 1968). In further studies
of the predictive validity of the examina-It can be argued that a scholastic apti- tions, reported in a note prepared by the Psy-
tude test may still have certain advantages chological Service of the Ministry of Educa-in certain situations. First, being neither tion, performance on the verbal-reasoningcriterion-referenced nor based on a single test was the best predictor of success at sec-concept of subject excellence, it could allow ondary school. Performances in the Englishlocal school authorities to experiment with certification examination and on the nonver-curriculum. Second, it might have a par- bal reasoning test were the next best predic-ticular value in countries in which school tors. In particular, and not unexpectedly,inputs and facilities vary widely, insofar as the spatial component of the nonverbal testit is less sensitive than achievement tests to was found to have "a very high correlation"the differential opportunities of students to with performance in mathematics and physi-learn (Heyneman 1987). cal sciences. Furthermore, performance on
the nonverbal reasoning test was found toIn Africa, two countries used aptitude correlate "very well with success at techni-
tests for a number of years but discontinued cal courses at tertiary level." Unfortunately,the practice. An aptitude test was adminis- we do not have more details of these stud-tered as part of the Ethiopian School-Leav- ies-for example, the magnitude of the cor-ing Certificate Examination between 1967 relations between performance on the apti-and 1974. In Lesotho, verbal reasoning and tude tests and later school performance-butnumerical ability tests were introduced in further information about this researchthe 1970s to be used in conjunction with the would obviously be of considerable interestachievement tests of the primary-school cer- in the context of the selective functions oftificate examination to select students for sec- public examinations.ondary school. Sebatane (1985) reports thatthe "predictive validity" of the tests was School-based Components of"professionally established" but does not re- Examinationsport any details. The results of the abilityand achievement tests were combined, a A few systems currently include assess-practice described by Sebatane as "unten- ments conducted by students' teachers asable." The practice of administering the abil- part of the final examination grade. Theity tests has been discontinued. junior-secondary cycle examination in Le-
sotho has elements of school-based assess-The most consistent use of aptitude tests ment, while in Ethiopia, a total of 30 per-
has been in Zambia. Verbal and nonverbal cent of the total marks for the primary
26
school-leaving examination (G6) is awarded In other countries, notably Kenya, Le-for school-based assessments. sotho, Mauritius, and Zambia, multiple-
Despite the potential advantages of this choice tests are machine-scored. Ethiopia'sform of assessment, especially from the va- ESLCE has been machine-scored by the Uni-lidity perspective, a number of problems versity of Addis Ababa since 1977.have been observed in its use. A lack of Ethiopia's other national examinations arevariance in school-based marks has been scored by large groups of university studentsnoted; marks tend to be inflated. Some or members of youth organizations.schools fail to submit students' school-basedassessment marks in time. In such instances, Systems that require students to completecandidates in Ethiopia are awarded the same projects and/or administer practical exami-percentage marks for school-based work as nations tend to have assessments conductedthey achieved in the public examination. by outside teachers. In this case, and alsoSince school-based marks are usually higher where practicals are conducted by the stu-than examination marks, these candidates dents' own teachers, the process is oftentend to be penalized due to the failure to monitored by the school inspectorate. Areturn school-based marks. number of examination systems (Lesotho,
Mauritius, Swaziland) have their senior-cycleMarks derived from school-based assess- secondary school-leaving examinations set
ments, at one time used to contribute to the and scored by an agency outside the coun-total marks for the Baccalaureat in Guinea, try (the University of Cambridge Local Ex-are no longer used because serious discrep- amination Syndicate and London Universityancies were found between the school-based Schools' Examination Council).and non-school-based elements of the ex-amination. Discontinuation of the practice No formal reliability studies of public ex-in Guinea has contributed to an observed aminations appear to have been conducted.increase in discipline problems, less regular To ensure the reliability of the scoring pro-attendance, and a nonchalant attitude to- cedure of essay-type examinations, severalwards school examinations. different procedures have been used. In
Togo, papers are corrected once except inScoring "dcases where the score distribution shows a
large deviation from the norm alreadyIn most systems, scoring or marking of known by the committee." (Unfortunately
state examinations is carried out by practic- the country report does not indicate theing teachers. In fact, this exercise is consid- source of this norm or elaborate on the termered a most useful form of in-service teacher "score distribution"). In such instances, pa-education for its participants since it high- pers are recorrected. In Mauritania andlights the essential elements of the curricu- Madagascar, each paper is corrected twice;lum, standards expected, and marking pro- in Madagascar, the average mark is assignedcedures. In Cape Verde, scoring is entrusted except where the difference is "very large,"to the candidates' teachers. Responsibility in which case the paper is scored a thirdfor scoring some public examinations is as- time. The scoring process adopted in Guineasigned to regional or local committees of clearly indicates that factors other than levelteachers in Togo and Madagascar. In Swazi- of performance on the set questions in theland, moderators are used to scrutinize a examination affects a candidate's final mark.sample (10 percent) of scripts marked by Scorers are made aware of candidates' schoolteachers and to recommend, if necessary, grades prior to correction. According to thechanges in the scoring. national report "these records allow the jury
27
to more fairly mark those pupils who have to be discontinued as the process producednot received a complete education covering too many errors. In Ethiopia, the inabilityall of the questions posed at the examina- to obtain appropriate scanner parts due to ations." It seems reasonable to assume also lack of appropriate arrangements for secur-that a student who performed well on the ing foreign currency resulted in a return toexamination could have his or her mark re- hand scoring of the Grade 6 and Grade 8duced in line with expectations derived from national examinations.school grades.
Functions of ExaminationsTechnology
In theory, public examinations are ex-The use of computers and optical-scan- pected to serve a number of different func-
ning machines allows for the processing of tions. For teachers and pupils, they can actlarge amounts of examination data in a short as an incentive for greater effort. They areperiod. Furthermore, if multiple-choice regarded as a fair means of distributingitems are used, computerized item analysis scarce educational benefits. They can pro-facilitates the provision of feedback infor- vide feedback to teachers and students onmation to schools. their strengths and weaknesses. Public ex-
aminations can help unify the teaching ef-To date, the use of computers and opti- fort throughout an entire country around
cal scanning devices appears to have been common goals and themes. For students,confined in the main to Ethiopia, Kenya, Le- they provide formal evidence of educationalsotho, Swaziland, and Zambia. Dependence achievement (certification). Public exami-on modern technology is most pronounced nations may also serve an accountability rolein Kenya. Over 360,000 primary-school leav- by offering evidence of standards attaineding-certificate examinations are scanned and by individual teachers and schools. Lastly,analyzed in a year. Each year, through its and most important in the present context,KCPE Newsletter, schools are informed of the educational systems, governments, and otherpercentage of candidates in the country who employers use examination results for se-answer each item correctly and in many lecting appropriate candidates for the nextinstances the percentage opting for each level of schooling or for employment.distractor. Computers are also used to cal-culate national order of merit lists. Selection
Some examination systems rely on out- In practice since educational opportuni-side agencies for computer services. Pro- ties form a pyramid with fewer places avail-cessing of the Ethiopian ESLCE is done by able at each successive level, the primarythe University of Addis Ababa; in Zambia, function of almost all public examinationsmachine-scoring of multiple-choice tests is reviewed in the fourteen countries in ourundertaken by the Department of Finance, study is selection. While most of the exami-while other processing, including the pro- nations also serve a certification function,duction of certificates, is done at the Uni- some, most notably the concours in Franco-versity of Zambia. phone countries, are designed solely to se-
lect students for the next highest level of theSwaziland and Ethiopia have encoun- educational system. The highly competitive
tered serious problems that have thwarted concours is used, for instance, in Togo at thetheir efforts to use modern equipment to end of the secondary junior-cycle to selectscore large numbers of examinations effi- students for the three-year senior-cyde pro-ciently. In the former, machine-scoring had gram and again at the end of this cycle to
28
select candidates for colleges and universi- retention, and expulsion cut-off points.ties and for foreign study. Thus, it is reasonable to assert that a stu-
dent who was expelled from school on theThe capacity of educational systems basis of examination performance could
rather than the academic achievement of have merited either retention or promotionstudents plays the crucial role in determin- had an accurate measure of student achieve-ing the percentage of pupils who advance ment been available.to the next highest level. Lack of schools,especially secondary schools, inevitably re- Since the selection function of examina-sults in a severe bottleneck at the end of tions relates to the identification of studentsprimary school. The number of selected or who are most likely to succeed at the nextpromoted students depends ultimately on highest level, the lack of studies of the pre-the number of available places. Pressure for dictive validity of examinations is surpris-a place in a secondary school is most pro- ing. Clearly it would be possible to increasenounced in Rwanda, Uganda, Chad, and the efficiency of the selection process byZambia. In Ethiopia, on the other hand, using shorter tests. Such tests would re-little selection pressure at the end of pri- quire less supervision and less time in mark-mary school is due to the ready availability ing and in analysis and so could result inof places; the vast majority of students can substantial financial savings. Overempha-transfer to secondary school. Here, as in sis on increasing the efficiency of the selec-the other cotntries, however, the selection tion function, however, would inevitably runfactor is emphasized at the end of senior the risk of reducing the content coverage ofcycle. Evidence for this claim is provided the examinations (see Chapter 5). As a con-by the large number who repeat the ESLCE sequence, we would expect that over timein the hope of improving their grade-point sections of the curriculum not being exam-average to qualify for one of the relatively ined would be neglected in schools. Thefew available places in third-level educa- combination of selection efficiency and cur-dional institutions. ricular validity in examinations involves a
delicate balance.In a number of Francophone countries,
rigid marking systems are employed at end- Certificationof-year grade examinations, administered toselect or to promote students to the next The emphasis on the selection functionhighest grade. Among the most severe is of examinations is so pervasive that there isthat employed in Chad secondary schools a danger of losing sight of their certificationin which a mean mark of ten or more is function. With the exception of the concoursrequired for promotion, greater than seven and those taken in some countries prior toand less than ten for repetition, while a mark the primary school-leaving examinations,of seven or less results in expulsion. Given virtually all of the examinations have a cer-what is known about the extent of marker tification function. Formal certificates thatunreliability in examinations, this procedure indicate candidates' academic achievementsobviously presents problems. If, for ex- may be important credentials for gainingample, a student obtained a score of 8.5 on employment or for qualifying for furthera test with a standard error of measurement training. In Ethiopia, for example, posses-of 0.8 we could say that the individual's sion of the Grade 6 and Grade 8 certificates"true" score lies between 10.1 and 6.9 with may be required as a minimum qualifica-a 95 percent level of confidence. This range tion for a range of occupations, for exampleencompasses each of the critical promotion, drivers and filing clerks. Lower level cer-
29
tificates, however, as the numbers possess- officials may congratulate officials of schoolsing them increase, are losing their currency in which students have performed well orin the labor market. require less successful schools to explain
poor examination performance.There has been a growing tendency in
recent years to use the same test to meet Performanceselection and certification requirements. InMauritius, for instance, prior to the intro- Attempting to summarize and provideduction of the present Certificate of Primary figures comparable across countries for ex-Education (CPE), a primary school-leaving amination performance (such as are pre-certificate served certification requirements sented in Table 3.2) is problematic. Such awhile a different examination (Junior Schol- summary fails to take into account differ-arship) administered around the same time ences in criteria for awarding grades or insatisfied the selection requirement. A real- the structure of performance (that may in-ization that the two formal examinations volve compensation across subjects or reach-contributed to excessive pressure on ten- and ing a particular level in specific subjects) thateleven-year olds led to the dropping of both can merit a pass or a particular grade. Evenexaminations in favor of one (CPE) that now within the same examination, different cri-serves both functions. teria can be used to indicate that a student
has "passed." For example, in the school-An examination may be an efficient se- leaving certificate examinations of a num-
lection device but may fail to serve the im- ber of countries, such as Kenya and Lesotho,portant certification function. Certification, to obtain a school certificate, a candidatewhich is recognized as a function of exami- must have passed in a specified number ofnations in all countries, is primarily con- subjects. However, candidates who do notcerned with providing a measure of curricu- obtain the specified number of passes maylum coverage. Adequate curriculum cover- obtain a certificate in the subjects which theyage is more likely to be provided in broad- did pass. In our table, we provide figuresbased examinations that extensively sample for the students who pass the total certifi-the prescribed syllabus than in examinations cate examination and, when available, fig-that are confined to narrow content domains, ures for candidates who failed to get a gen-which cover material that is normally not eral certificate but passed some subjects indirectly prescribed in school curricula, such the examination. Thus, problems in compa-as nonverbal reasoning ability, or that are rability can arise from the fact that what isconfined to one or two assessment proce- meant by a pass can vary from examinationdures. to examination, not only between but also
within countries. Further, in a number ofAccountability countries (Kenya, Zambia), a distinction is
not made between pass and fail on the pri-Public-examination results are frequently mary school-leaving examination.
perceived as providing evidence of schooleffectiveness. In a number of countries, To complicate matters, success or failureschools and teachers may be held account- in the eyes of students, parents, and teach-able for their students' achievements as re- ers may not be determined so much by ex-flected in performance on public examina- amination-performance level as by whethertions. The use of examination results to hold or not a student qualifies for a place at theteachers or schools accountable is magnified next highest level of the educational system.when results are published. Even when re- A mark of 80 percent on a primary school-sults are not formally published, education leaving examination may gain admission in
30
Table 3.2PUBLIC-EXAMINATIONS PASS RATES
Country/linguistic
status Primary Secondary
AnglophoneKCPE KCE KACE
Kenya (1987) Not applicable 78.0% 56.3%
PSLCE CE COSCLesotho (1987) 91.0% 65.1-68.8% 19A%-31.7%
(1980-87) (1980-87)62.2%-74.7%
(GCE)
CPE SC HSCMauritius 58.4% 58.1% 52.6%
PC JC COSCSwaziland (1988) 79.6% 67.3% 54.0%
PLE UCE UACEUganda (1987) 82.2% (1989) 80.1% 66.4%
G7 G9 SCZambia (1983) Not applicable 35.8% 63.6%
+ 32.4(GCE)
FrancophoneCEPE/CEPEA Concours BEPC Bac
Chad (1988) 56.8% 35.3% 49.8% 12.9% (1987)
7th yearEntrance BEPC Bac
Guinea (1988) 56.6% 48.8% (i) 21.0%(ii) 30.6%
CEPE CFEPCES BacMadagascar (1988) 49.0% (1987) 26.7% (i) 20.1%
(ii) 18.2%
CEE 1st year Sec. BEPC BacMauritania (1988) 22.0% 29.2% 37.6% 25.6%
ConcoursCEP for Sec. NA 93.2% (Public)
Rwanda (1988) NA 7-9% NA 70.0% (Private)
CEPD BEPC BacTogo NA NA NA
OtherEnd of Basic First Cycle Second Cycle
Elementary Complementary General Complementary
Cape Verde (1987) 67.3% 56.9% NA NA
Ethiopia G6 G8 ESLCE(1987 82.5% 68.3% 97.7% (Amhric)
51.9% (Math)
31
one country but may be insufficient to do so also recorded on the junior and senior-cyclein another where there is a severe shortage certificate examinations. In almost all coun-of secondary-school places. tries, pass rates were highest on the primary
school-leaving examination and lowest onMost systems tend to use some form of the secondary senior-cycle examination. In
preset achievement standard to award general, the secondary-school pass ratesmarks. While the percentage of students seem to have been slightly lower in Franco-getting various grades under this approach phone than in Anglophone countries.varies relatively little from year to year,notable variation does occur. In 1986 in Within individual countries, substantialGuinea, for example, 27.4 percent of candi- variation in performance rates has been re-dates passed the Grade 7 entrance examina- corded between different regions. For ex-tion. Two years later, this figure had more ample, the pass rate on the primary school-than doubled to 56.6 percent. leaving examination (CEPE) in Madagascar
was approximately 60 percent for candidatesA number of systems transform exami- living in cities and as low as 20 percent for
nation marks into normalized scores, for ex- rural candidates.ample Ethiopia and Lesotho. Marks aretransformed into standard scores so that Repeat students tend to lower the meanindividual students' reported scores reflect pass rates. In 1988, repeat students repre-their performance relative to the perfor- sented 59.4 percent of all candidates in Ethio-mance of other students. Such an approach pia for the ESLCE secondary leaving exami-can tell us little about what a pass means in nation. Their mean score (which was con-terms of a student's level of performance or siderably below a D grade) placed themknowledge. The standardization of scores lower than other candidate categories. Onensures that, irrespective of level of achieve- the 1983 Zambian Junior Secondary-Schoolment, percentages passing, failing, or gain- Certificate Examination, the success rate ofing distinctions are predetermined. Undue internal candidates (52.5 percent) was morereliance on normalized scores can lead to than six times that of external candidatesthe adoption of grading practices that bear (8.4 percent).little resemblance to educational realities(Sebatane 1985). For example, an indepen- Examination Repetitiondent study of the secondary school-leavingexamination in Ethiopia reported that the Policies regarding examination repetitioncut-off raw score that distinguished between vary. Students who fail public examinationspass and failure on three of the 1986 sub- in Ethiopia may repeat the year in schooljects, English, history, and chemistry, ap- provided there is sufficient accommodation;proximated the number of items that could otherwise they must study outside of thebe attributed to guessing the answer to each public-school system. Madagascar permitsitem (Cambridge Educational Consultants multiple repeats on its primary-leaving ex-1988). amination. On the other hand, Guinea and
Mauritius limit the number of repeats whileThe summary data in Table 3.2 reveal Zambia and Lesotho require the public-
that large numbers of candidates were un- school candidates to pass some public ex-successful in public examinations. In a aminations on the first attempt. Both thesenumber of countries, fewer than one in four countries, together with Kenya, Cape Verde,passed the primary school-leaving examina- and Mauritius accept repeat private candi-tion. High failure rates (60 percent +) were dates.
