Part III: Achievement, Assessment, and Evaluating Learning

Comparative School Achievement
National Case Study Report
International Study of School Achievement
Reflections
The Two Faces of Examinations
Tradeoffs in Examination Policies: An International Comparative Perspective
Secondary School Examinations: International Perspectives on Policies and Practice
An International Perspective on National Standards
A Comparative Assessment of Assessment
An International Comparison of End-of-Secondary School Examinations

Source:   Max A. Eckstein, "Comparative
School Achievement."  Reprinted with permission
of Macmillan Library Reference USA, a
Division of Simon & Schuster Inc., from 
Encyclopedia of Education Research, 5th Edn.
Vol 1:  323-329.  Copyright 1982 by the
American Educational Research Association.

 

COMPARATIVE SCHOOL ACHIEVEMENT

Achievement has been a central topic of interest to comparative educators since Marc-Antoine de Jullien first proposed his comprehensive schema for studying foreign educational systems in 1817 (Fraser, 1964). Throughout the nineteenth century, as school systems were developing in the more industrialized nations of the world, observers traveled abroad to study those practices and policies that might explain differences in the achievements of students and the contributions of a nation's schools to the well-being of their respective societies. American and other educators described and commented upon schooling in other countries and on their presumed outcomes, limited severely by the data available and the lack of research sophistication of their times.

A century later, in the aftermath of two world wars and of the dissolution of great European-based empires, new nations as well as old again concerned themselves with the potential of their school systems to serve their interests: economic growth, political stability, social development, and educational advancement. Comparative study was stimulated by the twin desires to learn from foreign examples and to seek yardsticks against which to measure performance. With the growth in communications through national and international organizations, the accumulation of educational and social data, and the rapid advances in research concepts and techniques, the possibilities of cross-national study of educational achievement were considerably enhanced.

Substantial progress may be observed from the early statistical studies and impressionistic observations of the later nineteenth century to the more systematic empirical studies of the mid-twentieth. Studies of curricula, examinations, textbooks, teacher training, and instructional practices compared across several countries, began to appear with increasing frequency, as did efforts to assess pupil attainment in such areas as arithmetic and reading (Eckstein, 1977). Nevertheless, despite its centrality as a topic for comparative investigation, achievement was relatively neglected in contrast to other aspects of education. The reasons are clear: cross-national assessment of student school performance is fraught with problems of equivalence and comparability, complicated by differences in national objectives and practices, and confounded by verbal and conceptual ambiguities. Small wonder then that comparativists leaned heavily upon such system variables as retention rates and promotion figures from one level to the next. Enrollment or attendance figures are generally available, rates can be calculated from official statistics, and thus seemingly reliable and objective measures may be used. That such figures are themselves not altogether unambiguous is acknowledged, but for international comparison, they appear to be far less troublesome than curriculum content, student performance, and instructional methods.

The IEA Project

The first concerted effort to compare achievement levels according to internationally accepted measures is represented by the massive research project of the International Association for the Evaluation of Educational Achievement: the IEA Project. It began in the late 1950's when researchers from a dozen countries convened under UNESCO auspices to consider the feasibility of conducting such research. The report on the pilot study (Foshay et at., 1962) concerned itself with many of the administrative and methodological problems involved in international collaboration on this scale, while the mathematics study (Husén, 1967) presented the results of the first completed survey of student achievement in twelve countries.

Subsequent phases of the project encompassed six additional school subjects: science (Comber & Keeves, 1973), literature (Purves, 1973), reading comprehension (Thorndike, 1973), English and French as foreign languages (Lewis & Massad, 1975; Carroll, 1975); and civic education (Farnen, Oppenheim, & Torney, 1975). Twenty-one nations participated, though not all were involved in each subject. In addition to the achievement data of samples of students at several school grade-levels, information was gathered on the students' home and school backgrounds through questionnaires administered to principals, teachers, and the students themselves in each country. The IEA Project was an ambitious attempt along the lines of the Coleman study in the United States and the Plowden report in Britain to perform simultaneous national replications. And the central purpose of all this activity was to answer the question "What factors best explain differences in student achievement?" (Postlethwaite, 1974; Härnqvist, 1975).

