19742 A View from LA7HR No. 33 PRIMARY SCHOOL QUALITY IN MEXICO by Juan Carlos Palafox Juan Prawda Eduardo Velez Human Resources Division Technical Departnent Latin America and the Caribbean Region The World Bank November, 1992 "A View from LATHR ' is a series of occasional flyers produced by the Human Resources Division of Latin America and the Caribbean Technical Department of the World Bank for the purpose of stimulating discussion among staff on key issues facing the sector. The views expressed here are those of the authors and should not be attributed to the World Bank. ABSTRACT. This paper assesses the determinants of academic achievement for primary student completers in Mexico, as a function of different student, family, and school characteristics. Although several scholars have inserted an ability variable in the achievement functions to control for innate characteristics in the determinants of individual cognitive achievement, almost all the literature on this issue refers entirely to advanced industrial countries, and little has been done elsewhere. In this article we contribute to the exploration of the achievement-ability nexus in Mexico. The main findings are the following: (a) the quality of primary education in public schools in Mexico is low; (b) although ability is the best predictor of cognitive achievement, other education inputs, like preschool experience, have significant effects on achievement; (c) age, probably caused by repetition, drop-out, and/or late primary school entry, has a negative effect on achievement; (d) males do better in math than females; (e) students coming from smaller size households attain better cognitive achievement scores; (f) parents with more schooling can have a positive effect on their children's cognitive achievement; and (g) students living in two-parent household tend to do better in primary school. As a result, it is recommended for the Mexican educational sector to: (a) increase the supply of formal and non-conventional early stimulation programs, specially targeted for children living in poor urban and rural areas; (b) increase and strengthen parental awareness concerning the importance of their children's education achievement; (c) implement proven policies and enhance the allocation of proven educational inputs that may decrease repetition and dropout rates in targeted public primary schools, specially in the first grades and in the afternoon shifts; (d) reduce late primary school entry through early stimulation programs and parental awareness dissemination campaigns; and (e) strengthen the educational managerial capacity in the educationally less developed Mexican States. Juan Carlos Palafox is Director of Programacidn Educativa del Distrito Federal de la Direcci6n de Planeacidn, Programaci6n y Presupuesto de la Secretarta de Educacidn P4blica in M6xico. Juan Prawda and Eduardo Velez are staff members of the World Bank. The authors are grateful to Honguy Yang for the technical assistance in preparing the random sample for statistical analysis, and to Julian Schweitzer for having reviewed earlier drafts and contributed with valuable insights. -3- - DnRODUCTION Latin America and the Caribbean (LAC) countries have placed great stock in achieving universal primary education by the year 2000. According to UNESCO's latest data (UNESCO- OREALC, 1990), average primary education net enrollment rate for the LAC region in 1987 was 86%, about 5% higher than at the beginning of the eighties. Yet, in spite of this commendable enrollment increase, school quality, especially in those schools serving rural and urban poor children, is unacceptably low in the region. Average repetition rates in the first two primary grades in LAC, estimated at 42 and 27% respectively is extremely high (Schiefelbein and Wolff, 1992), and academic achievement remains extremely low'. Enhancing school quality is one of the major concerns in almost all the LAC countries. In spite of the growing research literature, there is considerable debate concerning the differential effects of school inputs on student achievement. While in developing countries, education inputs seem to be more important than socioeconomic origin to explain achievement (Heyneman and Loxley, 1983; Fuller, 1987; and Lockheed and Verspoor, 1991), in developed countries, the contrary seems to be the case - family background is more important than school and teacher characteristics in explaining achievement (Coleman et al, 1966 and 1982). Research on educational quality in LAC has been quite weak and scarce, and there has been little analysis of efficiency (Lockheed and Verspoor, 1991). However, the available 2 empirical evidence systematically shows low levels of education achievement in the region , including Mexico at the primary education level (SecreWia de Educaci6n Pdbhca - SEP, 1992, and Guevara, 199 1). Low quality of education is f1requently associated with high repetition and dropout, and as such, its financial implications are enormous, estimated at one billion US dollars for the LAC region (Schiefelbein and Wolff, 1992). THE ASSESSNMW TEST During the administrative decentralization process undertaken by the Mexican Government in the education sector in 1978, a standardized achievement test was implemented to stream primary public school graduates applying for public secondary education in the I According to a recent cognitive achievement test administered by the United Nations Latin America and the Caribbean Economic Commission (CEPAL) in a sample of fourth gmders in Uruguay, half of the students sampled had mastered less than half of the reading and mathematical items of the national core curriculum (CEPAL, 1991). In the rural municipal primary schools in Chile, students have scored around 44% of the national core curriculum in math and Spanish CMe World Bank, 1991). Academic achievement in rural primary and secondary schools in Costa Rica remains extremely low: 20 and 23 % in mathematics and Spanish (fourth graders), and 2 and 7 % in mathematics and Spanish (secondary seniors) (Ministerio, de Educaci6n Pdblica, Memoria, 1987). 