32
Governance and Administration There is a trend in Anglophone coun-tries, though not in Francophone ones, to-
Public examinations are organized and wards the devolution of responsibility foradministered either by the Ministry of Edu- the administration of examinations to ancation (or special sections within the Minis- authority other than the Ministry of Educa-try) or by an independent body established tion. However, even in countries in whichspecifically for this purpose. In general, in an examinations authority has responsibil-Francophone countries, responsibility is ity for examinations (for example, Ethiopia,vested in the Ministry, while in a number of Lesotho and Zambia), the authority stillAnglophone countries, it is vested in an in- maintains close links with the Ministry anddependent examination board or council. In may be dependent on other bodies for a va-a relatively small number of countries, such riety of services, such as marking of testsas Ethiopia and Togo, the Ministry has re- and computer facilities.sponsibility for primary and junior-cycle sec-ondary examinations, but the senior-cycle Localization of Examinationsschool-leaving examination (or part of it inthe case of Togo) is administered by univer- A trend in countries in which studentssity authorities. Appendix 2 identifies the take foreign examinations as part of theiragency or body that has been given respon- normal schooling is to localize all examina-sibility for individual public examinations tions, that is, to have examinations controlledor shares responsibility with external exami- within the country. There are only threenation organizations, for example the Uni- countries (Lesotho, Mauritius, and Swazi-versity of Cambridge Local Examination land) in which the leaving-certificate exami-Syndicate. nation at the end of secondary schooling is
designed and marked abroad (in Britain),Examination systems in Kenya, Lesotho, and steps are being taken to change that situ-
Swaziland, Mauritius, Zambia, and Uganda ation. Localization frees foreign currencyare run by independent boards. Member- for other activities, though whether theship of these boards tends to be drawn from money will be made available for improv-a broad educational constituency. Ministries ing the resources available for examinationsof Education are represented by one or more is another matter. It also, however, raisesmembers. The membership of the Uganda questions about the maintenance of stan-National Examinations Board, for example, dards and the acceptability of a country'sincludes four ex officio members: the Vice- educational qualifications in other countries.Chancellor of the University, the Chief Edu- The localization of examinations, particularlycation Officer, the Chief Inspector of Schools, at the end of secondary school, has beenthe Director of the National Curriculum viewed with apprehension in some coun-Development Centre; seven elected members tries in case standards should fall as a con-from professional bodies; and thirteen min- sequence. Governments, examination bod-isterial appointees (mostly principals of dif- ies, and the general public are very keen toferent types of schools and colleges). see that this does not happen and that their
school qualifications will have currency notIn general, examination sections within only nationally but internationally. To
education ministries and examination boards achieve this, it will be necessary for govern-publish syllabi and regulations for their ments to invest more in evaluation and re-examinations, appoint appropriate person- search activities relating to examinations.nel (markers, invigilators, etc.), and issue cer-tificates.
33
Funding country's deteriorating financial position andthe difficulty of raising foreign currency
Information from a limited number of place the continued operation of the exami-countries that have made appropriate data nation in jeopardy. Figures from Mauritiusavailable suggests that levels of state fund- for 1986-87 indicate that foreign examina-ing of public examinations vary consider- tion syndicates acquired $1.16 million forably. At one extreme, Madagascar devotes fees for secondary-level examinations; Cam-10 percent of its total educational budget to bridge University Local Examination Syndi-public examinations. Chad and Rwanda, on cate received $970,000 of this total. Otherthe other hand, spend less than 0.5 percent countries that pay substantial amounts ofon this exercise. Size of government expen- foreign currency, mainly for senior-cyclediture on public examinations expressed in leaving-certificate examinations, are Lesothomonetary terms or as a percentage of cur- and Swaziland.rent government expenditure on educationmay represent only part of the total exami- While the inevitable increase in candi-nation expenditure. Where Ministries of date numbers will increase the cost of fund-Education meet all examination costs as they ing public examinations, economies of scalearise, many, if not all of the costs, are ab- should lessen unit costs. In Kenya and coun-sorbed into the general operating budget of tries dependent on externally produced andthe Ministry. corrected examinations, local printing, ex-
amination production, and correction couldThe difficulty of estimating or conduct- reduce costs.
ing a comparative analysis of examinationcosts is compounded by the fact that the Some examination authorities raise fundsnumber of cost-incurring items differs through their publishing activities, notablyamong countries. In Cape Verde, for ex- by publishing copies of previously adminis-ample, teachers who construct examination tered public-examination papers.papers or who supervise at examinationsessions are not paid. Furthermore, at the Feesbasic elementary level, markers are paid onlyfor correcting external candidates' scripts. At In general, fees for end of primary-schoolthe secondary level, markers are paid by the examinations are either not charged or areschools. kept to a low level (see Table 3.3). This can
be attributed to government policies of en-In at least five of the countries reviewed couraging pupils to persist in school up to
(Kenya, Lesotho, Mauritius, Swaziland, and the end of the first or compulsory cycle.Zambia), public-examination bodies allocate Senior-cycle secondary-school examinationsconsiderable portions of their budgets to the are generally much more expensive for stu-purchase of foreign currencies. For security dents than those offered at the end of juniorreasons, Kenya devotes roughly 15 percent cycle. Candidates for examinations offeredof its total budget to external printing of its by overseas examination boards or syndi-primary and secondary examinations. In cates are required to pay high fees. The feeZambia, approximately 8,000 students take for the senior-cycle terminal examination inthe GCE 0-level examination each year, at a Lesotho ($124.19) represents considerablycost of $782,500 to the Examinations Coun- more than the mean monthly income percil. Since the amount that the Council is capita in the country.permitted to charge candidates is not nearlysufficient to meet the GCE costs, it must sub- Fees do not cover costs. In Kenya, forsidize the fee paid by each student. The example, although the National Examina-
34
Table 3.3CANDIDATE FEES FOR PUBLIC EXAMINATIONS
Countries/linguisticstatus Scale
AnglophoneKenya KCPE - $3.54; KCSE $42.49 (minimum)Lesotho PSLCE - $1.86; JCE - c. $32.09; Senior Cert/COSC -
c. $124.19Mauritius CPE - none; Second level examinations - $67.00 average
Swaziland PC - $1.25 per subject; JCE - Reg. fee - $3.33 +$2.92 per subject; COSC - Registration fee - $13.13 + $11.46
per subjectUganda PLE - $1.71; UCE - $13.14 (8 subjects); UACE
- $13.43 (3 principal, 1 subsidiary subject)Zambia None - G7 and G9 (except external); SC - $12.87 (8 subjects)
FrancophoneChad No feeGuinea No feeMadagascar CFEPCES - $0.19; Bac.- $1.58Mauritania NARwanda None apart from Examen de reclassement andLe Jury
Central - $12.53
Togo CEPD - $0.36; BEPC - $1.80; Bac.1 - $7.30;
Bac2 - $12.90
OtherCape Verde Basic elementary - none; Basic complementary -
$2.20; General secondary - $10.20; ComplementarySecondary - $2.00 per subject
Ethiopia Grade 6 - none; Grade 8 - none; ESLCE -Gov. schoolsfree; Others: Registration fee - $2.50 + $2.00 per paper
tions Council effects considerable efficiencies Securitythrough computer-scoring its primary leav-ing examination, its income from fees for the A marked feature of public examinationsexamination represents less than 50 percent in Africa is the pronounced emphasis on se-of the total cost. In Uganda, the ratios of curity. The security dimension is empha-unit costs to unit fee for the three major pub- sized throughout all stages of the examina-l1c examinations are 1.34:1 for the PCE, 1.42:1 tion process: setting questions, printing,for the UCE, and 1.44:1 for the UACE. proofing, packaging, distribution, supervi-
35
sion, and correction. The security emphasis facilities. Elsewhere, papers are printed insimply reflects the importance of examina- secure premises, often as near as practical totion success in the lives of students and their the date of the examination to avoid unnec-families. Success offers some hope of secur- essary storage problems or security breaches.ing valued scarce employment or of advanc- Distribution of papers is undertaken undering to the next highest educational level. Fail- secure conditions.ure, on the other hand, can lead to a rela-tively frugal existence or at best a lost year Some systems use identification numbersdevoted to preparing for the same examina- to ensure that markers do not have access totion. candidates' names. In the case of oral exami-
Commentators have noted "the ease with nations, care is taken to ensure that candi-which malpractices can occur" (Kelly 1986, dates do not hear the questions beforehandp. 552). In Togo and Madagascar, concern or come in contact with others who havehas been expressed that security is not well heard the questions. When there is evidencemaintained. In Zambia in 1980, it was re- of misconduct at an examination centre inported that candidates in three regions re- Uganda, the entire examination may be can-ceived help from invigilators during the GCE. celled. In Kenya, serious malpractice in rela-Copying has been a problem in Ethiopia. Col- tion to the conduct of an examination canlusion between candidates and supervisors, lead to a lengthy prison term.though rare, has been reported in Swaziland.Premature opening of envelopes containing Feedbackexamination papers has been noted inRwanda. In Uganda, types of examination Public examinations can be used to pro-irregularities reported include leakage, im- vide feedback to teachers, curriculum bod-personation, external assistance, smuggling, ies, and educational policymakers. Severalcopying, collusion, and substitution (Ongom of the countries in our review provide some1990). kind of mechanism to make information on
examinations available to interested parties.In general, however, examination authori-
ties are confident their security precautions Past examination papers are importantare effective. Machine scoring appears to sources of information in some countries andcontribute to test security because it reduces considerable use is made of them in schools.the personnel requirement for scoring. Con- Such papers, however, are not available ev-cern has been expressed, however, that due erywhere, such as in Lesotho. The wide-to the development of sophisticated devices, spread involvement of teachers from variousthe security of machine-scored data is by no parts of the country in item-writing, for ex-means guaranteed. In Ethiopia, the typing of ample in Lesotho, has also been an importantGrade 6 and Grade 8 examinations is en- source of information to schools about thetrusted to one typist. In many countries, skills and knowledge tested in examinations.people entering and leaving centres in whichexamination papers are printed are searched. Comparisons among Schools/DistrictsSupervisors in Zambia must take an oath ofsecrecy. Similarly, employees of the Kenyan Another source of feedback provided byNational Examination Council sign docu- examination authorities is the publication ofments binding them to the terms of the Offi- the examination results of students on acial Secrets Act. Papers for examinations in school-by-school or district-by-district basisKenya are printed in the United Kingdom (Ethiopia, Kenya, Rwanda, and Swaziland).due to the absence of secure local printing
36
Analysis of Results letter in their classes. For example, theyanalyze with their students the characteris-
Some examination systems have devoted tics of "good" and "poor" essays presentedresources to providing detailed analyses of in the newsletter.students' performance on public examina-tions. For example, in Lesotho, the test items Timelinessfor the primary-school mathematics exami-nation, together with performance item sta- Prompt release of examination results ististics for the population of examination can- necessary to allow for appropriate planningdidates, was sent to all schools. Item-analy- by schools and individuals. Results returnedsis information has also been prepared in after the commencement of the new schoolEthiopia, but the information was given only year represent an inefficiency, as they clearlyto Ministry of Education or Examination delay the assignment of pupils to appropri-Board personnel. Countries in which item- ate classes. The lag between the time theanalysis data have been used for post-ex- examination is given and the formal publi-amination reports are Kenya, Mauritius, and cation of results depends on the efficiencySwaziland. Chief Examiners may also pro- of collection procedures and the size of thevide descriptive reports on the overall per- examination enterprise.formance of students on individual exami-nations and even on individual examination Even within countries, the timeliness ofquestions. In some countries, such as Zam- feedback varies from one examination to thebia, every school receives a copy of the Chief next. Evidence from Chad suggests that theExaminer's Report. results of the CEPE and the Baccalaureat are
available as soon as one week after the ex-Information about examination perfor- amination while Grade 6 concours results
mance can be helpful in identifying aspects may take up to three months to be pub-of the curriculum that appear to have been lished.poorly taught or ignored by teachers. Wenmight expect the information to contribute A disadvantage associated with the Ke-to improvement in the construction of ex- nya KCPE Newsletter is that it does not ap-arninations and in future examination per- pear until the final term of the year follow-formance. ing the examination, thereby lessening the
potential value of its feedback for teachers.The KCPE Newsletter published by the Since, however, curriculum content tends to
Kenyan National Examinations Council is remain constant over the years, the overallan example of a systematic approach to- utility of the newsletter is probably not se-wards informing teachers about how stu- riously diminished.dents have performed on examinations. Thenewsletter, which is circulated to all schools, Challenges to Examination Resultscontains item--analysis data for each exami-nation paper. It also includes suggestions A number of examination bodies permitas to why students may have selected par- appeal of examination results. In both Le-ticular incorrect options. Examples of poor, sotho and Swaziland, appeals are allowed,average, and good answers to selected ques- on payment of a fee, for a period of six weekstions are presented and teaching implica- following the publication of the primary andtions are described. Teachers make frequent junior-cycle results. In Lesotho, if the ap-use of the information provided in the News- peal is successful, the candidate is refunded
37
half the fee. Ethiopian authorities discour- Outside-School Tuitionage appeals, claiming that the process is bothexpensive and time-consuming, but the As happens elsewhere throughout theBoard does recheck results. Appeals in Zam- world, students in examination classes fre-bia must be submitted through the princi- quently seek additional academic help out-pal teacher; recheck fees are not charged. side of school or in tuition offered in schoolsSimilarly, in Swaziland, requests for outside of regular school hours. The empha-rechecking the results of the Cambridge sis on private tuition is particularly pro-Overseas School Certificate Examination nounced in Mauritius. One study reportedmust be made through the principal. In this that 73 percent of Standard 6 pupils in theinstance, the examination authorities do little primary school-leaving examination classto encourage candidates to seek a recheck. took private tuition. As many as 11 percentIt is stressed in an explanatory note that re- of Standard 1 pupils also took private tu-ports will be limited to one subject per school ition. Tuition frequently was provided inon any one occasion of the examination and groups of twenty-five or more. Approxi-to the work of not more than six candidates mately 60 percent of secondary-school stud-in that subject. The service is not made ents also enrolled in tuition groupsavailable to private school candidates. (Joywathsing and others 1988).
38
5 Issues in Examination and Assessment
In this chapter, we consider trends that Expanding Enrollmentsappear in the examination systems of at leastsome of the countries included in this study Perhaps the most obvious trend in theand outline issues to which education au- development of examinations in recent yearsthorities might direct their attention when in many of the countries studied for thisconsidering possible areas for reform in as- report, has been the enormous increase insessment and examination procedures. the numbers of students sitting for exami-
nations. For example, between 1968 andIn considering these issues, sight should 1988, Swaziland registered an increase of
not be lost of the fact that considerable differ- over 200 percent in numbers taking its pri-ences exist between the examination systems mary-certificate examination and an increaseof the fourteen countries described in this of over 1,000 percent in the numbers takingreport. In highlighting issues, one can easily the junior-secondary school-certificate ex-give the impression that contextual factors amination. In Ethiopia, over a much shorterthat operate in the individual countries are time span (1979-87), primary-certificate can-less important than they actually are (see didates have shown a 240 percent increase,Little 1990). Of obvious importance are dif- junior-secondary certificate candidates a 260ferences in the educational systems of the percent increase, and senior-secondary cer-countries. For example, there are differences tificate candidates a 400 percent increase. Inin the numbers of students who are retained general, the higher the level of the examina-at varying levels and differences in the use of tion, the greater has been the rate of increase.a vernacular language for instruction. Bothhave implications for the assessment of learn- As we noted in Chapter 1, increases ining. Among the differences in examination the numbers taking examinations, as wellsystems are differences in traditions, not only as in enrollments generally, have been asso-in the colonial period but also in the post- ciated in people's minds with declining stan-independence period, which reflect consid- dards in education. This has been particu-erable diversity in the way in which national larly so when resources to deal with thissystems of examinations have developed expansion have not been available. The(Eisemon 1990). Examination systems also increased numbers have also forced a re-differ greatly in the use of technology , for consideration of the content of curricula inexample, in the use of computers and opti- schools. An academic type of curriculum,cal-mark readers. Such differences have im- which might have been suitable when ac-plications that go beyond the technology. For cess to education was limited, is not likelyexample, they affect the types of items used to be suitable when all or most of the popu-in examinations, methods of marking, costs, lation attend school. Curriculum change hasand the ease with which information on ex- obvious implications for the nature of ex-amination performance can be provided to a aminations, implications that relate not justvariety of publics. to content but to function. It is increasingly
recognized that the dominant selective func-
39
tion of examinations must give ground to According to one national report (Ethiopia),the certification function. Just as curricu- ... among the weaknesses observed inlum change has implications for examina- some papers were unclear instruc-tions, examinations may have implications tions, poor grammar, placing of ques-for curriculum. For this reason, changes are tions some pages from the passagebeing introduced to the examinations of on which they are based, use of in-several countries, including content and terrelated questions, (for example,skills appropriate to the wider range of the stem of a question prompts theachievements, needs, and interests repre- answer to an earlier question), andsented in the growing numbers of students inadequate supporting art work" (p.taking examinations, to stimulate beneficial 40).effect on curricular practice in schools.
Governments and examination bodies willExpanding enrollments will continue to be a have to address the issues raised by staff short-
pressing issue for many examination systems. ages and lack of trained personnel.For educational authorities they will lead to ademand for greater resources and to changes both Effects of Examinations on Curriculumin the content and function of public examina- and Teachingtions.