That differences in achievement existed was clearly evident. In each subject at each age and grade, achievement was compared at several levels: among students within nations, among schools within nations, and among nations. In mathematics, for example, Japanese students scored higher than those in all other nations at the same age/ grade levels. Although the differences were not great, younger students (primary level) from Sweden and Italy performed better in the reading comprehension tests, while lower secondary students from New Zealand and Italy did well. In the same subject, average national scores of older students differed considerably, closely associated with the extent to which nations retained students of the appropriate age-group through the final year of secondary schooling. In English as a foreign language, Swedish students performed better than those of the other nine nations participating. And in science, secondary school students from New Zealand and Germany led those of other nations involved in the study.

It was inevitable that some educators would respond to the fact that students from their own nation performed higher or lower than students from other countries in a subject. However, the researchers properly insisted that national averages could in no way be regarded as the results of an international competition, for obvious reasons. Their quest was for the associated factors that might explain the differences observed.

The six volumes on achievement in the several subjects cited above contain a host of data and interpretations relating to their own areas, but three additional IEA publications review the project as a whole: a technical report on the methodology (Peaker, 1975), a summary discussion of the findings as a whole (Walker, 1976), and a review of the findings in relation to differences of school-system structure and organization and socio-economic characteristics of the twenty-one nations involved (Passow et al., 1976). Looked at in its entirety, the project confirmed and extended much that was known or suspected about the factors affecting student achievement: home background, comprising essentially the educational and social status of parents, by far the most influential force; school characteristics of various kinds; and features of the national educational system. But the relative significance of individual factors and of groups of variables was found to vary considerably among different countries, age levels, and subjects of study, provoking new questions about why achievement levels vary.

A number of more specific points may be selected from the total project to illustrate its capacity to illuminate and to provoke new questions. Although home background factors tended generally to be more important than schoolrelated characteristics, this fact varied greatly by subject, by grade level, and by nation. In science and foreign languages, for example, school factors were generally of greater significance than they were in reading comprehension and civic knowledge. Although, as expected, nations differed in curriculum content and curriculum emphases in particular subjects, they also differed in their orientations toward the subject matter itself and the very nature of learning. The mathematics study had revealed, for instance, that students of the United States were more inclined to guess than those from Belgium; the literature survey showed that students' attitudes toward literature and their approaches to interpreting and evaluating it also differed according to nation or culture. Certain achievement differences (science, foreign languages) were linked with sex. The study as a whole confirmed the gap that separates the less developed countries of the world (Chile, India, Iran, and Thailand participated in several phases of the study) from the rest, in school resources and student performance as much as in economic development. Finally, among the several more obviously policy-related aspects of the study, it was evident that national achievement norms tended to be lower in those nations that maintained nonselective school systems and relatively open access to upper secondary education.

Although the IEA Project is a landmark in quantitative comparative research -- rigorous in design, catholic in perspective, comprehensive in scope (Inkeles, 1977) -- its very strengths may also account for some of its limitations. The decision to perform a cross-sectional study and use linear multiple regression as the primary analytic tool imposed certain limitations from the beginning. Inevitably, explanations for the phenomena observed could only be based upon correlations among variables, a far cry from causal associations (Bridge, Judd, & Moock, 1979). The variables themselves comprised groups of separate but presumably associated indicators for samples of specified student Populations, expressed in average values. And the cross-sectional approach made it impossible to investigate the developmental aspects of education, except by tentative inference. Just as causation cannot be assumed from correlation, the cumulative effects of schooling cannot be adequately considered merely by comparing the achievements of three (or more) different age levels at one time. Average values may obscure suggestive or significant relationships, and the difficulties of actual or proxy variables remain intractable. However, these deficiencies are generic to this particular research mode and common to all similar social science research. The IEA Project, despite its limitations, goes a long way to realizing many of the objectives of comparative study of education as outlined by Jullien over a century and a half ago.