We have already noted in Chapter 1 thatResources for Examinations whenever the results of external examina-
tions are an important determinant of fu-Examination authorities appear to devote ture education and life chances or are used
virtually all of their time and resources to to evaluate teachers, the examinations willthe administration of the national examina- be taken very seriously by teachers and stu-tions. Frequently they are understaffed. dents. In this situation, teachers pay par-With a few notable exceptions, relatively ticular attention to the "tradition of pastlittle or no time is given to systematic analy- examinations." Having identified the con-ses to improve examination quality. Indeed, tent and skills covered in previous exami-most authorities do not have the research nation questions, they coach students in thecapability to undertake this task. Fre- identified content and skills. This can havequently, skilled personnel are not available positive effects on curriculum. For example,within the country. In a number of in- in Lesotho, the marks (25.0 percent) in thestances, however, people with formal train- Junior-Certificate Examination awarded foring in psychometric techniques have been practicals help to ensure that this aspect ofassigned to othfer sectors within education the curriculum is covered throughout theministries. In Zambia, staff shortages within school year. In Kenya, by including itemsthe Examinations Council have resulted in that relate to local conditions and that at-members of its Department of Research and tempt to measure higher-order cognitiveTest Development having to devote most of skills in the KCPE, the Kenyan National Ex-their time to item-writing and pretesting. aminations Council has drawn teachers' at-
tention to a new range of knowledge andIn general those responsible for the con- skills.
struction of examinations have had little orno relevant formal training. Many, of However, when people talk about thecourse, acquire skills through experience. "tradition of past examinations" definingHowever, weaknesses could be avoided if curriculum in schools, very often they areadequate training facilities were available. referring to negative effects of this tradition.
40
What they mean is that teachers are prob- efforts to obtaining good examination re-ably spending inordinate time in simply sults, "out of a concern for their pupils, andworking for the certifying tests, often at the to safeguard their own reputations" (Kelly,expense of subjects and skills not tested in Nkwanga, Achola, Kaluba, and Nilssonthe examination. 1986, p. 551). Teachers in Cape Verde make
extensive use of the examination papersThe narrowing effects of the public ex- from previous years; virtually all of the time
aminations on curriculum have been com- spent in school is devoted to preparing stu-mented on in most of the national reports dents for examinations. Practice in exami-and reinforce the findings of research in a nation-taking techniques is also a feature ofvariety of countries relating to the effects of primary education in Mauritius and Mauri-examinations that we outlined in Chapter 1. tania. In Ethiopia and Madagascar, routineIn Ethiopia, curriculum areas that tend to classroom examinations reflect closely thereceive relatively little attention in the class- item types used in the official examination.room include writing, vocational, oral and Teachers in Ethiopia and Kenya use the mul-aural skills, use of reference material, and tiple-choice format of the national examina-practical work in science. In Madagascar, tion in their regular classroom tests as earlysubjects such as music and physical educa- as the junior primary grades. Indeed, teach-tion, which are not examined, tend not to be ers have been observed to use the multiple-taught in primary schools. In Zambia, local choice format not only in examining but inlanguages, practical subjects, and production teaching, as early as in the first grade.activities are not formally examined and con-sequently are undervalued in schools. In The amount of time teachers in a num-Zambia as well as in Ethiopia, where the ber of Francophone countries are requiredprimary school-leaving examinations are to devote to formal assessment must of ne-limited to multiple-choice items, writing cessity limit teaching time. For example, inskills receive inadequate attention. In Cape Togo, Madagascar, and Guinea, teachers areVerde, the status of subjects such as visual expected to administer monthly, trimester,education, handicraft, and physical educa- and end-of-year examinations.tion is diminished as these are not testedwhen ascertaining who is to be promoted. Examinations are often regarded as anThe Mauritius report observed that the important source of motivation and as en-highly competitive Certificate of Primary couraging teachers and students to workEducation Examination has led to "a neglect hard. By targeting important elements ofof broader curriculum goals not included in the curriculum, they can focus student learn-the CPE syllabus" (p. 34). ing and may raise students' levels of achieve-
ment. And success on a public examinationThe importance of examination results is an achievement of significance not alone
for the pupil and for the teacher is reflected for students but also for parents and teach-in the amount of time spent fostering test- ers. The significance of passing the Kenya:aking strategies. A recent report from CPE examination in the eyes of one youngUganda observed that "the system is so candidate is well portrayed in an essay writ-exam ridden that the entire teaching and ten during the examination (Appendix 3).learning is geared to passing the exam andgetting the good marks needed for entry to Evidence from a number of countries,higher levels of education" (Uganda, Minis- such as Mauritania and Zambia, however,try of Education 1989, p. 7). In Zambia, indicates that examinations tend to promoteI even against their better professional judg- a passive concept of learning. Furthermore,ment," teachers feel compelled to direct their since scoring high marks in examnina-ions
41
becomes the dominating interest of pupils, ample in the assessment of practical or"pupils who cannot secure places at the next project work, or it could be more generaleducational level are stigmatized as failures. and overlap with elements of the externalThere is even some tendency to judge the assessment.personal worth of an individual on the ba- Examinations in Ethiopia (at the primarysis of his examination performance" (Kelly level) and in Lesotho (at the junior-second-1986, p. 26). ary level) already have elements of school-
based assessment. There is also a definiteExamination authorities must take into ac- commitment in a number of countries (Le-
count that while motivating students and teach- sotho, Swaziland, Uganda, Zambia) to makeers, important public examinations also help to school-based assessment an important fea-ensure that curriculum content likely to be tested ture of public examinations. However, thiswill be covered by teachers; that subjects that is being done with caution. For example, inare not tested tend to receive relatively little Zambia, the recommendation is that continu-teaching emphasis; that considerable class time ous assessment in the public-examinationis devoted to fostering test-taking strategies; and system should be confined to performancethat passive concepts of learning are promoted. in practical subjects (Zambia, Ministry ofIn addition, society at large and, in particular, Education 1977). A recent report in Ugandapupils themselves frequently equate examination contained a strong recommendation thatperformance with personal worth. school-based continuous assessment should
account for 20 percent of the marks allo-Reducing the Burden of Examinations and cated for each subject in the primary school-Introducing School-based Assessment leaving examination (Uganda, Ministry of
Education 1989). The assessment, whichIn a number of countries, efforts are be- would be moderated by the school
ing made to reduce the burden of examina- inspectorate and officials of the Uganda Na-tlions. Three broad strategies to achieve this tional Examinations Board, would includecan be identified, although it is not usually assessment of performance in class, of par-clear in the case of individual countries what ticipation in community projects, and of cre-strategy will eventually be adopted. ative activities. In Swaziland, it is hoped
that school-based assessment will eventuallyOne strategy involves the reduction of receive a weighting of 50 percent of the to-
the number of public examinations through tal mark in the primary-certificate examina-the abolition of examinations and certifica- tion.tion at particular points in the educationalsystem. For example, in Kenya, public ex- A third possible strategy is to base certi-aminations and certification no longer oc- fication entirely on school-based assessment.cur at the end of the junior cycle of second- This is not envisaged in any country in ourary schools, while other countries (Cape study in the immediate future.Verde, Ethiopia, Guinea) plan to reduce thenumber of their examinations. Arguments can be advanced in favor of
and against school-based assessmentIn the second strategy, procedures are (Heyneman 1988; Pennycuick 1990). Argu-
already in place or proposals exist to allo- ments in favor go as follows:cate a percentage of public-examination *Since assessment by teachers is a cru-marks on the basis of school-based assess- cial component of good learning andment. The school-based assessment could teaching, every effort should be madebe used in areas in which a terminal exter- to improve teachers' competence in thisnal examination is inappropriate, for ex- area (Crooks 1988). If school-based as-
42
sessment becomes part of the certifica- relationship between teachers and stu-tion process, it is likely that greater ef- dents towards making the judicial as-fort will be invested in improving pect of the teacher's role more promi-teachers' general competence in assess- nent.ment, and this should have beneficial * Marking standards in school-basedeffects on teaching and learning. assessment are likely to vary both* School-based assessment provides within and among schools. Whileimmediate feedback information to moderation procedures can help, theyteachers on student achievement and tend to be expensive.teaching effectiveness. * School-based assessment can subject* Since school-based assessment is car- teachers to considerable parental pres-ried out over time and by a person who sure, especially during the periodsknows students well, it is likely to pro- leading up to and immediately aftervide a more valid and reliable ap- critical public examinations.praisal of a student's achievements * School-based assessment would re-than can a single external terminal quire teachers to devote more time toexamination. In this context, one com- assessment and recording.mentator in Zambia has observed that * School-based assessment gives risethe school is "the only place where to a variety of administrative problemsthere is enough information to do rea- for schools, such as what to do whensonable justice to a pupil" (Kelly 1986, students are absent for tests or whenp. 20). students transfer from one school to* School-based assessment permits an another.extension of the range of curriculum * Teachers' assessments are subject totopics which are examined. The present a variety of biases.system of examinations limits the range * In many instances, it is difficult ifof achievements that can be assessed not impossible to apply school-basedand must narrow the curriculum in assessment to non-school-based candi-schools. Aspects of achievement that dates.cannot be satisfactorily assessed in aterminal examination include a Such considerations led the participantsstudent's ability to plan and organize in a conference on school-based assessmenta project and persevere with it over in Swaziland to conclude that if such as-time. While the assessment of oral and sessment is to be used in combination withpractical skills may be carried out in a external-examination results in the finalterminal examination, inevitably it will national summative assessment of students,be limited, artificial, and expensive. a number of conditions should be fulfilled* School-based assessment reduces the to at least mitigate, if not completely remove,undesirable back-wash effects of exter- the problems to which its use is likely tonal examinations. give rise. First, national item banks of tests* School-based assessment, if spread should be available to teachers. Second,over the year, can increase the level of clear guidelines for assessment should bepupil motivation and application provided. Third, procedures for recordingthroughout the year. and interpreting should be common to all
schools. And fourth, standards betweenSome of the following arguments ad- schools "should be controlled statistically"
vanced against the use of school-based as- and should be moderated and controlled bysessment in the certification of students: school inspection (Swaziland, Ministry of
* Its use can change the nature of the Education 1987).
43
Clearly advantages and disadvantages are quire extensive pre-service and in-service teacher-associated with both school-based assessment training programs as well as other resources.and external examinations. Individual coun-tries should consider the appropriateness of Use of Multiple-Choice Tests and Theirthese procedures for the conditions in which Associated Technologytheir educational systems operate-the pres-sure for selection, the level of training of Partly as a result of the growth in the num-teachers, and the availability of resources to bers of students taking examinations, theresupport either a school-based or external sys- has been an increase in many countries in thetem of assessment. When selection is an im- use of objective multiple-choice tests. In ourportant issue, any consideration of a system study, half the countries used such tests inin which teacher assessments are a basis for their primary-school certificate examinations.important decisions about students must rec- Such tests are cheaper to score when largeognize that the assessments should not only numbers of candidates are involved. They arebe accurate but should be perceived to be so. likely to be more reliable and to provideTo achieve this situation, teachers would re- broader curriculum coverage than essay tests.quire the support of a high level of training They also make the provision of feedback onand a range of services, including guidance student performance easier. Related to theand secretarial support. In many countries use of objective tests has been the growth inoutside Africa in which school-based assess- pretesting of items, in mechanical marking,ment is used, teachers also have available a and in the use of computers to process results.range of aptitude and achievement tests that Despite the apparent advantages of these pro-can be used to set objectives, to monitor stan- cedures, the use of objective tests and the tech-dards, and to provide a basis for student guid- nology usually associated with them, some ofance. The problems that can arise when which can be applied to more traditionalschool-based assessment is introduced into a kinds of examinations, can cause a number ofsystem in which classes are large and the nec- problems.essary resources are not provided have beendocumented in Lesotho (Sebatane 1985) and The first problem arises from the fact thatin Nigeria (Ali and Akubue 1988). the selective role of examinations may be re-
inforced by the introduction of computerThe issue of school-based assessment is a technology and the ease with which statisti-
very live one in many countries. As educa- cal procedures can be used to standardizetional systems develop and expand and as the marks and determine cut-off points for gradespressure for selection eases, we may expect a on the basis of the number of students obtain-decline in the need for external examinations. ing certain scores. This can happen both withCountries will reach that point at different examinations that use multiple-choice testsdates. As they work towards it, investment and ones that do not.in the development of the required infrastruc-ture for an effective system of school-based A second problem is that the new proce-assessment (including teacher training) dures, especially the selection of items thatshould have the short-term effect of improv- have high indices of discrimination for mul-ing the quality of assessment and instruction tiple-choice tests, tend to reinforce the viewin schools. that the latent-trait model of classical psycho-
metric theory adequately represents theConsideration should be given to increasing school achievements of students. The as-
the role of school-based assessment. This will re- sumptions on which this approach is based-
44
that when many discrete behaviors co-vary, Examination authorities should be sensitive tothey have a single underlying determinant, of the fact that while multiple-choice tests offer thewhich individuals possess varying benefits of efficiency, test reliability, and potentialamounts-can be challenged even in the case forfeedback to teachers, they are subject to a num-of western students (Christie and Forrest ber of serious limitations. These include unduly1981). Their validity is more doubtful in the arbitrary use of standardized marking schemescase of students who live in two cultural and cut-off scores attributable to the ready avail-worlds, with all that implies in terms of lan- ability of computers, an assumption of a unitaryguage use, modes of cognitive functioning, achievement trait, psychometric procedures thatand methods of learning (Kellaghan 1961, reduce the content validity of a test and the neglect1968). of important educational skills not amenable to
multiple-choice test assessments. Since decisionsRigid adherence to a policy of discarding about students are likely to be more valid and fair
items with low item-test correlations runs the when based on multiple sources of information, in-risk of eliminating items that prove to be ei- sofar as it is feasible, candidates should be requiredther very easy or very difficult (Greaney to demonstrate their knowledge and skills in a va-1980). Thus, the routine discarding of items riety of contexts.that have low correlations with the total testscore also runs the risk of making the test less Validityrepresentative of the defined universe(Cronbach 1971; Kwasna 1974; Quansah Validity is concerned with the appropri-1985). ateness of descriptions, inferences, classifica-
tions, or decisions that are made on the basisA further problem arises from the fact that of examination results. As far as curriculum
the introduction of the new technical devel- validity is concerned, the responsibility foropments led in some countries to dropping determining that questions in public exami-essay examinations and to a heavy reliance on nations reflect adequately the official curricu-multiple-choice tests. An examination sys- lum rests ultimately with the body entrustedtem made up only of multiple-choice items with the operation of the examination system.has several disadvantages. For one thing, it If an examination is to possess curriculum va-cannot measure some of the outputs of edu- lidity, individual questions should representcation that most people would regard as im- a balanced and adequate sampling of the cur-portant, such as oral and practical skills, in- riculum course. While some systems, such ascluding the ability to write continuous prose. in Swaziland, take the precaution of usingFurther, multiple-choice tests are most effi- moderators to establish the extent to whichcient in measuring recall and recognition of the elements of the prescribed curriculum arefactual information; it is extremely difficult to covered, others, for example Madagascar andconstruct such tests to measure higher-order Togo, simply select questions from lists sub-skills. Limiting an examination to multiple- mitted by teachers.choice items is likely to have undesirable ef-fects on what goes on in schools since, in pre- Part of the reason for the lack of adequateparing students for the examination, the curriculum coverage in examinations is thattemptation will be to pay little attention to student selection is a major function of theskills other than those which can be measured examinations. An examination can be anby the test. If higher-order and writing skills efficient selection instrument even if largedo not form part of public examinations, they sections of a curriculum are excluded fromare not likely to receive the attention they it. Further, since an examination designedshould in schools. for selection will have as a major interest
45
discrimination between students close to the However, discarding items that do notcut-off point for selection, questions in- exhibit intermediate levels of difficulty willcluded may be too difficult for the general also affect a test's instructional validity sincebody of candidates, something that is borne material that is well taught will be excludedout by the poor level of student performance from the test. It is thus important that de-on many public examinations in Africa. cisions to include items in a test should not
be based on the common practice of select-Despite the emphasis on selection in pub- ing only items of intermediate level of dif-
lic examinations, study of the predictive va- ficulty, which will maximize a test's abilitylidity of performance on the examinations to discriminate between student perfor-has received little attention. The limited mances, but, at the same time, will reduceavailable evidence is not very encouraging. its curricular and instructional validities.In one study in Ethiopia, students' perfor-mance on the ESLCE mathematics exami- An alternative to norm-referenced stan-nation did not predict success in first-year dardized tests was proposed in the Mauri-mathematics in the university (Asmerson tius report. It suggested that a pass shouldand others 1989). A somewhat similar study be redefined in terms of essential compe-conducted by the University of Zambia con- tencies for primary-school graduates rathercluded that the low predictive validity of than in terms of standardized scores. Eachexaminations rendered their use for selec- letter grade would be defined in criterion-tion at all levels problematic (Kelly and oth- rather than in norm-referenced terms, anders 1986). grades would indicate the performance lev-
els of students. The approach would alsoAttempts have been made in a number have the advantage of lending itself more
of countries to improve aspects of the valid- than norm-referenced procedures to theity of examinations. One such attempt in- monitoring of changes in achievement stan-volves the pretesting of items used in mul- dards over time. Since a criterion-referencedtiple-choice tests, a device that, it might be approach might not adequately serve the se-argued, should improve curricular validity. lection function of the examination, the Mau-However, pretesting may not have this ef- ritius report goes on to recommend that stu-fect (Madaus, Airasian, and Kellaghan 1980). dents in the top grades, such as A and B, beSelection of items of intermediate levels of ranked for selection purposes. While thedifficulty for the final test, like the use of proposal has much to recommend it, the dif-high discrimination indices, runs the risk of ficulties of specifying essential competenciesmaking a test unrepresentative of the de- should not be underestimated.fined universe. Such a practice is more likelyto contribute to instructional validity (what A body responsible for the construction ofis taught in schools) than to curricular va- examinations has a clear obligation to attend tolidity (what is laid down in the curriculum), validity issues. In this, the need to achieve bal-since only items that a given proportion of ance between the certification and selection func-pupils (about half) get right will be retained tions of the examinations must be kept in mind.in the test. Thus, only items pupils have The examinations body should ensure that thehad the opportunity to learn (presumably procedures it uses in the construction of exami-in school) will find their way into the test. nations are more likely to contribute to, than toIf schools consistently ignore aspects of the interfere with, the examination's validity. If ex-curriculum, items relating to those aspects aminations are used for selection, the examina-will not be included in the test, creating a tion body should accumulate evidence to showsituation in which certain curriculum areas how well performance on examinations reflectsare not tested (Kellaghan 1990). real-life educational performance.
46
Accountabi]lity schools or districts on the basis of rankingsmay be inaccurate and unfair. First, it has
Examination results are sometimes used been shown in the United States that rank-as a measure of school effectiveness. When ings can vary greatly depending on thethis occurs, teachers and others tend to be school outcome that is used (Guskey andheld accountable for the performance levels Kifer 1990). Second, comparisons betweenof students. Frequently, successful perfor- schools on the basis of examination resultsmance is equated with the percentage of can- generally fail to take into consideration thedidates who advance to the next highest social or even physical conditions underlevel of the educational system. which schools operate. However, if statisti-
cal adjustments to school outcomes are madeIn Kenya, national and provincial on the basis of student characteristics and
rankings of schools based on public exami- their backgrounds, level of teacher qualifi-nations are published annually to provide cation, and availability of material resources,the public with an opportunity of compar- a very different order of merit might emerge.ing schools. In districts that occupy the Third, whether or not adjustments are madelower end of the merit order, teachers may to school-output measures, errors in the mea-be accused of being lax, administrators of surements on which rankings are based arenot performing their supervisory duties, and not likely to be taken into account whenparents of not being "development-minded" judgments of merit are being made about(Oloo 1990). Zambian teachers also tend to schools. For example, will observers knowbe blamed if too few of their pupils are se- that the difference between a rank, say, oflected on the basis of examination results 35 and one of 38 may only be a matter offor further schooling. "Good" teachers in chance? Finally, in cases in that parents orthe Prai region of Cape Verde are those who students have a choice of school, the publi-attain a promotion rate of 50 percent at the cation of results may lead to schools thatend of the school year. Examnination perfor- are perceived to be doing well to attract stu-mance is used by the Ministry of Education dents of high levels of scholastic ability,in Swaziland and by the general public as a aspiration, motivation, and parental supportmeasure of school effectiveness. In Lesotho while those that are perceived to be doingalso, examinations serve an accountability badly, even though they may be more "ef-function; teachers and schools with high fective" than schools with better results, willexamination marks are usually regarded as be avoided by such students. This situation"good." In Uganda, where most of the in time could lead to low morale in indi-school fees are paid by parents, examina- vidual schools and to a ghettoization of thetion results can have a fairly rapid impact; school system."good" results tend to increase the pressurefor school places. The effects on schools of using examination
results as an accountability mechanism demandsThe ranking of school districts or schools immediate attention from education authorities.
is more complicated than current practice In particular attention should be directed to thewould seem to acknowledge. In fact, infer- identification of "successful" and "failing"ences made about the "effectiveness" of schools and teachers on the basis of examination
results and to the publication of school rankings.
47
Language of Examination candidates' poor examination performancesis due to a failure to understand questions set
In thirteen of the fourteen countries in in English. The low pass rate (38.8 percent)our study for which data were available, recorded by Madagascar students in thethere was an average of 4.7 principal lan- Technical Baccalaureat has been attributed toguages (i.e., indigenous languages spoken the extensive use of French in the educationalby at least 10 percent of a country's popula- system. In Ethiopia, where English is the me-tion and any other language that serves as dium of instruction in secondary schools, itan official language, lingua franca, or me- is accepted that there will be a move towardsdium of instruction) (World Bank 1988, Table assessment in Amharic in the Grade 8 andB-12). ESLCE public examinations.
This situation creates serious problems re- The choice of language for examinations haslating to language of instruction and language to be considered in the wider context of choice ofof examinations. Countries have adopted a language of instruction. The decision is a diffi-variety of strategies to deal with the situation cult one for many countries and requires contin-and, in some cases, have changed these strat- ued research, analysis, and review.egies over time. The most common practiceis to use a local language for instruction in Comparability of Performancethe early grades of primary school. In somecountries, the relevant international language Possession of a certificate, such as a pri-(French or English) is also used at this stage. mary school-leaving certificate, would implyThe medium of instruction in the higher that students had attained an acceptable andgrades of primary schooling in most countries comparable level of performance in definedis officially either French or English. In all achievement areas. Further, one might as-countries, French or English is the medium sume that the certificated students had ad-of instruction in secondary schools, except in dressed themselves in examinations to theCape Verde, where it is Portuguese. In some same tasks under similar conditions. How-countries, some schools also provide instruc- ever, current practices in some countries re-tion in an alternative language, for example, lated to "compensation" raise a number ofArabic in Mauritania. In all cases, with the issues that suggest that these assumptionsobvious exception of language subjects, the may not be tenable. Basically, the practice oflanguage of examinations is the same as the compensation involves allowing credits orofficial language of instruction. marks that a student has achieved in one cur-
ricular area to make up for a relatively poorFor the majority of pupils in the countries performance in another area. Questions ob-
in our study, the language used in instruc- viously arise about the comparability of stu-tion and in examinations (English or French) dents' achievements when different achieve-was not their mother tongue. We know rela- ments are involved or when the achievementstively little about the difficulties that African are measured in different ways.students experience in the acquisition of a sec-ond language or about how education InGuineaadjustmentsaremadetoexami-through the medium of a second language nation marks to take into account the fact thataffects the quality of that education. Not sur- some teachers may not have covered certainprisingly, some commentators have attrib- elements of the curriculum, again raising theuted low levels of achievement in public ex- issue of comparability of results. Cape Verdeaminations to the language difficulties of stu- students are exempted from an oral exami-dents (Eisemon 1990). In Mauritius, for ex- nation if they achieve a certain level of per-ample, where the lingua franca is Creole, some formance in the written test. In fact students
48
may gain exemption from each of the three Examination bodies that use compensationnational examinations on the strength of their procedures should ensure that these do not giveperformances on internal school-based ex- rise to ambiguities in the interpretation of exami-aminations. A Togo candidate who scores less nation performance.than 10 but at least 9 out of 20 can compen-sate for this relatively low mark by taking an Quotasoral examination. In the Ethiopian Grade 6and Grade 8 examinations, it is possible to To meet social or political objectives, quo-compensate for poor performance levels in tas are used in some countries, along withcertain subjects. For example, candidates examination results, to control student flowwho score poorly in Amharic, English, and to the next highest level. In effect, quota sys-mathematics, three important subjects at sub- tems guarantee that a certain number of can-sequent educational levels, can qualify for a didates from a particular category, who oth-certificate by moderate to good performance erwise would not advance to the next level,in political education, general science, and so- do so. Where this happens, the success ofcial science. these candidates occurs at the expense of
other candidates who would have advancedIn Lesotho a first-class pass is awarded to had no quota system been in operation.
primary certificate candidates who obtain agrade of at least 60 percent in English, Followingthe correction of theconcours atSesotho, and imathematics and at least 60 per- the end of primary schooling, the Rwandancent in the examination as a whole. To ob- Ministry of Education selects students for sec-tain a second-class pass, a grade of at least 50 ondary schools, bearing in mind the need topercent has to be obtained in these three sub- have regional, ethnic, and gender balance.jects and at least 50 percent in the overall ex- Mauritius also uses a gender quota; the topamination. A third-class pass is awarded to 2,000 boys and 2,000 girls on the primarycandidates receiving 40 percent in any three school-leaving examination are offered placessubjects and also on aggregate. A student in secondary school. Both Mauritania andwho obtains 90 percent in all subjects with the Ethiopia set pass marks that vary from regionexception of, say, English, in which he or she to region. In Ethiopia, the selected passinggets 40 percent, would merit only a third-class raw score is transformed into a percentilepass. It coulcd be argued that the overall per- rank derived from the distribution of aggre-formance of this candidate is superior to that gate scores for all candidates. Regional offic-of another who obtained a first-class pass by ers are instructed to determine the passingscoring 60 percent in all subjects. raw scores for their regions by identifying the
aggregate score that corresponds to the na-In these examples, it is often difficult to tionally selected passing percentile rank. The
determine what possession of a formal cer- effect of this on the Grade 6 National Exami-tificate means in terms of levels of scholastic nation in 1987 was that while the passing rawachievementsince students' achievements are score in Eritrea was 42, in Gojam it was asassessed in different ways. However, the use high as 47. Thus, for many Ethiopian stu-of different modes of assessment does not dents, success or failure depended on placenecessarily give rise to this problem. While of residence.it may have this effect, multiple measures ofachievement can be useful in allowing a per- One effect of the operation of quota sys-son to exhibit a particular skill, trait, or tems is that students change their place ofachievement. residence or their names to improve their
chances of selection.
49
The quota system enables examination been selected on their examination perfor-authorities to examine the predictive validity mance but gained entry to secondary schoolof selection procedures. To do this, the sec- because of quota requirements.ondary-school performance of students whowere selected on the basis of their examina- The social, political, and educational implica-tion performance is compared with the per- tions of quotas merit serious consideration andformance of students who would not have monitoring.
50
6 Using Examinations to Improve Education
In conside]ring the role of examinations ticular the inability of educational systemsin educational reform, one should not lose to accommodate all students who wish tosight of the range of problems facing educa- remain at school. In this situation, exami-tion in Africa. The extent of these problems nations are used as the means of allocatingvaries from country to country, but all coun- places to students. Whatever dedsions aretries to a greater or lesser extent, face prob- taken about assessment and examinationslems relating to accommodation, the rel- procedures, it has to be recognized that theevance of curricula, availability of books and need will remain, for some time at any rate,materials, and teacher training. Further- to operate a system of selection that is per-more, these problems in the educational sys- ceived to be fair and equitable. And if theretem co-exist with, and often are caused by, has to be selection, it has been argued, for aeconomic and social problems in the wider long time in European countries (Curricu-society. Problems such as these are not go- lum and Examinations Board - CEB 1985)ing to be solved by improvements in exami- and currently in developing countriesnation systems alone. However, examina- (Heyneman 1985), that examinations are ations could play a role in the reform of edu- more equitable way of doing the job thancational systems and in improving the qual- other procedures that have been tried. Evenity of education in schools. At the very least, if examinations continue to be used for se-good assessment procedures must be better lection, this does not mean, of course, thatthan bad ones. Furthermore, despite the efforts should not be made to improve themproblems that will arise in implementing as instruments of certification also.some reforms, such as in measuring andteaching higher-order skills and in provid- In this chapter, for the fourteen coun-ing feedback information to schools, at- tries in our study, we present the proposalstempts to do these things are preferable to which were set out in action plans forcontinuing with examination systems that, changes in examination systems which, itin many countries are regarded as obviously was believed, would lead to an improve-defective in many ways. Besides, there are ment in the quality of examinations and ofsome recommendations that can be made, education. Following that, we provide somewhich can hardly be regarded as controver- guidelines which, if followed in the imple-sial, such as the need to include in examina- mentation of action plans, should result intions assessments of students' ability to write the improvement of examinations. Afterif writing is to receive the attention it merits that, we consider problems and undesirablein schools. effects that can arise from certain kinds of
examination reform. Finally, we outlineWe have seen that examinations occupy conditions that should be present in the
a key role in the educational systems of the administration of examination authorities ifcountries considered in this report. This is these authorities are going to be in a posi-partly because of a resource problem, in par- tion to support examination reform.
51
Action Plans tions which they offer. Ethiopia, for ex-ample, has indicated its intention to elimi-
Each of the fourteen countries prepared nate the Grade 6 National Examination whenan action plan designed to improve the qual- the shortage of junior-secondary schoolity and functioning of its examination sys- places is overcome.tem. A tabular summary of the plans isprovided in Table 5.1, separately for Anglo- Centralize Examination Authority.phone, Francophone, and other countries. Three countries (Cape Verde, Ethiopia, Mau-The objectives of the action plans can be cat- ritania) have indicated a desire to centralizeegorized as relating to Administration, Tech- responsibility for public examinations withinnical and Training Requirements, Equip- one agency. In this way it is hoped to makement, and Construction/Rehabilitation. It the system more efficient through a sharingshould be pointed out that the objectives of scarce resources. In Ethiopia one centralwere not stated with similar clarity in all system is planned as a replacement for thethe reports. Ethiopia School-Leaving Certificate Exami-
nation Office run by the University of AddisAdministration Ababa and the National Examination Board
of the Ministry of Education which has re-Expand Examination System. The prob- sponsibility for the Grade 6 and Grade 8
lems caused by the substantial increase in national examinations.enrollment figures in general, and in thenumber of candidates taking public exami- Centralize Operation of Examinations.nations in particular, are reflected in the Assigning printing and packaging of exami-plans to increase the administrative capa- nation papers to an external agency increasesbilities of public-examination authorities. the likelihood of a breach of test security.Five of the six Anglophone countries and Three large countries (Kenya, Uganda, Zam-Ethiopia indicated a need to provide for the bia) propose to assign responsibility for allexpansion of their examination systems to aspects of public examinations to onecope with anticipated increases in numbers. agency. At the moment, to cite but one ex-
ample, the Kenya National ExaminationLocalize Examinations. Three countries Council could achieve substantial savings if
(Lesotho, Mauritius, Swaziland) plan to lo- it were in a position to transfer the printingcalize public examinations by developing al- of examination papers to an in-house facil-ternative examinations to those prepared by ity.overseas agencies. The challenge for thesecountries is to develop within their exami- Create a Technical Bureau/Support Unit.nation systems the ability to design, print, The need for the creation of a technicalmark, and validate examinations while bureau or support unit to develop tests, tomaintaining public confidence by ensuring provide training, and to evaluate and carrythat the new examinations retain the status out research on aspects of the examinationof external examinations such as those pres- system has been identified in five of theently offered by the Cambridge Overseas Francophone countries (Chad, Guinea,School Certificate Examination Syndicate. Madagascar, Mauritania, Togo.) In Mada-
gascar, for example, it is hoped that this newPhase Out Examinations. In 1987, the unit might take responsibility for improv-
Kenya National Examination Council dis- ing school-based assessment techniques, de-continued the Kenyan Certificate of Educa- veloping standardized tests, and carryingtion. Guinea, Cape Verde, and Ethiopia plan out specialized studies designed to informto reduce the number of public examina- those responsible for decision making.
52
Table 5.1OBJECTIVES IN DRAFT PLANS
Administration Kenya Lesotho Mauritius Swaziland Uganda Zambia Chad Guinea
1. Expand system to dealwith increasing numb
x x x x x x
2. Localize all exams X X X
3. Phase out exam(s)
4. Centralize responsibilityfor exams within one agency
5. Allow agency to assumeresponsibility for all aspectsof exams (e.g. printing) X
6. Create technical bureau/support unit to develop tests,provide training/evaluation X X X
7. Improve efficiency of examsby strengtheningmanagement structures X X X X X
Technical and Training
1. Improve quality of exams(improve exam-syllabus linkage,introduce criterion-referencedtests) X X X X X X X X
2. Improve assessment, researchand development capacityof systems by improving technicaltraining/resources (e.g. library) X X X X X X X X
3. Involve teachers more inassessment as a replacement for, orin conjunction with, public exams X X X X
4. Develop capacity to providefeedback information on public-exam performance to schools X X X X
5. Provide training for MOEofficials (school inspectors)/teachers/counsellors inassessment aLnd exams X X X X X X
Equipment
1. Improve efficiency of examsthrough provision of equipment X X X X X X X X
Construction/Rehabilitation
1. Extend/improve accommodationfor administration of exams X X X X
53
Table 5.1 (continued)
OBJECTIVES IN DRAFT PLANS
Administration Madagascar Mauritania Rwanda t Togo CapeVerde Ethiopia
1. Expand system to dealwith increasing numbers x
2. Localize all exams
3. Phase out exam(s) X X
4. Centralize responsibilityfor exams within one agency x x
5. Allow agency to assumeresponsibility for all aspectsof exams (such as . printing)
6. Create technical bureau/supportunit to develop tests, providetraining/evaluation X X x
7. Improve efficiency of exams bystrengthening management structure x
Technical and Training
1. Improve quality of exams(improve exam-syllabus linkage,introduce criterion-referencedtests)b X X X
2. Improve assessment, researchand development capacity ofsystems by improving technicaltraining/resources (e.g. library) X X X X x
3. Involve teachers more inassessment as a replacement for,or in conjunction with, public exams X X x
4. Develop capacity to providefeedback information on public-exam performance to schools x
5. Provide training for MOEofficials (school inspectors)/teachers/counsellors inassessment and exams X X x x x
Equipment
1. Improve efficiency of examsthrough provision of equipment X X X X X
Construction/Rehabilitation
1. Extend/improve accommodationfor administration of exams x
a/ Being finalizedb/ Recommendations to produce tests of "creativity," "intelligence," "visualization," are not included here.