That such issues have been taken to heart is indicated by certain new approaches in IEA work currently under way: revised replication of the mathematics study, which will incorporate a longitudinal dimension; smaller-scale observational studies of classroom interactions, which will attempt to remedy the neglect of subtler, qualitative aspects of teaching-learning processes; and a movement from largely uniform international measures to tests that contain both an international core and sections tailored to the concerns and the curricula of individual participating nations.

Comparative Research Findings

The possibilities of the IEA six-subject study have not yet been fully exploited. The data bank with its mass of educational and related information is available for new and revised analyses by the scholars of the world. Since the publication of the original IEA volumes, a slow but steady stream of studies has used these data for various purposes, often in conjunction with the results of additional research, both in a single nation and in several countries.

Home and school influences The IEA results confirmed what had already been concluded from similar national-survey research studies and a host of more limited ones: on the whole, home background matters more than school-related variables in accounting for differences in student school performance. With one exception, specific school characteristics did not clearly or consistently relate to variance in student cognitive achievement.

Some have used this broad generalization to argue that in the absence of political or economic reforms that would radically alter the nature of home and society, changes in educational policy and practice can have little effect. Yet the evidence suggests otherwise. Although most school factors examined in the IEA Project bore no direct, consistent relation to achievement, the indicators of "opportunity to learn," that is, time devoted to teaching and studying a particular portion of knowledge in a subject, were important cross-nationally. Furthermore, school factors in general were substantially more influential in certain subject areas rather than in others -- in science and foreign languages, for instance, rather than in reading comprehension in the native language and in civic education. The relative influence of school compared with home factors tended to increase with the age of the student. And, finally, the amount of influence ascribed to school factors varied from nation to nation. The conclusion can only be that differences in school policies and practices do matter, but that much remains to be explained: when, where, how, and why?

Several writers have raised the question, "Why do the IEA studies and similar large-scale surveys, such as the Coleman report, shed no more light on the influence of home and school factors on student achievement?" One answer proposed (Marjoribanks, 1973) is that the two environments are rather broad concepts, crudely measured and aggregated , and therefore unlikely to account reliably for more than small proportions of variance. The author reanalyzed a number of small-scale studies from England, the United States, Canada, and Australia, focusing upon the "learning sub environment," those psychosocial factors that appear most likely to affect a student's cognitive performance at school. These include such factors as the intellectuality and "achievement press" of parents and their ambitions, aspirations, and attitudes toward school achievement. Marjoribanks found these to be highly predictive of children's school achievement in their different national settings.

The argument is taken a step further in the analysis of data from one less developed nation (Uganda) compared with certain evidence from the IEA Project (Heyneman, 1976). Although home background appears more influential in the developed countries, the effects of schooling on cognitive achievement appear relatively greater in the case of less developed nations. One inference to be drawn from such work is that there may be a threshold effect at work: or certain levels of social and educational development, school factors are highly influential; at more advanced levels, home background becomes more significant. Furthermore, this effect may be true not only between nations but also between segments or strata of society within a given nation.

On this issue of relative influence, it can be concluded that, in all likelihood, home backgrounds differ from each other far more than schools do, at least for the measures used in large-survey research. Secondly, the very nature of the research design may obscure rather than highlight the associations among factors: the need to use large numbers of indicators; the decision to aggregate a number of measures for a particular variable. The very comprehensiveness of studies that seek to test many elements of a complex explanatory model at one time is also a deficiency. Fine distinctions, qualitative dimensions, and exceptional relationships -- all potentially highly significant -- may be lost in the manipulation of average values. It is left to researchers without the resources and scope of the IEA teams to reanalyze portions of data already collected and, in complementary and exploratory smaller-scale studies, to investigate portions of the conceptual model. In this way, they can direct attention to specific variables in both the home and school environments of students and their relation to one another under specific conditions.