54
Improve Management Efficiency. A to- In most instances, it is envisaged that for-tal of ten of the fourteen countries specified eign training will be required to enable per-the need to improve the efficiency of public sonnel to acquire the appropriate researchexaminations by strengthening management skills. In addition to training, research, andsiructures. In many instances, the lack of development, personnel will need to haveautonomy of the examination body was cited access to libraries equipped with appropri-as a factor which restricted the efficiency of ate technical books, journals, and sample as-the examination system. There is also a need sessment instruments.for a substantial recruitment and trainingprogram to equip management personnel Involve Teachers in Assessment.with relevant skills and knowledge. Salary Slightly over half the countries identified thescales should be sufficiently attractive to dis- need to increase teachers' involvement in thecourage skilled personnel, such as in the area assessment process. In some instances it isof computing, from seeking employment envisaged that assessment by teachers wouldelsewhere. eventually replace some external examina-
tions. Most systems, however, favor the in-Technical and Training Requirements volvement of teachers in providing part of
the overall assessment in a particular sub-Improve Examination Quality. Consid- ject examination. For Mauritius for example,
erable agreement is to be found in the ac- it is suggested that, following the trainingtion plans regarding the need to improve of teachers in assessment techniques, thethe quality of public examinations. Each of Certificate of Primary Education wouldthe fourteen countries specified that it show the teacher's grade in addition to theneeded technical assistance in this area. external examination grade for each subject.Mlany recognized the need to improve thevalidity of their public examinations, in Develop Feedback Capacity. Four An-particular to improve the link between the glophone countries (Kenya, Lesotho, Swazi-content of examination questions and the land, Uganda) and Ethiopia stressed thesyllabus. Psychometric support was also need to develop the capacity of the public-called for to help, among other tasks, to examination system to provide feedback in-develop item banks and construct adequate formation to schools on aspects of students'assessment instruments. A number of ex- performance on examinations. It is believedamination bodies recognized the need to in- that such information could contribute to thetroduce new modes of examining, involv- quality of teaching and learning in theing oral, aural, and practical tasks. schools by identifying aspects of the curricu-
lum that appear to be misunderstood or sim-Improve Research and Development ply ignored by examination candidates.
Capacity. The need to introduce or improve Feedback information can also be useful tothe capacity of the examination system to educational authorities by identifying re-conduct research directly related to public gional disparities in achievement and aspectsexaminations was recognized in each of the of the curriculum where candidates farefourteen countries. In Swaziland for ex- poorly, possibly due to lack of appropriateample, it was proposed that this research equipment, for example in physics andwould include validity studies, monitoring chemistry.the implementation of a proposed continu-ous-assessment system, and the effects of Provide Training for Inspectors andchanges in the system on the internal and Teachers. Virtually all countries recognizeexternal efficiency of the educational system the need for the training of Ministry of Edu-(dropout, repetition, levels of achievement.) cation officials, school inspectors, and teach-
55
ers in assessment terminology and proce- sotho and Madagascar have indicated thatdures. In-service training seems particularly their examination offices are badly in neednecessary to familiarize teachers with the of repair and improvement.techniques required to conduct school-basedassessment. Zambia's proposal envisages Cross-National Cooperationthat as many as 10,000 secondary-schoolteachers would attend a ten-day in-service A number of countries referred to thetraining course over a four-year period. Cape benefits of cross-national cooperation in ex-Verde authorities propose that a foreign ex- aminations but did not actually include pro-pert should train a cadre of people who in posals for such cooperation in their actionturn would become trainers of others. plan. Two possible approaches to cross-na-
tional cooperation are outlined here.Equpmenet
First, associations of examining bodiesAll action plans indicate a need for equip- might be formed to enable those involved in
mernt. Specified needs relate mostly to the examinations to exchange information andproduction of examination papers, such as experiences and to provide short-term train-printing equipment, and the processing of ing courses. Such an association (The Asso-examirnations. Data-processing equipment is ciation of Heads of Institutions Responsiblerequired for scoring multiple-choice exami- for Examinations and Educational Assess-nations, the preparation of results, and for ment in East and Southern Africa) alreadystoring of assessment information. Main- exists in the Anglophone areas of eastern andframe and micro-computers have been re- southern Africa. This association, however,quested. Books and journals on testing and is in need of financial support. There is alsomeasurement have also been listed. Other a need to cater for countries in non-Anglo-equipment that is frequently requested in- phone areas.cludes ty,rpewriters, photocopiers, and paper.A niumber of countries have asked for fur- The second proposal for improving cross-niture, (Madagascar and Togo), and motor national cooperation relates to the establish-vehicles, (Lesotho and Mauritania). ment in a university of research facilities and
post-graduate courses relating to measure-Const i,xtio iRehabilitation ment, assessment, research, and evaluation.
It would probably be necessary to make thisFour Anglophone countries, three Fran- provision in both Anglophone and Franco-
cophone countries, Cape Verde, and Ethio- phone areas.pia include specific requests for constructionand/or rehabilitation of buildings. Kenya Before this step is taken, there will be aand Uganda require the construction of large need, as indicated in individual country ac-buildings to house printing equipment. The tion plans, to have individuals go to univer-Ugandan proposal outlines the need for a sities in Europe or North America for post-building to house, in addition to printing graduate studies. Consideration should beequipment, a computer facility as well as given to having individuals from severalrooms for desktop publishing. Other coun- countries attend the same institution as thistries, for example Togo and Chad, require could help establish relationships whichthe construction of buildings to accommo- would form a useful basis for future co-op-date all examination board personnel. Le- eration when the individuals return to their
own countries.
56
Improving Assessment in Public important and will receive special attention.Examinations Hence, the areas selected for inclusion in
examinations should cover importantA number of guidelines should be fol- curricular topics.
lDwed to improve assessment in the public-examination system. Before presenting them Second, insofar as it is possible, thetwo general points about examination reform modes of assessment (written, practical, andthat have already been made might be re- oral) which are used should be diverse andpeated. First, such reform has to be consid- should reflect the goals of curricula. Whatered in the general context of the educational is tested is what is likely to be taught andsystem, not as an isolated element in it. what is not tested is likely to be ignored.Changes in the assessment system are likely Hence, if examinations do not require stu-to depend on, and to affect other aspects of dents to write continuous prose, the skillsthe educational system. The dependence of involved in this activity will probably notexamination reform on other aspects of an receive much attention in schools. Similarly,educational system is perhaps most obvious examinations that do not include oral orwhen one considers the resources that will practical components are likely to lead to abe required to implement reform. On the neglect of skills in these areas. However, costother hand, the effects of reform are likely to is a serious problem if external examinersexhibit themselves in the quality of students are used, particularly in the case of oral andthat are selected for further education, as well practical skills. For this reason, the possibil-as in the character of curricula offered in ity of using assessments by teachers withinschools. Second, in considering examination the school is worth exploring. If done prop-reform, one has to be conscious of the fact erly, such assessments are likely to be lessthat there is no perfect system of assessment expensive than external ones and are alsoand that there is no point in looking for one. likely to be more valid. However, some sys-Any system will have advantages and dis- tem of moderation may be required ifadvantages. This will be particularly obvi- teacher assessments are incorporated into theous when an assessment system is used to public-examination system. In this case, itserve more than one purpose. It is the task should be borne in mind that moderationof individual countries, in the light of their can be cumbersome and expensive and, intraditions, resources, values, and needs to some circumstances, quite impractical.decide what balance of options best suitsthem. Since it is not usually possible to an- Third, assessment instruments shouldticipate precisely the advantages and disad- pay adequate attention to the measurementvantages of any system in advance, it is im- of higher-order thinking skills. Many tests,portant that the effects of examination re- and these are the easier ones to construct,form, particularly on curricula and teaching require little more than the recall or recog-practice, should be monitored. With these nition of information. However, if it ispoints in mind, we may now turn to more agreed that "there is a need to make deepspecific guidelines for the improvement of learning a central goal of education, and toassessment in the context of public examina- foster development of this goal through thetions. evaluation of students," then it will be nec
essary to "...place emphasis on understand-First, assessment should be based on the ing, transfer of learning to untaught prob-
official curriculum. However, since not lems or situations, and other thin kingskills,everything included in a curriculum can be evaluating the development of these skillsassessed, the areas which are selected will through tasks that clearly must involve morebe regarded by teachers and students as very than recognition or reca" (Crooks 1988). .7t
57
is assumed in this view that the inclusion apply learning to new material or situationsof tasks requiring higher-order skills in tests (Gronlund 1973; Terwilliger 1989).will lead to greater emphasis on teachingthese skills in classrooms. There is in fact Even with a simplified taxonomy, it issome evidence from a study carried out by impossible to be sure that the taxonomicOgundare (1988) in Nigeria, in which pri- level of an item will evoke the targeted tax-mary-school students were taught social onomic level in the candidate answering thestudies using a problem-approach method. item. Whether or not it does will dependThis suggests that when this emphasis is on the learning experiences of the candidatepresent, not only do students acquire more (Torrance 1986). Bloom (1956) was awarefacts, they also comprehend material better of this when he observed that if a problemand are better able to apply their knowl- designed to measure application is too sim-edge to the solution of new problems and ilar to problems used in class, it will mea-to evaluation activities. sure comprehension rather than application.
Various taxonomies have been designed A fourth guideline to be followed in theto help test constructors elicit higher-order construction of examinations is that exami-thinking skills in students. Some of these nations should test students' ability to ap-contain as many as six levels (such as Bloom ply their knowledge and skills not just in1956, Perrott 1982). However the construc- scholastic contexts but in situations outsidetion of a test which distinguishes clearly be- the school as well. Thus the examinationstween that many levels might not be pos- should take into account the everyday lifesible. Problems arise particularly at higher conditions of students and the kinds of prob-levels designed to measure synthesis and lem they are going to meet at home, on theevaluation. For example, in Britain, the farm, in the shop, or in the market place.specification of achievement levels withinsubject areas where higher-order or general The Kenya Primary School Certificate Ex-skills were involved has proved very diffi- amination provides examples of items whichcult in the context of developing a system relate to situations outside the school. Theof assessment for the new national curricu- following are examples of such items:lum (Nuttall 1990). Because of this, it isperhaps better in test construction to think Diarrhoea is a disease that kills manyin terms of a simpler taxonomy containing babies in Kenya. When babies havetwo levels rather than six. The first level diarrhoea they lose a lot of water andcould cover what might be called "minimal" foods. One correct way to treat ba-objectives, which most students might be bies with diarrhoea is to:expected to achieve, and would relate tothe acquisition of knowledge and basic skills A. keep them wrapped up and warm(recall and comprehension in Bloom's tax- so that they sweat out the sickonomy). The second level would involve nesshigher-order outcomes, defined in terms of B.give them drinks of boiled coldmore complex cognitive processes. While water containing some sugar andtests of minimal objectives would examine a little saltstudents' ability to deal with familiar con- C .give them solid foods containcepts and rehearsed skills, the higher-order ing plenty of carbohydratesones would examine students' ability to water until the diarrhoea stops.
58
Wangari lives in a village with many by guidance on the design and implementa-people. From which one of the fol- tion of possible remedial actions.lowing sources can Wangari collect The most serious efforts to provide feed-the best drinking water? back information to schools have been made
by the Kenya National Examinations Coun-A. a big river nearby cil. While the Newsletter of the Council isB. the rain from her mabati roof commended by Lewin (1984) as providingC. the dam near the village "iexcellent examples" of the kind of feedbackD. a swamp on her farm. that is likely to be helpful to teachers, it does
not in fact provide teachers with the kind ofA fifth guideline recognizes that exami- information recommended by Lewin. Since
nation scores used for selection should dif- the information in the Newsletter is limitedferentiate between candidates on the basis to national statistics about performance,of characteristics relevant to the opportuni- there is no way a teacher can ascertain howties being offered. Thus, there is a need to his or her class performed on particularcarry out studies of the predictive validity aspects of the examination. Thus, the basisof selection instruments. for remedial action is limited. While the
provision of information on examinationThe fact that examination scores are im- performance on a school-by-school basis
perfect measures of students' achievements would be necessary to meet Lewin's condi-is the basis of a sixth guideline. It is desir- tions, the logistical problems to which thisable that the past performance and relevant would give rise are formidable.experience of students be taken into accountin making important decisions about stu- The provision of detailed information ondents. This guideline points to the need to the performance of students in the hope thatdevelop the capacity of teachers and schools teachers will direct their efforts towards theto carry out school-based assessment that topics and skills covered in examinations andwill be acceptable outside the classroom. thereby raise students' test performance
raises a broader issue which merits attention.A seventh guideline takes up the recom- It is not surprising that research evidence in-
mendation of several commentators that de- dicates that the test performance of studentstailed feedback information on the perfor- will improve if they have been taught themance of students in examinations should specific material covered by a test and if theybe provided to schools and teachers. For ex- have been frequently exposed to the test for-ample, Lewin (1984) has pointed out that mat (Cooley and Leinhardt 1980; Le Mahieutelling a teacher that 20 percent of a class 1984). However, are we justified in conclud-has failed does not communicate the causes ing on the basis of improved student test per-of either the teacher's or the students' fail- formance that students' actual achievementure. However, if teachers were told which (in terms, for example, of problem-solvingquestions and tasks students had completed skills), as distinct from their test performance,satisfactorily and which ones they had not, has also improved? The answer would seemthe information could be used as a basis for to be no, at least until we have satisfied our-remedial action. The information would be selves that some of the variance in test scoreseven more useful if accompanied by an in- is not attributable to the form of the test usedterpretation of student difficulties as well as (method of measurement) rather than to the
individual characteristics (achievement) that
59
the test is designed to measure (see Bolger struct validity (LeMahieu 1984). If students'and Kellaghan 1990; Campbell and Fiske examination scores are used to assess the ef-1959). Students are taught in such a way fect of providing information to schoolsthat the match between instructional pro- about students' test performance, the sort ofcesses and test items is very close, then in- learning that is revealed in the examinationferences about students' actual skills and score may be made up largely of skills thatknowledge become very difficult to make. are more properly considered test-takingThe problem is particularly acute if we want skills than achievement as indicated by stu-to test whether a student can apply skills and dents' ability to use higher-order skills. In-knowledge to solve new problems, since in deed, as noted above, the more effective thethis case, the problems must be new to the treatment is in terms of having teachers teachstudent and not ones that were taught in to the examination, the greater will be theclass. Practice on a problem may improve match between instructional processes andstudents' test scores; however, not only will test items, and, as a consequence, the moreit eliminate the possibility of drawing un- limited will be our ability to draw inferencesambiguous conclusions about students' abil- about students' actual skills and knowledgeity to solve problems, it may actually have as distinct from test-performance skills. Thisthe effect of interfering with the development guideline then specifies the need to use al-of understanding by students (Linn 1983a, ternative methods of measuring achievementb). when assessing the effects of an examination
reform which involves providing detailed in-This situation creates a dilemma for formation on students' performance to
which there is no obvious solution. To what schools.extent should one improve the match be-tween testing and instruction, not only to The final guideline, concerns the need topromote appropriate curriculum coverage develop effective assessment strategies, dis-but also to ensure that what is tested is what tinct from public examinations, to assess theis taught, if, at the same time, the ability to performance of the educational system andmeasure application and problem-solving of individual institutions. One type of strat-skills is going to be compromised? On the egy will probably not meet the requirementsother hand, should one include new prob- of these two purposes. A procedure involv-lems and types of problems on examinations ing testing in a sample of schools may bewith the objective of assessing higher-order adequate for the assessment of the perfor-skills, knowing that this will create the risk mance of the educational system, while aof failing to measure what is taught? (Linn more school-specific approach, which might,1983b). for example, be incorporated into the func-
tions of the school inspectorate, will be re-An eighth guideline relates to the assess- quired in the assessment of individual
ment of the impact of examination reform schools.on student achievement. On the basis of theprevious discussion, which implies a distinc- Anticipating Problems andtion between test performance and achieve- Undesirable Effectsment, it should be clear that the use of testinstruments, which are part of an instruc- In considering proposals to improve as-tional treatment, to monitor the effects of that sessment procedures and, in particular, totreatment, as in Somerset's (1987) Kenyan design examinations which will have a ben-study, cannot be regarded as satisfactory. eficial effect on curricula and studentThe problem in this approach concerns con- achievement, it is important not to
60
underestimate the problems involved in the lastic competition, they may be more accept-task (see Airasian 1988). It is also important able now than they were in the past. Fur-to guard against possible unintended conse- ther, the idea of relevance in curricula canquences that may arise. be defended without reference to the future
lives of children. Tasks that are related to aThe first problem relates to the task of child's previous experience and are seen to
specifying desired outcomes of instruction have relevance to everyday life are moreon which curricula and examinations might likely to motivate children to learn than arebe based. The task involves important value more abstract tasks.judgments; hence it may not be easy to reachagreement on what the outcomes of school- A second problem in the design of ex-ing should be. Besides, such outcomes aminations relates to the complexities in-vary, even within a single country, accord- volved in teaching stated objectives. Theing to the different circumstances in which mere positing of goals, or even their incor-schools operate and in which their students poration into examinations, does not guar-live. antee their achievement. Indeed, as we have
seen above, it may inhibit them. The prob-Of interest in this context are proposals lem is likely to be most acute if the exami-
to make education more relevant to the nations, as test designers are increasinglyneeds of rural children (the majority of chil- striving to do, place emphasis on measur-dren in Africa). Such proposals have been ing higher-order behaviors rather than, ormade repeatedly in the history of African in addition to, general minimum lower-leveleducation (see, for example, Gifford and behaviors. The earlier discussion alluded toWeiskel 1971; Jones 1922, 1924). On the the problem of measuring higher-order be-whole, they failed to gain acceptance, prob- haviors. Evidence is also available to indi-ably largely because of parents' antipathy cate that teaching such behaviors is differ-to them. Since parents saw traditional aca- ent in many ways from teaching lower-leveldemic education as the route to occupations rote behaviors. For example, higher-levelwith the greatest prestige and reward, a behaviors take longer to teach, developfailure to provide children with such educa- gradually over time, are less amenable totion was seen as limiting their occupational direct instructional approaches, are often dif-opportunities (see Blaug 1973; Bray, Clarke, ficult to locate in a curriculum, and may beand Stephens 1986; Foster 1965; Kellaghan too diffuse to drive instruction (Airasian1991). It remains to be seen whether the 1988).current emphasis being placed on local rel-evance will be more acceptable than past Faced with such difficulties, teachers mayefforts (see Lockheed and Verspoor 1990). resort to teaching students test-taking strat-It may be, as the number of children attend- egies which may undermine the objectivesing school increases, that the prospects of of the examination (Madaus 1988; Madausthe more desirable occupations to which and Kellaghan, forthcoming). Over time,parents aspired for their children in the past teachers become familiar with the formal re-will be seen to be less realistic. Much too quirements of the examination and preparemay depend on the degree of localization students to deal with these rather than withthat is introduced into curricula. If curricula curricular objectives. Strong evidence of thisfor children in rural areas are limited to has been observed in a number of Africanteaching basic principles relating to health countries, and the evidence is not confinedand agriculture and do not deviate from the to Africa. Selection-type questions, whentraditional curriculum to the extent that rural used in examinations, become a particularchildren would be at a disadvantage in scho- focus of attention. Not only are students
61
taught how to cope with such questions, but Conditions to Improve the Administrationnormal teaching may be presented in the of Examinationsform of statements accompanied by a rangeof options from which students are required In this section, some conditions relatingto choose, (for example, The capital of France to the administration of examinations andis: Brussels, Paris, Lyon, Toulouse). This examination authorities are outlined. Theseform of teaching may be used, not just in are necessary if the efforts to develop publicthe examination class, but can filter right examinations which will have a positivedown the school, as has been observed in effect on school curricula and studentreception classes in African schools. achievement are to be effective.
A further possible undesirable effect of Examination Authoritypublic examinations, particularly if the re-sults of schools are publicized, is that schools In the first place, the body entrusted withmay require pupils to repeat grades, either the construction and administration of theto take more time to prepare for the exami- examinations must be adequately financednation or to avoid taking it at all (Madaus and staffed. The body should have sufficientand Greaney 1985). In the case of Kenya, resources and competencies to develop ex-Oloo (1990) states, though he does not pro- amination instruments, to analyze examina-vide supporting data, that "it is fairly obvi- tion performance, and to carry out researchous that the schools which perform best in on examinations. Resources should indudeKCPE are those which enter a few selected access to foreign exchange to allow for pur-candidates." Further, pupils, as a result of chase, maintenance and repair of equipment.academic pressures, may leave school at an At present, most examination bodies in Af-early stage. Grade repetition and drop-out rica have insufficient resources and are hardare, in the United States at any rate, inter- pressed just to administer the system, leav-related (Catterall 1989) and are probably the ing no time for reflection or research.result of a variety of factors, which includefamily and student characteristics, low per- Salary levels should be sufficiently attrac-ceived benefits from education, and school tive to ensure that skilled personnel are notineffectiveness (Lockheed and Verspoor forced to seek employment elsewhere. Fre-1990; Swaziland, Ministry of Education quently in the past, personnel with high-1986). Care should be taken to ensure that level skills have been attracted to other jobs,the pressures to perform well on public ex- particularly in private industry, where sala-aminations do not lead to an exacerbation ries are higher.of problems of repetition and inefficiency inschools. In some countries, making examinations
largely self-financing through examinationFinally, as has already pointed out in fees has helped provide financial resources
Chapter 4, the ranking of schools in terms to enable the authority to operate with someof examination success, especially when no degree of independence.account is taken of the entry characteristicsof pupils, their home circumstances, or the Whatever the location of an examinationscharacteristics of schools, can be unfair and body, it is crucial that it should be indepen-can have disastrous effects on the morale of dent in the exercise of its functions. In someschools.
62
countries the examination body is located Teachers can also be involved in con-in a Ministry of Education; in other coun- structing examination questions. For ex-tries, it is a separate body, though working ample, teachers might contribute items to ain close co-operation with the Ministry of national item bank, from which examina-Education. Elsewhere, the examination body tion questions might be selected. Apart fromis dosely associated with a university; within involving teachers directly in the examina-some countries, in fact, universities and other tion process, this practice might also serveauthorities administer different public ex- to integrate public examinations with theaminations. There are factors in favor and internal assessment procedures of schools.against each arrangement. The decision that Teachers would come to know what thingsis made in any individual country about the are assessed in public examinations andlocation and control of the examination body would follow similar practices in their ownwill probably be largely determined by the in-school assessments. At the same time,country's tradition and experience. teachers would have an input into the ex-
amination system ensuring that what isNational Cooperation asked in the public examinations is relevant
and appropriate to their students.Examinations represent one part of the
educational process. Structures should be de- Communication between examinationveloped to allow the examinations author- boards and teachers could also be improvedity to work in close liaison with bodies re- considerably through the introduction ofsponsible for educational administration and newsletters which would include detailedcurricula. Cooperation can help ensure that results of pupil performance and how to dealthe public examinations serve the needs of with problems.all students in the system, improving thegeneral quality of education and satisfacto- Other forms of communication from ex-rily serving the functions of certification and amination boards include providing infor-selection. It can also help to ensure that mation about standards through the schoolexaminations adequately reflect curricula inspectorate and selected teachers, the pro-and not only that, but that they are seen to vision of in-service courses, and respondingdo so. Public examinations that reflect na- to individual queries from teachers and othertional curricular priorities can serve as a agencies. Pre-service teacher training shouldpowerful positive change agent in the edu- include courses on assessment. Participantscational system. in these courses should be introduced in a
practical way to a variety of assessment pro-The examinations body should also cedures.
work closely with teachers. Most examina-tion bodies make use of teachers to mark International Cooperationexamination scripts. This can be a good wayof letting teachers know what is important In considering the reform of examina-in examinations. Obviously, it is important tions, the need for cooperation and sharingto have as many teachers as possible in- of scarce resources should be recognized. Ex-volved in this exercise so that first-hand in- changes of item writers, markers, and otherformation about examinations is generally technical personnel, as is commonly carriedavailable in the educational system. out in a small number of countries, should
63
be fostered. Security threats and procedures ary leaving-certificate examinations. Im-should be reviewed. While countries are not proving the quality of public examinationslikely to administer common public exami- will of necessity require examination au-nations, international cooperative studies thorities to change existing modes of assess-might be conducted on common curricular ment. In particular, examination officialsareas. Perhaps the main advantage of such should be requested to improve the mea-comparative studies is the focus it could surement of higher-order cognitive skills,bring to bear on existing curricular goals and and skills related to everyday life, and toobjectives in individual countries. Regional develop techniques of oral, aural, and prac-organizations such as the recently founded tical assessment.Association of Heads of Institutions Respon-sible for Examinations and Educational As- Fornal Trainingsessment in East and Southern Africa shouldbe encouraged and supported. Such orga- Well developed, advanced-degreenizations can provide a structure within courses in measurement and assessmentwhich regional and international coopera- should be offered by at least one universitytion can be promoted. in an Anglophone area and one in a Franco-
phone area. In addition to recruiting stu-Number and Format of Public Examinations dents in the normal manner, special leave
arrangements should be made to allowConsideration should be given to reduc- members of national examination authori-
ing the number of public examinations ad- ties to enrol. Courses offered should coverministered. Economies incurred should be item construction, psychometric theory, re-devoted to improving the quality of the pri- search design and methodology, curriculummary leaving and the senior-cycle second- theory and design, computer applications,
and management.
64
7 Conclusion
In considering the reform of public ex- sion for school supervision, will limit theaminations, it has to be recognized that some options that are available to the examina-of the problems associated with examina- tion reformer.tions are not due to the format or adminis-tration of the examinations per se, but to Having said that, it should be also rec-structural aspects of the educational systems. ognize that unreformed examination systemsFor example, the severe shortage of suffi- are likely to cause damage to the quality ofcient secondary-school places would, irre- education offered in schools. It obviouslyspective of the quality of the examination makes sense to take steps to change suchused for selection purposes, inevitably have systems and to develop examinations whicha strong limiting influence on teaching style are more in keeping with principles of goodand the nature of student learning towards assessment practice. Well designed and re-the end of primary school. Thus, in the final sponsibly used assessment can be an impor-analysis, it needs to be stressed that the role tant source of information about what andof public examinations in contributing to im- how well students are learning and whichprovements in achievement levels will be ones are most likely to make the most oftempered by a range of factors over which opportunities that cannot be provided forexamination authorities have little or no con- all (see NCTPP 1990.)trol. Unless governments and ministries ofeducation tackle the problems created by a Most countries, to varying degrees, havelack of places in schools, poor quality of in- in recent years actually implementedstruction, inadequate supplies of books and changes in their examination systems. Theseequipment, and the inability of students to changes relate to the techniques of assess-understand ithe language of instruction, re- ment, to a broadening of the scope of as-forms in public examinations, no matter how sessment and, in turn and related to this, afar-reaching, are unlikely to be very effec- shift in responsibility for assessment.tive.
In Anglophone countries, the mainWhile factors other than examinations af- change in techniques relates to the introduc-
fect school quality, they also place con- tion of multiple-choice tests to replace free-straints on the kind and extent of examina- response ones.tion reform that is possible. Thus, those con-cerned with proposing changes in examina- But there have been other changes. Intions have to take into account the context some countries, the scope of the examina-in which the changes will be implemented. tions has been broadened considerably byIn particular, the conditions which operate combining school-based assessments within schools, which vary considerably across the results of a terminal examination. Thisthe countries considered in this report, and has increased the involvement of teacherswhich include school facilities, the quality in the assessment process, though theseand availability of text books, number of kinds of assessment are accompanied byunqualified teachers, class size, and provi- strict monitoring procedures.
65
At one level, changes represent an at- is it to do both, and if it is, what kind oftempt to improve efficiency by applying new balance should be struck between the dif-technology. For example, the use of ma- ferent functions? The rhetoric in most coun-chine-scored multiple-choice items rather tries would suggest that the provision ofthan person-scored essays helps to reduce basic skills for the total population is morescoring costs, as non-multiple choice exami- important; present examination practicenations are between two and five times more would often seem to suggest that selectionexpensive to design and grade than mul- is more important.tiple-choice examinations (Heyneman 1986.)The changes may, however, represent more These considerations should not be takenfundamental shifts in the control and role of to imply that an ideal assessment systemexaminations. The use of coursework as- will neatly fall into place once resourcessessment, for example, involves a shift in have been assessed and the values and aimsthe control of examinations and of the selec- of an educational system have been agreed,tion process, which is based on it. Some of insofar as these can be agreed. As has beenthe other changes may herald a shift in em- noted, no examination system is without itsphasis in the role of examinations from one defects. The design of a system will inevi-of selection to one of certification, as well as tably involve a series of trade-offs so thatan attempt to replace assessment-led cur- final decisions will represent a set of provi-ricula by curricula-led assessment. Attempts sional compromises among competing val-to provide a more adequate sampling of cur- ues (Eckstein and Noah 1988; Noah andricula in examinations suggest, in addition Eckstein 1990.) Old and new values willto a concern for curricular validity, a con- most likely continue to compete. For ex-cern with the performance of all children ample, even if greater emphasis is placed inrather than just the ones that are likely to be policy on the education of all children, it isselected to remain in the educational sys- unlikely that the use of examinations fortem and an effort to restructure testing so purposes of selection, even at a relativelythat talent is promoted rather than merely early stage in the educational careers ofscreened or classified. children, can be abandoned.
These comments suggest that important Compromise too will be forced on coun-values may underlie changes in examina- tries by lack of resources. Thus, the guide-tions. Consideration should be given to lines set out in this report to improve as-these values and how they fit with a sessment in public examinations are meantcountry's aims and priorities for its educa- to be just that-directions in which to traveltional system. Insofar as any country can rather than objectives to be achieved in theachieve consensus in its aims for education, immediate future. It is our belief, that if re-it would seem important that, in consider- forms are instituted which are pointed ining examination reform, an attempt should the directions indicated in the guidelines, ifbe made to reach some level of agreement the necessary administrative steps, set outand establish some set of priorities in val- also in the report, are taken, and if prob-ues pertaining to education. For example, lems and undesirable effects are anticipatedis the function of the educational system pri- and steps taken to deal with them, then themarily to select a small cohort of students quality of examinations and of educationalfor higher levels of education or is it to pro- systems in general will improve in Africa.vide basic skills for the total population? Or
66
8 References
Airasian, P. W. (1988). Measurement driven Campbell,D.T.,andFiske,D. (1959). Conver-instruction: A closer look. Educational gent and discriminant validation by theMeasurement: Issues and Practice, 7, 6-11. multitrait-multimethod matrix. Psycho-
Ali, A., and Akubue, A. (1988). Nigerian pri- logical Bulletin, 56, 81-105.mary schools' compliance with Nigeria Catterall, J. S. (1989). Standards and schoolnational policy on education. An evalua- dropouts: A national study of tests re-tion of continuous assessment practices. quired for high school graduation.Evaluation Review, 12, 625-637. American Journal of Education, 98, 1-34.
Asmerson, K., and others (1989). CEB (Curriculum and Examinations Board).Discontinuations of students from institu- (1985). Assessment and certification. Ations of higher learning in Ethiopia. Unpub- consultative document. Dublin.lished document. Addis Ababa: Univer- Christie, T., and Forrest, G. M. (1981). Defin-sity of Addis Ababa. ing public examination standards. London:
Blaug, M. (1973). Education and the employ- Macmillan Education.ment problem in developing countries. Cooley, W.W., and Leinhardt, G. (1980).Geneva: International Labour Office. The instructional dimension study.
Bloom, B. S. (Ed.) (1956). Taxonomy of educa- Educational Evaluation and Policy Analy-tional objectives. The classification of educa- sis. 2, 7-25tional goals. Handbook 1. Cognitive domain. Cronbach, L. J. (1971). Test validation. In R.London: Longman Green. L. Thorndike (Ed.), Educational measure-
Bolger, N., and Kellaghan, T. (1990). Method ment (2nd ed.). Washington DC: Ameri-of measurement and gender differences can Council on Education.in scholastic achievement. Journal of Educa- Crooks, T. J. (1988). The impact of classroomional Measurement, 27, 165-174. evaluation practices on students. Review
Bray, M., Clarke, P. B., andStephens, D. (1986). of Educational Research, 58,438-481.Education and society in Africa. London: Eckstein, M. A., and Noah, H. J. (1988). TheEdward Arnold. form and functions of secondary school
Brooke, N., and Oxenham, J. (1984). The leaving examinations: Internationalinfluence of certification and selection on variations. Paper presented at 32nd An-teaching and learning. In J. Oxenham nual Conference of Comparative and In-(Ed.), Education versus qualifications? A ternational Education Society, Atlanta,study of relationships between education, Georgia.selection for employment and the productivity Eisemon, T. 0. (1988). Benefitting from basicof labour. London: Allen and Unwin. education: School quality and functional
Cambridge Educational Consultants. (1988). literacy in Kenya. Comparative and Interna-The Ethiopian School Leaving Examination: tional Education Series, 2. Oxford:A strategyfor improvement. Report for the Pergamon.Commissionfor Higher Education, Addis Eisemon, T. 0. (1990). Examination policiesAbaba. Cambridge. to strengthen primary schooling in Afri-
67
can countries. International Journal of Edu- Hallak, J. (1990). Investing in the future: Set-cational Development, 10, 69-82. ting educational priorities in the developing
Eisemon, T. O., Patel, V. L., and Abagi, J. world. Oxford: Pergamon.(1987). Read these instructions carefully: Hawes, H. (1979). Curriculum and reality inExaminationreformandimprovinghealth African primary schools. Harlow, Essex:education in Kenya. International Journal Longman.of Educational Development, 8, 55-66. Heyneman, S. P. (1985). Investing in educa-
ERGESE (Evaluative Research of the General tion: A quarter century of World Bank expe-Education System in Ethiopia, Secretariate). rience. Seminar Paper No.30. Washing-(1986). A quality study: Summary report ton, DC: World Bank.presented to the Executive Committee of Heyneman, S. P. (1986). Testing to maximizeERGESE. Addis Ababa. national economic performance. In S.
Fafunwa, A. B. (1974). History of education in Heyneman and I. Fagerlind (Eds.), Maxi-Nigeria. London: Allen and Unwin. mizing national economic performance: The
Foster, P. (1965). The vocational school fal- role of educational testing. Washington, DC:lacy in development planning. In C. A. World Bank.Anderson and M. J. Bowman (Eds.), Educa- Heyneman, S. P. (1987). Uses of examina-tion and economic development. Chicago: tions in developing countries: Selection,Aldine. research, and education sector manage-
Fredericksen, N. (1984). The real test bias: ment. International Journal of EducationalInfluence of testing on teaching and learn- Development, 7,251-263.ing. American Psychologist, 39,193-202. Heyneman, S. P. (1988). Improving univer-
Fuller, B. (1986). Is primary school quality sity selection, educational research, anderoding in the third world? Comparative educational management in developingEducation Review, 30, 491-507. countries: The role of examinations and
Fuller, B. (1987). What school factors raise standardized testing. In S. P. Heynemanachievement in the third world? Review of and I. Fagerlind (Eds.), University examina-Educational Research, 57, 255-92. tions and standardized testing. Washington,
Fuller, B., and Heyneman, S. P. (1989). Third DC: World Bank.world school quality. Current collapse, Heyneman, S. P., and Loxley, W. (1983). Thefuture potential. Educational Researcher, effectof primary school qualityon achieve-18, 12-19. ment across twenty-nine high and low-
Gifford,P.,and Weiskel, T. C. (1971). African income countries. American Journal ofeducation in a colonial context: French Sociology, 88, 1162-94.and British styles. In P. Gifford and W. R. Heyneman, S. P., and Ransom, A. W. (1990).Louis (Eds.), France and Britain in Africa. Using examinations and testing toNew Haven, CN: Yale University Press. improve educational quality. Educational
Greaney, V. (1980). Developing national Policy, 4,177-192.attainment measures: A study of Irish Heyneman, S. P., and White, D. S. (Eds.).experience. International Review of Educa- (1986). Thequalityofeducationandeconomiction, 26, 3-16. development. Washington, DC: World
Gronlund, N. E. (1973). Preparing criterion- Bank.referenced tests for classroom instruction. ILO. International Labour Office. (1972).New York: Macmillan. Employment, incomes, and equality: A strat-
Guskey, T. R., and Kifer, E. W. (1990). Rank- egy for increasing productive employment ining school districts on the basis of state Kenya. Geneva.wide test results: Is it meaningful or mis- Irvine, S. H. (1968). A five-year follow-up ofleading? Educational Measurement Issues secondary school selection procedures inand Practice, 9, 11-16. centralAfrica, 1962-1967. BritishJournalof
68
Educational Psychology, 38, 208-211. Kwasna, K. B. (1974). Content validity andJones, T. J. (1922). Education in Africa. New reliability of domain-referenced tests.