Teachers and instruction The data base and the research model of the IEA Project enhanced the possibilities of systematic comparative study of teacher characteristics and instructional methods. For example, the first global study of reading and writing (Gray, 1956), a UNESCOsponsored project, laid the foundations for comparisons of educational conditions and standards of literacy in the world. Nearly two decades later, a collection of essays and studies (Downing, 1973) delved further into the subject, although still limited by the lack of a unified set of concepts and information. Examples of individual efforts in various subject areas may also be found in Halsall's thoughtful attempt to study attainment differences in French in Holland, Belgium, and England (Halsall, 1963); in Trace's comparison of curriculum and textbook content in the United States and the U.S.S.R. (Trace, 1961); and in Wiersma's study of academic achievement of prospective teachers (Wiersma, 1969). Yet even the evidence from IEA studies provided no clear answers. On the whole, teacher and instructional variables bore no general, consistent relationship with variance in student achievement that was statistically significant.

It was left to subsequent analyses of IEA data and to reviews of additional information to cast light on the subject. One study concluded (Avalos, 1980) that neither higher academic qualification nor longer preservice preparation of teachers were in themselves important in explaining differences in student achievement, although they might be in conjunction with other variables. The same author also found that differences in instructional method were not influential, although she found discovery methods more effective than expository teaching at higher levels of intellectual achievement. This study, of data relating to less developed nations, is in part substantiated by two additional works (Husén, 1978; Simmons & Alexander, 1980). Although neither found clear and consistent, significant relationships among teacher training, several other school-related variables, and achievement, Husén's analyses firmly rejected the null hypothesis that the sixteen teacher-related variables studied were unrelated to achievement. Four characteristics were rather more important than others: qualifications, experience, amount of education, and knowledge. In addition, two demographic characteristics were important under specified conditions: teacher's sex and teacher's age (older teachers may be more successful with older, that is, upper secondary students). Finally, positive teacher expectations, so far as they could be identified, tended to produce positive results.

Simmons and Alexander (1980) found that teacher certification and academic qualifications were not so important at primary and lower secondary as at upper secondary levels and in certain subject areas (notably science), However, in their search for evidence to influence educationalpolicy decisions in less developed nations, they found teacher experience did have a positive effect on academic achievement in the lower grades (although it was not so important at upper secondary levels). In general, they concluded that gross expenditures on teacher salaries and school facilities were not significant, but that teacher motivation (as indicated by time spent on preparation and by membership on curriculum-reform committees) was a positive factor in student achievement. Finally, Simmons and Alexander found that the amount of homework done, the physical conditions at home, and the amount of reading done were all important predictors of student achievement. The conclusions are that increasing the quality or quantity of most of the traditional inputs to schooling, such as teacher training or expenditures per student, is not likely to increase student achievement. However, affective skills taught by the schools may be more important than cognitive skills, especially for post-school benefits (higher earnings and satisfactions in work).

Finally, still with reference to practice in less developed nations, Heyneman (1978) reviewed the published evidence on the relationship between availability of textbooks and academic achievement. Studies covering twelve nations were reviewed, including the IEA Project, Heyneman's own investigations in Uganda, and a number of other works. The availability of books is a consistently good predictor of academic achievement, Heyneman concluded, although the reasons why the associations are stronger in some cases and weaker in others are not at all clear. He recommends that investment in reading materials is likely to improve cognitive achievement in less developed nations.

All of the studies reviewed in this section exemplify the potentialities and the limitations of quantitative analysis made possible by IEA and similar investigations. They raise questions about the conventional wisdom on which educators base their actions; they are suggestive about possible and conceivably unanticipated association among variables; and they indicate most clearly that smaller-scale, rigorous studies of particular sets of phenomena are necessary to complement the broader surveys.

Sex and achievement Much remains to be discovered about the relationship between sex and school achievement. Although the issue of separation of the sexes in school is quiescent in many countries, it remains controversial in various places because of social and religious values and customs. However, comparative study reveals that even where schools have been integrated, achievement often depends on the sex of the student. The sex of the teacher is probably also an important factor.

The IEA studies extended the general awareness that, on the average, sexes perform differently in given school subjects; what was known to be true in certain instances, was found to be true internationally. Boys do better than girls in civic education, mathematics, and science, with the exception that, in some countries, girls excel in biology. On the other hand, although the differences are sometimes slight, girls tend to do better in foreign languages, reading comprehension, and literature. Such findings add to what is known about sex differences in education globally: literacy skills as estimated by United Nations statistics, enrollment and attendance figures, curricular and vocational choices, and achievement in specified school subjects (Finn, Dulberg, & Reis, 1979).

It can be no coincidence that male teachers appear to be more successful with their students' science achievement too, while their female colleagues are better able to enhance their students' foreign-language performance (Husén, 1978). As one comprehensive review of the topic makes evident (Finn, Reis, & Dulberg, 1980), patterns of performance are inextricably bound up with behavior models suggested in schools by teachers and textbooks, by curriculum exposure, by academic supports, and by vocational expectations and opportunity -- all of which are deeply rooted in society's ideas and practices.

Yet the evidence is not all in, and the psychological and sociological dynamics remain unclear. Why, for instance, does it appear that male teachers influence male students positively in the middle grades, but negatively at upper secondary levels (they also influence female achievement negatively at the upper secondary level) (Simmons & Alexander, 1980)? In a four-nation study of reading achievement (Johnson, 1973-1974), boys scored higher than girls in Nigeria and England, while the reverse was true of samples of primary pupils from Canada and the United States. And a selection of IEA data (Passow et al., 1976) suggests that in those more developed nations of the world, where primary-school teaching is perceived as a career for females, primary-school achievement in basic skills (mathematics, reading comprehension) tends to be lower; where primary-school teaching is seen as a career for either sex, such achievement is higher. Comparative study demonstrates that school-achievement differences between the sexes are not easily or quickly reduced, even as social practices develop, economic conditions change, and school practices vary.

National School Policies The continuing debates over how particular school policies affect performance are unlikely to be stilled easily, for they are often rooted in fundamental differences of philosophy, political ideology, and social values. It should make a difference whether compulsory schooling begins earlier or later, how long it lasts, whether classes are smaller or larger, at what age students are moved from primary to secondary levels of schooling, and what mechanisms, if any, are use at important transfer points. The controversies over the effect of classroomgrouping policies or the merits of selective versus comprehensive schooling cannot be assessed with respect to student achievement alone, for they are also involved with political and social effects.

These and many similar issues have been studied both in one-nation case studies and in cross-national investigations. In the IEA Project especially, in addition to the several volumes devoted to the separate subjects of study, the National Case Study volume (Passow et al., 1976) discussed associations between such issues and student achievement. However, since the nations were not selected as a sample for particular research purposes (participation was on a voluntary basis) and the ranges they represented on such variables were limited (land unrepresentative), no clear associations were found between such factors as national average class size, forms of compulsory schooling, and achievement. However, a number of suggestive points emerged, indicating that further analyses and added data would be fruitful.

The question of the effect of selective versus comprehensive secondary schooling may demonstrate how comparative study may dispel some confusion. The IEA Project produced national achievement norms for each participating nation in different subjects for students at different levels. On the whole, achievement norms were demonstrably lower in those countries that retained larger proportions of their youth in the education system by means of nonselective transition from primary to secondary levels and by providing various forms of comprehensive secondary schooling. It is to be expected, therefore, that such countries would have a wider range of student ability in the samples tested at middle and upper secondary school levels, and that national averages would consequently be depressed in comparison with those countries restricting advancement through the school system. However, as IEA analyses demonstrate, if comparisons are made among the top 5 to 9 percent of achievers in each country, the differences among countries are sharply reduced. The best students tend to achieve at very similar levels in different countries, regardless of whether the school system is more selective or more comprehensive.