York: Phelps-Stokes Fund. African Journal of Educational Research, 1,Jones, T. J. (1924). Education in East Africa. 73-79.
New York: Phelps-Stokes Fund. Le Mahieu, P. G. (1984). The effects onJoywathsing, M., and others. (1988). The achievement and instructional content of
private costs of education in Mauritius. a program of student monitoring throughReduit: University of Mauritius. frequent testing. Educational Evaluation
Kahn, M. J. (1990). Some questions concern- and Policy Analysis, 6, 175-187.ing the standards of external examina- Lesotho. (1982). The education sector survey.tions. Studies in Educational Evaluation, 16, Report of the Task Force. Maseru: Govern-513-527. ment of Lesotho.
Kellaghan, T. (1961). Some implications of Lewin, K. (1984). Selection and curriculumbilingualism for education in Nigeria. reform. In J. Oxenham (Ed.), EducationIbadan, 11, 31-33. versus qualifications? A study of relation-
Kellaghan, T. (1968). Abstraction and catego- ships between education, selection for employ-rization in African children. International mentand theproductivityof labour. London:Journal of Psychology, 3, 115-120. Allen and Unwin.
Kellaghan,T. (1990). Examinationsystemsin Linn, R. L. (1983a). Curriculum validity:Africa. In P. Broadfoot, R. Murphy, and Convincing the courts that it was taughtH. Torrance (Eds.), Changing educational without precluding the possibility of mea-assessment: International perspectives and suring it. In G. F. Madaus (Ed.), The courts,trends. London: Routledge. validity and minimum competency testing.
Kellaghan, T. (in press). Examination sys- Boston: Kluwer-Nijhoff.tems in Africa: Between internationaliza- Linn, R. L. (1983b). Testing and instruction:tion and indigenization. In M. A. Eckstein Links and distinctions. Journal of Educa-and H. J. Noah (Eds.), Examinations: Com- tional Measurement, 20, 179-189.parativeand internationalstudies. NewYork: Little, A. (1982). The role of examinations inPergamon. the promotion of "the Paper Qualification
Kellaghan, T., and Greaney, V. (1989). A Syndrome". In ILO. Jobs and Skillsstudy of examination systems of five African Programme for Africa. Paper Qualifica-coun tries. Report of a study carried out for tions Syndrome (PQS) and unemployment ofHigher Education for Development Co- school leavers. A comparative sub-regionaloperation. Dublin: Educational Research study. Addis Ababa: InternationalLabourCentre. Office.
Kelly, M. J. (1986). An overview of education in Little, A. (1984). Combating the diplomaZambia. Lusaka: University of Zambia. disease. In J. Oxenham (Ed.), Education
Kelly, M.J. (1991). Education in a declining versus qualifications? A study of relation-economy. The case of Zambia, 1975-1985. ships between education, selection for employ-Washington: World Bank. mentand theproductivityof labour. London:
Kelly, M. J., Nkwanga, E.B., Achola, P.P.W., Allen and Unwin.Kaluba, L.H., and Nilsson, K. (1986). Little, A. (1990). The role of assessment, re-The provision of education for all. Towards examined in international context. In P.the implementation of Zambia's educational Broadfoot, R. Murphy and H. Torrancereforms underdemographicand economic con- (Eds.),Changingeducationalassessment. In-straints, 1986-2000. Lusaka: University of ternational perspectives and trends. London:Zambia. Routledge.
KenyaNationalExaminationsCouncil. (1986). Lockheed, M. E., and Komenan, A. (1989).K.C.P.E. Newsletter 1987. Nairobi. Teaching quality and student achievement
69
in Africa: The case of Nigeria and ing and Public Policy. Boston College.Swaziland. Teaching and Teacher Educa- Chestnut Hill, MA.tion, 5,93-113. Nigam, S. B. L. (1982). Preface. In ILO. Jobs
Lockheed, M. E., and Verspoor, A. M. (1990). and Skills Programme for Africa. PaperImproving primary education in developing Qualifications Syndrome (PQS) and unem-countries. A review of policy options. ployment of school leavers. A comparativeWashington, DC: World Bank. sub-regional study. Addis Ababa: Interna-
Lulsegged, A. (1988). Assessment of impact of tional Labour Office.SIDA assistance to the Swaziland govern- Noah,N.J.,andEckstein,M.A. (1990). Trade-ment. Unpublished manuscript. offs in examination policies: An interna-
Madaus, G. F. (1988). The influence of testing tional comparative perspective. In P.on the curriculum. In L. N. Tanner (Ed.), Broadfoot, R. Murphy and H.Critical issues in curriculum. Eighty-seventh Torrance (Eds.), Changing educational as-yearbook of the National Society for the Study sessment: International perspectives andof Education, Part 1. Chicago: NSSE. trends. London: Routledge.
Madaus,G. F.,Airasian,P. W., andKellaghan, Nuttall, D. L. (1990). Proposals for a nationalT. (1980). School effectiveness. A reassess- system of assessment in England andment of the evidence. New York: McGraw- Wales. International Journal of EducationalHill. Research, 14, 373-381.
Madaus, G. F., and Greaney, V. (1985). The Ogundare, S. (1988). Curriculum develop-Irish experience in competency testing: ment: A description of the developmentImplications for American education. of the national curriculum for primaryAmerican Journal of Education, 93, 268-294. social studies in Nigeria. Educational
Madaus, G. F., and Kellaghan, T. (in press). Studies, 14,43-50.Curriculum evaluation and assessment. Oloo, A. (1990). KCPE ranking is wrong.In P. Jackson (Ed.), Handbook of research on Daily Nation (Nairobi), February 3.curriculum. New York: Macmillan. Ongom, D. L. (1990). Malpractices in ex-
McNamara, V. (1982). The long-term effects aminations. Paper presented to seminarof "carrots". School system structure and on Malpractices in Examinations. Zomba,curriculum, SSCEP and the problem of Malawi.motivating learning that is functional. O'Rourke, B., Martin, M. O., and Hurley, J. J.International Journal of Educational Devel- (1989). The Scholastic Aptitude Test as aopment, 1, 49-60. predictor of third-level academic perfor-
Mingat, A., and Psacharopoulos, G. (1985). mance. Irish Journal of Education,23, 22-39.Education costsandfinancingin Africa: Some Oxenham, J. (1983). What examinations testfacts and possible lines of action. Discussion and emphasize: An analysis of the Le-paper. Report No. EDT13. Washington, sotho Primary and Junior Secondary Ex-DC: World Bank. aminations. In Lesotho, The education sec-
Myeni, A. D. (1985). Standardisation and the tor survey. Annexes to the Report of the TaskJunior Certificate programme. SIER Bul- Force. Maseru: Government of Lesotho.letin, 6, 30-40. Pennycuick, D. (1990). The introduction of
Nafziger, E. W. (1989). Inequality in Africa. continuous assessment system at second-Political elites, proletariat, peasants and the ary level in developing countries. Inpoor. Cambridge: Cambridge University P.Broadfoot, R. Murphy and H. TorrancePress. (Eds.), Changing educational assessment.
NCTPP (National Commission on Testing London: Routledge.andPublicPolicy). (1990). Fromgatekeeper Perrott,E. (1982). Effective teaching: Apracti-togateway: Transforming testingin America. cal guide to improve your teaching. London:Report of the National Commission on Test- Longman.
70
Quansah, K. B. (1985). Current testing in Terwilliger, J. S. (1989). Classroom standardAfrica: Selection or measure of real per- setting and grading practices. Educationalformance? Prospects, 15, 281-288. Measurement Issues and Practice, 8(2),15-19.
Sebatane, E. M. (1985). Perspectives of edu- Torrance, H. (1986). What can examinationscational assessment in Lesotho. Studies in contribute to school evaluation? Educa-Educational Evaluation, 11, 269-278. tional Review, 38, 31-43.
Somerset, H.C.A. (1987). Examination reform Uganda, Ministry of Education. (1989). Re-in Kenya. Discussion paper. Washington, port of Education Policy Review Commission.DC: World Bank. Kampala.
Somerset, H.C.A. (1988). Examinations as an Unesco (1988). Statistical yearbook, 1988.instrument to improve pedagogy. In S. P. Paris.Heyneman and I. Fagerlind (Eds.), Univer- Unesco (1990). Basic education and literacy.sity examinations and standardized testing. World statistical indicators. Paris.Washington, DC: World Bank. Windham, D. M. (1986). Internal efficiency and
Swaziland, Ministry of Education. (1985). the African school. Discussion paper.Report of the National Education Review Washington, DC: World Bank.Commission. Mbabane: Ministry of Edu- World Bank. (1988). Education in Sub-Saharancation. Africa. Policies for adjustment, revitaliza-
Swaziland, Ministry of Education. (1986). tion, and expansion. Washington, DC.Wastage in the educational system. Survey on World Bank. (1990). Education in Sub-Saharanthe causes of repetition and drop-out in pri- Africa. Updated statistical tables. AFTEDmary and junior secondary schools in Technical Note No.1. Washington DC.Swaziland.. Mbabane: Educational Re- World Declaration on Education for All.search and Planning Department, (1990). AdoptedbytheWorldConferenceMinistry of Education. on Education for All, Meeting Basic Learn-
Swaziland, Ministry of Education. (1987). ing Needs. New York: UNICEF House.Final report on a management development Yussufu, A. (1989). Promise and problemsconference for senior management. A strate- associated with the assessment of a re-gic approach to continuous assessment and vised curriculum based on behaviouralinstructional leadership. Mbabane: Educa- objectives in the Republic of Kenya. Stud-tional Research and Planning Section, ies in Educational Evaluation, 15, 279-284.Ministry of Education. Zambia, Ministry of Education. (1977).
Tan, J-P. (1985). Private enrollments and Educational reform: Proposals and recom-expenditure on education: Some macro mendations. Lusaka: Department of Tech-trends. International Review of Education, nical Education and Vocational Training.31, 103-117.
71
9 Appendices
Appendix 1National Education Systems
Countries and Compulsorylinguistic school Entrance age and duration of first andstatus age second-level (general) education
5 6 7 8 9 10 1112 13 14 15 16 17 18 19 20
Anglophone
Kenya - P P P P PP P P S S S S
Lesotho 6-13 P P P P P P P Si Si Si S2 S2
Mauritius - P P P P P P S1 S1 S1 S2 S2 S2 S2
Swaziland - P P P P P P P S1 S1 S1 S2 S2
Uganda - P P P P P P P S1 S1 S1 S1 S2 S2
Zambia 7-14 P P P P P P P S1 S1 S2 S2 S2
Francophone
Chad 6-12 P P P P P P S1 S1 S1 S2 S2 S2
Guinea 7-13 P P P P P P S1 S1 S1 S1 S2 S2 S2
Madagascar 6-13 P P P P P S1 S1 S1 S1 S2 S2 S2
Mauritania - P P P P P P S1 S1 S1 S2 S2 S2
Rwanda 7-15 P P P P P P P P S1 Sl S1 S2 S2 S2
Togo 6-12 P P P P P P SI S1l 1S S2 S2 S2
Other
Cape Verde 7-13 P P P P P P S1 S1 S1 S2 S2
Ethiopia 7-13 P P P P P P Sl S1 S2 S2 S2 S2
P=Primary; S1=Secondary Junior Cycle; S2=Secondary Senior Cycle
Sources: UNESCO (1988), and national reports
73
--.Appendix 2TITLES AND DESCRIPTIONS OF EXAMINATIONS
Countries /LinguisticStatus
When Given Title Function Type Responsibility of
Anglophone
KENYA
1. End of primary KCPEb Certification/ Multiple- KNEC(8)a Selection choice,
Essay
2. End of secondary KCSE Certification/ Essay, KNEC(4) Selection Multiple-
choice,Practical,Oral, Aural
LESOTHO
3. End of primary Primary School Certification/ Almost all Ministry ofschool (7) Leaving Certificate Selection Multiple- Education
Examination choice,Letter and Essay
4. End of junior Junior Certificate Certification/ Essay, Examinationsccle secondary Examination Selection Short answer Council of(3) Practical Lesotho
5. End of senior Cambridge Overseas Certification/ Short UCLEScycle (2) Certificate Exam. Selection answer,
Essay
MAURITIUS
6. End of primary CPE Certification/ Multiple- MES(6) Primary Educ. Selection choice,
(CPE) Essay
7. End of first SC Certification/ Multiple- MEScycle secondary Selection choice and UCLESage 16+ (2) Essay,
Practical
8. End of second HSC Certification/ Multiple- MES incyde secondary Selection choice collaborationage 18+ (2) Essay, with UCLES and
Practical University ofLondonG:CE 0and A- Levels
a. Numbers in parentheses indicate the number of grades at a particular level.b. Acronyms defined on page ix.
74
Appendix 2 (continued)
When Given Title Function Type Responsibility of
SWAZILAND
9. End of Prirnary Primary Certificate Certification/ Multiple- Swaziland(7) Selection choice plus Examination
Essay, CouncilPractical,ListeningComprehension
10. End of first Junior Certificate Certification/ Multiple- Swazilandcycle secondary Selection choice plus Examination(Form 3) Essay, Council
Practical,ListeningComprehension
11. End of second Cambridge Overseas Certification/ Short Swaziland Exam.cycle secondary Certificate Selection answer Council and UCLES(Form 5) Essay
UGANDA
12. End of primary PLE Certification/ Short UNEB(7) Selection answer
13. End of first UCE Certification/ Short UNEBcycle secondary Selection answer(4) Practical,
ProjectsOral, Aural
14. End of senior UACE Certification/ Short UNEBcycle secondary Selection answer(2) Practical,
ProjectsOral, Aural
ZAMBIA
15. End of p:rimary Grade 7 Composite Certification/ Multiple- Examinations(7) Examination Selection choice Council of
Zambia
16. End of junior Grade 9 Certification/ Essay, Examinationscycle secondary Junior Secondary Selection Multiple- Council of(2) School Leaving choice, oral Zambia
Examination and aural, practical
17. End of senior Joint School Certification/ Essay, Multiple Examinationscycle secondary Certificate and Selection choice, Council of(3) General Certificate Practical Zambia
of Education Exam.