The issue of how increased access to more schooling may affect outcomes is but one of the current concerns in the United States as in other nations of the world. So too are the issues of how to provide for cultural, linguistic, and other "exceptional groups" whose achievement levels are demonstrably below national norms and whose participation in the mainstream of national life is limited. Bilingual education programs and compensatory school schemes of various kinds have developed in many countries of the world in order to meet similar kinds of problems: transient foreign workers in Sweden and West Germany, the Francophone communities of Canada, Asian refugees in the Netherlands, Indian and Jamaican immigrants in Britain, the poor in rural or urban areas everywhere. Such efforts to use the school system to achieve particular social as well as educational objectives are increasingly being described and analyzed in the comparative literature.

However, as these examples of policy questions indicate, student attainment in basic skills such as reading the native language or in standard school subjects is but one way of defining the outcomes of different school policies. School achievement may also be considered as the capacity of the school system to produce what the educators, citizens, and the leadership of a nation deem important. To evaluate the performance of an education system calls for some understanding of the goals, costs, demands, and needs of the nation for which it provides. Although societies may agree on certain broad social and personal objectives of education, the many varieties of practice, organization, and criteria for evaluation among the schools of the world indicate that, in fact, the ends, the means, and the processes connecting them may vary considerably. Thus, Coombs and Lüschen (1976) propose four criteria by which to assess system performance comparatively: effectiveness, efficiency, responsiveness, and fidelity. They acknowledge the existence of many output measures, some more usable than others, but note the problems involved that may explain why so little has yet been achieved in comparing the achievements of school systems relative to their particular respective priorities and objectives. They conclude with a number of hypotheses suggestive of policy-oriented research.

Valuable as the comparative research has been, its potential for informing specific policy has been severely limited. One cause for this fact may be that the researchers have not adequately translated their findings into forms that educational practitioners and policy makers can grasp. In this respect, comparativists may be at one with other researchers in education and the social sciences generally. A second cause, quite evident in the IEA Project, ties in the research strategy that resulted in conclusions about the relative importance of one factor compared with another over entire nations. Eckstein (1977) summarizes the argument:

Teachers, curriculum makers, and educational policy makers, however, usually wish to know something more specific. They are more interested in those variables over which they have some control than in those less amenable to their decisions. They need to know the effect upon achievement of varying a particular item under rather specific circumstances. They are less interested in influencing achievement on an average, national basis than in, say, rural as compared with urban settings, boys vis-à-vis girls, students in poor neighborhoods as distinct from wealthier communities. What provides the largest increments to achievement for low achievers? For average or high achievers? The answers to these and similar questions require analyses that partition the national samples (singly and across groups of nations) so as to investigate relationships among variables for specified groups of students, e.g., rural/urban, poor/wealthy, high achievers/low achievers. The potential of the I.E.A. studies to inform policy making in education was neglected because insufficient attention was given to policy questions and because authors did not take care to express their findings in appropriately concrete form. (pp. 354-355)

Conclusions

The scope of this article has been restricted in two important senses. First, the word "comparative" has been taken to mean cross-national, although this is by no means the only possible usage. Much educational research is "comparative" in the intra-national sense and faces similar possibilities and obstacles. Second, the term "school achievement" has been taken to mean student cognitive performance, rather than other contributions of a school system to individual or social benefit. As a result of these limitations, a substantial body of literature on both the politics and the economics of education has been excluded (e.g., Blaug, 1978; Merritt & Coombs, 1977; Messialas, 1977). It should also be noted that the discussion is based on two significant assumptions. The explanatory model for school achievement presupposes that there are explicit causes of differences in student performance and that they may be discovered in the patterns of relationships among a number of environments: the home, the school, and the larger contexts of the whole culture or society. And, it is held that comparative school research, whether of achievement or of other aspects of schooling, contributes substantially to our understanding of the complex processes of education and thereby to informing policy.

By moving outside the boundaries of a single nation, comparative study enables the researcher to include variables that may not be available at home. For instance, some practices, such as beginning age and duration of compulsory schooling, are generally uniform within a nation, so that the effect of varying this policy cannot be studied without comparison. Similarly, it may be possible to increase the range of variability of a given factor by using cross-national data when a given nation exhibits only limited variations. In fact, certain important policy differences and outcomes can only be investigated through cross-national comparison. As Wolf (1977) effectively demonstrates in his discussion of student performance in the United States in the IEA Project, reference to the larger number and wider range of variables represented in other countries can be most illuminating.

A second value of comparative study is its capacity to extend generalizations, to expand the scope and validity of a given finding. The conclusion of a one-nation study in education may hold true for a particular school system, but of necessity omits consideration of important national variables. Comparative study broadens the applicability of the conclusion, or, if not, poses new questions about the educational processes under investigation.

This is in fact the third value of comparative studies in achievement: their heuristic potential. They increase the number and variety of phenomena for study; challenge the conventional wisdom in education; and, most important of all, as Bloom (1976) suggests, enhance the theoretical models posited to explain achievement.

All questions concerning school achievement do not require cross-national treatment, although comparison at a lesser level is likely to be necessary. Nor can the conceptual and operational difficulties of cross-national research be avoided. But the history of comparative education over the past century provides ample evidence of progress in defining important educational questions, developing means for studying them, and revealing, with increasing degrees of specificity, the possibilities and limits of such investigations (Noah & Eckstein, 1969). The variety of educational practices and their outcomes in the many nations of the world may be regarded as a series of natural expenments created by different political, social, and economic circumstances. Comparative study investigates their meanings and seeks to relate them to persistent problems in understanding education.

NOTES

Avalos, B. Teacher effectiveness: Research in the Third World -- Highlights of a review. Comparative Education, 1980, 16, 4554.

Blaug, M. Economics of Education: A Selected Annotated Bibliography (3rd ed.). New York: Pergamon Press, 1978.

Bloom, B. Human Characteristics and School Learning. New York: McGraw-Hill, 1976.

Bridge, R. G.; Judd, M.; & Moock, P. The Determinants of Educational Outcomes. The Impact of Families, Peers, Teachers and Schools. Cambridge, Mass.: Ballinger, 1979.

Carroll, J. B. The Teaching of French as a Foreign Language in Eight Countries. Vol. 5 of International Studies in Evaluation. New York: Wiley; Stockholm: Almqvist & Wiksell, 1975.

Comber, L C., & Keeves, J. P. Science Education in Nineteen Countries. Vol. 1 of International Studies in Evaluation. New York: Wiley; Stockholm: Almqvist & Wiksell, 1973.

Coombs, F. S., & Lüschen, G. System performance and policy-making in West European education: Effectiveness, efficiency, responsiveness and fidelity. International Review of Education, 1976, 22, 133-153.

Downing, J. Comparative Reading: Cross-national Studies of Behavior and Processes in Reading and Writing. New York; Macmillan, 1973.

Eckstein, M. A. Comparative study of educational achievement. Comparative Education Review, 1977, 21, 345- 357.

Farnen, R. F.; Oppenheim, A. N.; & Torney, J. V. Civic Education in Ten Countries. Vol. 6 of International Studies in Evaluation. New York: Wiley; Stockholm: Almqvist & Wiksell, 1975.

Finn, J. D.; Dulberg, L.; & Reis J. Sex differences in educational attainment: A cross-national perspective. Harvard Educational Review, 1979, 49, 477-503.

Finn, J. D.; Reis, J.; & Dulberg, L. Sex differences in educational attainment: The process. Comparative Education Review, 1980, 24, S33-S52.

Foshay, A. W.; Thorndike, R. L.; Hotyat, F.; Pidegon, D. A.; & Walker, D. A. Educational Achievements of Thirteen-year-olds in Twelve Countries. Hamburg: UNESCO Institute for Education, 1962.

Fraser, S. Jullien's Plan for Comparative Education, 1816-1817. New York: Columbia University, Teachers College, 1964.