75
Appendix 2 (continued)
When given Title Function Type Responsibility of
Francophone
CHAD
18. End of primary CEPE Certification Written SEC(6) CEPEA Ministry of
Education
19. End of primary Primary Selection Dictation, SEC, Ministry of(6) Concours arithmetic Education
20. End of first BEPC Certification NA SEC, Ministry ofcycle secondary Education(4)
21. End of first Concours for Selection Essay and SEC, Ministry ofcycle secondary second cycle Oral Education(4)
22. End of second Baccalaureat Selection / NA Nationalcyle secondary Certification University of(3) Chad
GUINEA
23. End of primary Seventh year Entrance/ NA National(6) Selection Education
Ministry
24. End of first BEPC NA NA Nationalcycle secondary Education(4) Ministry
25. End of second Baccalaureat Certifiction/ Written, Nationalcycle secondary Selection Oral Education(O, NfMinistry
MADAGASCAR
26. End of primary CEPE Certification/ Written Ministry of(5) (Regional Exam) Selection Primary Ed. and
ProvincialDirector
27. Annual first School-based Selection/ Written, Schoolcycle secondary Promotion Oral(4)
28. End of first CFEPCES Certification/ NA Ministry ofcycle secondary Selection Pnima and(4) zy
Education
76
Appendix 2 (continued)When given Title Function Type Responsibility of
29. End of second Baccalaureat Certification/ Written Ministry ofcycle secondary Selection Higher(3) Education
30. After two years CFE/FP1 Certification/ Written Ministry ofof post-primary Selection Secondary and
Primary Education
31. After two years CFE/FP2 Certification/ Written Ministry ofof senior cycle Selection Secondary andeducation Primary Education
32. End of second Baccalaurdat Certification/ Written Ministry ofcycle secondary technique Selection Secondary and(3) Primary Education
MAURITANIA
33. End of primary CEE Certification Written Ministry of(6) Education
34. End of primary Secondary Selection Written Ministry of(6) entrance Education
35. End of first BEPC Certification/ NA Ministry ofcyde secondary (Not obligatory) Selection Education(3)
36. End of second Baccalaurdat Certification/ NA Ministry ofcycle (3) Selection Education
RWANDA
37. End of primary Concours Selection Written Ministry ofschool Primary and
SecondaryEducation
38. Term and end of School-based Feedback/ Written, Teacheryear Promotion Oral,
Practical
39. Third, fourth Examens de Placement of NA Committee forand fifth years reclassement students from Programs inof secondary private or Secondary Schools
foreign schools
77
Appendix 2 (continued)
When given Title Function Type Responsibility of
40. NA Le Jury Central Certification/ Oral Committee forSelection for Programs inprivate and Secondary Schoolsrepeatcandidates
41. End of secondary NA Certification Written Committee for(6) (Essay, Programs in
Multiple- Secondary Schoolschoice, Oral)
TOGO
42. End of 6th grade CEPD Certification/ Written DECSelection
43. End of first BEPC Certification Written DECcycle secondary(4)
44. End of first Concours Selection NA DECcycle secondary(4)
45. End of second Baccalaureat Certification Written and DEC andcycle secondary oral Directorate of(3) Third Level
Education
46. End of secondary Concours Selection for Written DECuniversity
Other
CAPE VERDE
47. End of basic Elementary cycle Certification/ Written, Nationalcomplementary examination Selection Oral Examination(2) (Exemption option) Committee,
Ministry ofEducation
48. End of general General secondary Certification/ Written, Nationalsecondary examination Selection Oral Examinationeducation (3) (Exemption option) Committee,
Ministry ofEducation
78
Appendix 2 (continued)
When given Title Function Type Responsibility of
49. End of Diploma of Upper Certification/ Written, Nationalcomplementary Secondary Educ. Selection Oral, Examinationsecondary (Exemption option) Practical Committee,education (2) Ministry of
Education
ETHIOPIA
50. End of primary Grade 6 National Certification/ Multiple- National(6) Examination Selection choice, Examination
School- Board, Ministrybased of Educationcomponent
51. End of Grade 8: Grade 8 National Certification/ Multiple- NEB, Ministry of(2 years - first Examination Selection choice Educationcycle)
52. End of Grade 12: ESLCE Certification/ Multiple- University of(4 years - second Selection choice Addis Ababacycle) (Essay
up to 1977)
79
Appendix 3
Kenya Certificate of Primary Education:Sample Essay and Marker's Comments
That evening my father came home looking Once resettled, he told us that I was amongunusually happy. He told me he had some- the top people in the school and that I wouldthing to tell us, but before he could do that, he get an award for that. He also told us that heasked me to go and call Heri. When we both was to go and check which school had chosencame in we found that everyone else was me tobe one of their students. Iwas overjoyedwaiting for the two of us. My father as usual in such a way that I could hardly move fromwas seated on his armchair with a cigarette in where I was now seated. I could just remem-his mouth. My mother was standing by him ber the way I was so nervous while doing theand my two other sisters were seated at the exams and all the words of good luck fromdining table. We all had an idea of what the my friends and relatives. I was glad that allgood news was but we were unsure. the messages in the cards had been fulfilled
and also the fact that I did not disappoint myWhen everybody had seated down, my fa- parents. This had happened before with Heri,ther began talking. There was suddenly a cry but I had challenged him and beaten him. Iof joy and everbody came rushing towards could remember all the promises made to meme. It was so sudden that I almost fell off my by my uncles and aunts and I felt as thoughseat. Everybody was hugging me and doing my heart was smiling inside.all sorts of funny things. I just could notbelieve my eyes. What I had been waiting for After having supper thatnight,Iwas called toso anxiously was now revealed. It was too my parent's room and asked what I wanted togood to be true. I do not know how I made it, have as a present. I named almost everythingbut passing my examination was the best one could think of and this made them laughnews I had ever heard. None of us could since they knew the situation I was in and incontrol ourselves. Some of us almost jumped which they had also gone through. Since theyhigh enough to touch the ceiling with our could not buy me everything, I was promisedheads. All that was ringing in my head was the most essential things. I would never for-that I had passed with flying colours and to getthatdaybecauseitwasthehappiestdayofwhich school I would be chosen to go. my life and having that feeling inside me
makes me feel that in future I will also do well.My father asked everyone to be silent since hehad not finished saying what he wanted.
Mark Awarded: 37
80
iMarker's Comments:
Although this piece of writing was not the Examples:very best, it did represent that top qualitywvriting which is not very easy to come by. Incorrect verb tenses:Candidates who reached this standard wereactually very few. Except for such errors as We all sit down and listen to him. (We all satthe usage of 'seated' instead of the simple past down and listened to him.)'sat' and such flaws as 'after having supperthat nights-a candidate as this would even My father did not agreed. (My father did nothave scored higher marks had he been more agree.)interesting and had he used a greater varietyof structures in his writing. They took the money and leave him crying.
(They took the money and left him crying.)From the brief analysis of errors in the com-positions included in this newsletter and in Numerous similar errors were made by can-other compositions written by candidates in didates.1985 KCPE we discovered that certain kindsof errors were being made repeatedly. Incor- Source: Kenya National Examinations Coun-rect verb tenses, wrong usage of words, spell- cil (1986)ing errors and errors in syntax were wide-spread.
81
Distributors of World Bank PublicationsARGENTINA FINLAND KUWAIT SOUTH AFRICA, BOTSWANAC.do. Hhc SRL Akatem m Xijuppa MEMRB RR.homdon S iAl O ICABlTW-Gaania Guano. P.O. Box 128 P.O. Boo 5465 Qln nvst esflaid. 165 4th Flor4f453/465 Htlawnkil0 1Od0UdadtyPe1333 Banoo. Aireo SF40101 MALAYSIA P.O. - 1141
AUSTRALIA. PAPUA NEW GUINEA. UniveydM.Iay.Coopve CeTodnPIT SOLOMON ISLANDSB, FRANCE PBookau1127 jai= -tdledb-pimdVANUATU ANDWESTERN SAMOA Wesid Bank Publihaooo. Ktil K-1. n tCA. Book. Ajotourna 66, avetaed'Ia L u Inbo-tiranil Sob-ipaon Svfr648WhitehoneRood 75116 Pai. MEXICO P.O. Bo. 41095
biNtcnn3132 INhOTEClVktmia GERMANY .lolt~~~~~~~~~~~~~~~~~~~NcyEjannesburg 32Z4
UNOiVGR A Aparado Prtl 22-601AUSTRIA ~~~~~~~~~~~~N-V0g1406O T
IaIpa.% Moot D.P SPAIN/t,STRIA PppebdadbAlleeS 46TlnMWrD rl
Cald nd Co. 0.520Boon MOROCCO Ctllo 37bW1 GREECE 3Odd dEhad Markting Mxtne 2501 Madrid
A-l(n1 Wim GREECE~110M I2m.uMwnat. d. dAn.
SEHRAIN 24 Ippodamoa Sonet Pl. RI.h C-b1nde Uoreri CnEtnadona AEDOSE-hnm Reah and Co yt q Athe11635 NETHERLANDS C0)D9 B 39tm.
A.d.tes Lid, DeUndeb:Ir/lnOr-PubSktimS
Eer-d Building No. 42 GUATEMALA P.O. wn SRI LANKA AND THE MALDIVESDPiBoot A5 S-b, Piedri 7SS 7480 AB Hksbegm Lake Hou e Booiahop
l.a Boo 3750 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~- ~~~~~~~ ~P.O. Eo. 244ManamTown317 Za 1 NEW ZEALAND 10p, Sir oitt am A.
Cu tern;d City Hill. Library nd h&o ton. Savien 10 di- MaorathABIANGLADESHE PhrxwNB~Cbob Micl induDebpmnt HONG KONG,MACAO Nv Mar Combo2
A -na -doty (MIDAS) A. 20M L Audad. SWEDENHol re. Rod 16 4 WyndtwnSrt SWEDENd
Dhlumondi R/Arme Wiontg Centr NIGERrIA tdter
Ohak. 1209 2nd Flo UIGERIA Poe Un.ibid RFeng- a 12 Boe 16356CftOi1 H-g Kn-g Th(C0ow0BfidingjdSB1 503 27 Sotcikhom
Bwk.ffcw Pnivatemad Bag 5095156 Nr Abmed Sk INDIA Ibadan 6 ni,Cbitt.gong 40M A1ied Puolihl Pv ate Ltd. n FW-ogr -Wil AB
751 Mount Road NORWAY B= AB
76, KDA Avenue Madr. -6001002 NarWAYt Infwtem cOlo 5-10425 S6ockhobKul" Bok Depatomto
B-.EtIU ofikan PO. Box 6125 Etmutd SWITZERLANDBELInUM 15JN.HesodiaMag N-0602 0s1o6 smg. IaiAiw.' IDeLno-y Ballard Etate I-raitleA. duRi202 Bnbay .409 038 OMAN 6.briao.t
1060 B-R fi rC pe13/14A3ag AltRoad MRE d C peale 3tANADA Ae d A11000d P0.3Boo1
61
3,eobAirport CH 1211 C-.o 11
CANADA Nor Dbihi - 11009D2M tLe DiffusMuCP. 5 1531BrueAmpere 17 Chittaranjan A-nue PAIaSTAN PostI ooaerB-usbemill-, c-b C Icutt - 700 072 MiB 'Y'Ubrauie Pyot
Boodronilbo Qodbeo Cabo&tta .709072 ~~~~~~~~~Mir. Book Ag-try so-r do. Aboon -to.14B SE6 ydvHtud 65, Shtthrmh+QudAmr C- p-. t3312
JCyIdA t HMi tod CnIdlng P.O Bo. No 729 CH 1002 L3e
China FinndI & Ernair Bangalore - 560 09Publihing Houe TANZANIA
D.Fo Si Dng lit 3-5-1129 Kadguda Cmn Rod PER U . A OPd.dUnivo y Pro.9
BeiFng Hydeobad - S33 027 Apartdo 3 S24 Du O SZ99
COLOMBIA ProathIans s. P .d n PioonrTHAILAND
Intoenli. I-tdu NearThako-eBau Nsranngpra pHaItPPINES roDol stAp.rtado Aro 34270 Ahotedabd -3S0C09 Intonatomtw Book Cootr 3S6 Slmn RodBogota DRE. Fifh - 15pm.. U. Buildig gk
P.6.1. Houne Ayala Avenue NbSd BaMakokCOTE ITIVOIRE 16-A A&hokMarg MtoMani TRINIDAD & TOBAGO ANMTIGUAC-to d'dition t deDjii*an Ludorw -226C01
Afdnonno ICEDA) BARBUDA. BARBADOS.04BP 541 LVDONESIA55 DOMWINCA. GRENADA. GUYANA,
AB.dp.04Pbirau Pt1 hOnSirA Pixnit UvORrUGAL JAMAICA. MONTSERRAT. ST.Abidjan 04 PIa,x,, Pt. i-ttangl 37-1in PRD tugal KITTS &k NEVIS, ST. LUCIA.
CfYRU PO RuA,gi 37 Ra Do C-7o0-74 ST. VINCENT & GRENADINESC-YPRUS P.O. Boo 181 10 tmSs-oSb iMvEMRB Lmation Se1vic20 k Syttnac Booties UnitPDO. Boo ENS 89 Watt. S6reetP4ica ISRAEL7 SAUDI ARABIA, QATAR 9-1 t
or-S tUAE Ltd JoorBookStare Trinidd. West idianY-tl.it-b-I.Ld. ~~~~~~~P.O. Box 3196DENMARK P.O. Bo 56055 YO 11P A BE TSanrf-d.Utbosb* Tet Aviv 61560 I UNITED ARAB EMIRATES
Rtoano 11 AtttI lle MEM RB Itdoaio Serice. P.O. Bo 5(97DK-1970Frdikb'org C B- tiP
ITALY Al Al. SU.rtDOMICAN REPUBUC Uico C xs Saer. Sanoi SPA Al Dthr Cato- UNTED KINGDOMEditoraTaler, C prA. Vi DuO Di Clabie. I/l Dih oee M-.WNtT It-Id.
Resurodo6e loabe Ib Cto6nS. 309 C dbo Pose 552 P.O. oP.. L t3Apurtado Port 2190 50125 FPl-oren Rydh AlP, Hmpi CUP34 2POSanto Demoingo Ryd le an ieC3 P
ARAB REPUBLIC OP JAPAN Hji Abdlih Ali-Bldig England
EGYPT. ARAB REpUBLIC OF Ekt Book So-vite K lwld Street VENEZUELA
Al Alrm 37-4 Hongo3.3-Cho Bunkyo.ku 113 PO B 3999 VENU dESeAAl CGl-s Street Tokyo O' AptdU . 60d337
KENYA 33, Motnro-d H-.m n A-vad SCants. 1060-AThe Middle ESot Oboer-eo Ain.c Book Servto (EAl Ltd. P.O. Boo dS7941. Shrif e Seet P.. Boo 45245 JPddhbC iro Nairobi
EL SALVADOR KOREA, REPUBUC OF SINGAPORE,TAIWANF-od. Pan Kcrea Book Crporoti MYANMAR BRUNEIAvenida Mnud Enriquer ujo 3530 P.O. Boo 101, Kwumgwbhamun Pnv.. Ltd.Eddio ISEA. la. Pin Seoul 02-06 lt PF. Pe-FuPlrd-iail
Swn S.lNadmr Bldg.
24 Now hndu hi.1 RoadSigapore 1953
RECENT WORLD BANK TECHNICAL PAPERS (continued)
No. 134 de Haan and Bekure, Animal Health in Sub-Saharan Africa: Initial Experiences with AlternativeApproaches
No. 135 Walshe, Grindle, Nell, and Bachmann, Dairy Development in Sub-Saharan Africa: A Study of Issuesand Options
No. 136 Green, editor, Coconut Production: Present Status and Prioritiesfor Research
No. 137 Constant and Sheldrick, An Outlook for Fertilizer Demand, Supply, and Trade, 1988/89-1993/94
No. 138 Steel and Webster, Small Enterprises under Adjustment in Ghana
No. 139 Environmnent Department, Environmental Assessment Sourcebook, vol. I: Policies, Procedures,and Cross-Sectoral Issues
No. 140 Environment Department, Environmental Assessment Sourcebook, vol. 1I: Sectoral Guidelines
No. 141 Riverson, Gaviria, and Thriscutt, Rural Roads in Sub-Saharan Africa: Lessons from World Bank Experience
No. 142 Kiss and Meerman, Integrated Pest Management and African Agriculture
No. 143 Grut, Gray, and Egli, Forest Pricing and Concession Policies: Managing the High Forest of Westand Central Africa
No. 144 The World Bank/FAO/UNIDO/Industry Fertilizer Working Group, World and Regional Supplyand Demand Balances for Nitrogen, Phosphate, and Potash, 1989/90-1995/96
No. 145 Ivanek, Nulty, and Holcer, Manufacturing Telecommunications Equipment in Newly IndustrializingCountries: The Effect of Technological Progress
No. 146 Dejene and Olivares, Integrating Environmental Issues into a Strategyfor Sustainable AgriculturalDevelopment: The Case of Mozambique
No. 147 The World Bank/UNDP/CEC/FAO, Fisheries and Aquaculture Research Capabilities and Needsin Asia: Studies of India, Thailand, Malaysia, Indonesia, the Philippines, and the ASEAN Region
No. 148 The World Bank/UNDP/CEC/FAO, Fisheries and Aquaculture Research Capabilities and Needsin Latin America: Studies of Uruguay, Argentina, Chile, Ecuador, and Peru
No. 149 The World Bank/UNDP/CEC/FAO, Fisheries and Aquaculture Research Capabilities and Needsin Africa: Studies of Kenya, Malawi, Mozambique, Zimbabwe, Mauritania, Morocco, and Senegal
No. 150 The World Bank/lJNDP/CEC/FAO, International Cooperation in Fisheries Research
No. 151 The World Bank/UNDP/CEC/FAO, Tropical Aquaculture Development: Research Needs
No. 152 The World Bank/UNDP/CEC/FAO, Small-ScaleFisheries: Research Needs
No. 153 The World Bank/UNDP/CEC/FAO, Small Pelagic Fish Utilization: Research Needs
No. 154 Environment Department, Environmental Assessment Sourcebook, vol. III: Guidelinesfor Environmental Assessment of Energy and Industry Projects
No. 155 Belot and Weigel, Programs in Industrial Countries to Promote Foreign Direct Investment in DevelopingCountries
No. 156 De Geyndt, Managing Health Expenditures under National Health Insurance: The Case of Korea
No. 157 Critchley, Reij, and Seznec, Water Harvestingfor Plant Production, vol. Il: Case Studiesand Conclusions for Sub-Saharan Africa
No. 158 Hay and Paul, Regulation and Taxation of Commercial Banks during the International Debt Crisis
No. 159 Liese, Sachdeva, and Cochrane, Organizing and Managing Tropical Disease Control Programs:Lessons of Success
No. 160 Boner and Krueger, The Basics of Antitrust Policy: A Review of Ten Nations and the EuropeanCommunities
No. 161 Riverson and Carapetis, Intermediate Means of Transport in Sub-Saharan Africa: Its Potentialfor Improving Rural Travel and Transport
No. 162 Replogle, Non-Motorized Vehicles in Asian Cities
No. 163 Shilling, editor, Beyond Syndicated Loans: Sources of Credit for Developing Countries
No. 164 Schwartz and Kampen, Agricultural Extension in East Africa
The World BankHeadquarters European Office Tokyo Office1818 H Street, N.W. 66, avenue d'1ena Kokusai BuildingWashington, D.C. 20433, U.S.A. 75116 Paris, France 1-1 Marunouchi 3-chome
Chiyoda-ku, Tokyo 100, JapanTelephone: (202) 477-1234 Telephone: (1) 40.69.30.00Facsimile: (202) 477-6391 Facsimnile: (1) 40.69.30.66 Telephone: (3) 3214-5001Telex: wuI 64145 WORLDBANK Telex: 640651 Facsimile: (3) 3214-3657
RCA 248423 WORLDBK Telex: 26838Cable Address: INTBAFRkD
WASHINGTONDC
ISBN 0-8213-2052-1