Gray, W. S. The Teaching of Reading and Writing. Paris: UNESCO, 1956.

Halsall, E. A comparative study of attainments in French. International Review of Education, 1963, 9, 41- 59.

Hänqvist, K. The international study of educational achievement. In F. N. Kerlinger (Ed.), Review of Research in Education (Vol. 3.). Itasca, Ill.: F. E. Peacock, 1975.

Heyneman, S. P. Influences on academic achievement: A comparison of results from Uganda and more industrialized countries. Sociology of Education, 1976, 49, 200- 211.

Heyneman, S. P. Textbooks and Achievement: What We Know (World Bank Staff Working Paper No. 298). Washington, D.C.: World Bank, 1978. (ERIC Document Reproduction Service No. ED 179 044)

Husén, T. International Study of Achievement in Mathematics: A Comparison of Twelve Countries (2 vols.). New York: Wiley, i967.

Husén, T. Teacher Training and Student Achievement in Less- developed Countries (World Bank Staff Working Paper No. 310). Washington, D.C.: World Bank, 1978.

Inkeles, A. The international evaluation of educational achievement. Proceedings of the National Academy of Education, 1977, 4, 139-200.

Johnson, D. D. Sex differences in reading scores across cultures. Reading Research Quarterly, 1973-1974, 9, 67- 86.

Lewis, E. G., & Massad, C. The Teaching of English as a Foreign Language in Ten Countries. Vol. 4 of International Studies in Evaluation. New York: Wiley; Stockholm: Almqvist & Wiksell, 1973.

Marjoribanks, K. Psychosocial environments of learning: An international perspective. Comparative Education, 1973, 9, 28-33.

Massialas, B. G. Education and political development. Comparative Education Review, 1977, 21, 274- 295.

Merritt, R. L., & Coombs, F. S. Politics and educational reform. Comparative Education Review, 1977, 21, 247- 273.

Noah, H. J., & Eckstein, M. A. Toward a Science of Comparative Education. New York: Macmillan, 1969.

Passow, A. H.; Noah, H. J.; Eckstein, M. A.; & Mallea, J. The National Case Study: An Empirical Comparative Study of Twenty-one Educational Systems. Vol. 7 of International Studies in Evaluation. New York: Wiley; Stockholm: Almqvist & Wiksell, 1976.

Peaker, G. F. An Empirical Study of Education in Twenty-one Countries: A Technical Report. Vol. 8 of International Studies in Evaluation. New York: Wiley; Stockholm: Almqvist & Wiksell, 1975.

Postlethwaite, N. Introduction, and target populations, sampling, instrument construction, and analysis procedures. Comparative Education Review, 1974, 18, 157- 179.

Psacharopoulos, G. Returns to Education: An International Comparison. San Francisco: Jossey-Bass, 1973.

Purves, A. C. Literature Education in Ten Countries. Vol. 2 of International Studies in Evaluation. New York: Wiley; Stockholm: Almqvist & Wiksell, 1973.

Simmons, J., & Alexander, L. Factors which promote school achievement in developing countries: A review of the research. In J. Simmons (Ed.), The Education Dilemma: Policy Issues for Developing Countries in the 1980s. Elmsford, N.Y.: Pergamon Press, 1980.

Thorndike, R. L. Reading Comprehension Education in Fifteen Countries. Vol. 3 of International Studies in Evaluation. New York: Wiley; Stockholm: Almqvist & Wiksell, 1973.

Trace, A. S. What Ivan Knows that Johnny Doesn't. New York: Random House, 1961.

Walker, D. A. The IEA Six-subject Survey. Vol. 9 of International Studies in Evaluation. New York: Wiley; Stockholm: Almqvist & Wiksell, 1976.

Wiersma, W. A cross-national comparison of the academic achievement of prospective secondary school teachers. Comparative Education Review, 1969, 13, 209- 212.

Wolf, R. M. Achievement in America: National Report of the United States for the International Achievement Project. New York: Columbia University, Teachers College, 1977.