2 See footnote 1. -4- Metropolitan Area of Mexico City. The streaming procedure allocated applicants to schools and school shifts of their choice according to test results, resulting in a better match between supply and demand. The policy implemented by the Mexican educational authorities allocated students with higher achievement scores to the morning shift of the secondary of their choice, and in the secondary afternoon shift, students with lower scores. This streaming procedure eradicated the traditional student queuing in front of the school days before registration, and eliminated corruption among school officials who would grant school shift change "services" to concerned parents. During the last twelve years, this selection instrument was extended to the large metropolitan areas of 14 Statese. For the 1991-1992 school calendar year, this instrument was administered to over 352,000 secondary freshman applicants, 32% of those applying to enter first grade in federally financed secondary schools throughout the country. The questionnaire provides information on student's educational and family background, such as: gender, age, number of preschool grades attended, number of people living in the same household, average schooling of both parents, and school shift of public primary attended. The streaming instrument comprises achievement tests in mathematics and in Spanish, and a test on reasoning ability using a set of items of Raven's Progressive Matrices (Raven, 1956). The reasoning ability test has been widely used in developing countries; it involves the matching of pictorial patterns for which literacy and numeracy do not provide an advantage (Psacharopoulos and Velez, 1992; Boissiere, Knight, and Sabot, 1985; Sinha, 1968). DESCRIPTIVE STAT[STICS A sample of 34,422 students, about 10% of the universe of applicants tested in 1991, was randomly selected. According to Table 1, 51% were male; the mean age of the sample was 12.5 years, almost coinciding with the theoretical age cohort demanding secondary education; 63% attended the primary school morning shift. The sample had a mean of 1.8 years of preschool education, which is not surprising given the high preschool coverage existing in Mexico of about 73% of the 5-year-olds and 59% of the 4-year-olds during 1991. About 87% of the sample lived with both parents in the same household containing an average of 6 persons. About 54% of the sampled student's parents and 63% of the mothers had at most primary level schooling ( six grades), whereas 22% of the parents and 20% of the mothers had at most secondary level education (nine grades). The mean averages in reading, mathematics, and reasoning ability were 12.6, 7.2 and 3.4, indicating that the sampled students mastered only 53 and 30% of the national core curriculum in Spanish and Math respectively, and 28% of the items in'the ability test. These low scores are a clear indication that there is ample room to improve the education quality of the public primary schools in Mexico. * Aguascalientes, Baja California, Baja California Sur, Durango, Guanajuato, Hidalgo, Jalisco, M6xico, Quer6taro, San Luis Potosf, Sonora, Tabasco, Tlaxcala, and Zacatecas. -5- Table 1. Mean of Sample Variables Characteristic Mean Value Males 51.0 % Age 12.5 years Morning shift 63.0 % Preschool experience 1.8 years Father's education None 8.0 % At most primary (six grades) 54.0 % At most secondary (nine grades) 22.0 % With more than twelve grades 16.0 % Mother's education None 11.0 % At most primary (six grades) 63.0 % At most secondary (nine grades) 20.0 % With more than twelve grades 7.0 % Living with both parents 87.0 % Family size in the same household 6 people Cognitive achievement in Spanish 12.6 Cognitive achievement in Math 7.2 Raven's ability score 3.4 Source: Sistema Automitico de Inscripci6n y Distribucid6n. Direcei6n General de Planeaci6n, Programaci6n y Presupucsto. Secretarfa de Educaci6n Pblica. M6xico, 1992. A key question for educational planners in Mexico would be to find out if the low educational quality in public primary schools is evenly distributed across the country or whether it is concentrated in some particular States. According to Table 2, the quality of primary education in public schools differs significantly by States. About 88% of the sample originates from educationally more developed States and the capital city, where secondary education enrollment is concentrated4. In the educationally more developed States the mean averages in " According to a composite ranking of Mexico's States based on terminal efficiency, dropout and repetition rates for primary schools during the 1988-89 school year (World Bank, 1990, Appendix 3). -6- reading, mathematics, and reasoning ability represented 53, 31, and 30% of the maximum potential scores. In the educationally developing States the corresponding mean averages were 45, 26, and 21 % of the maximum potential scores respectively. It is not a surprise that the highest scores were achieved in Mexico City, which has the best public primary schools in the country. What is a surprise is that the lowest scores were attained in Tlaxcala, classified as a more educationally developed States. One explanatory hypothesis is that a non-official automatic promotion policy for grades one through five or six, probably encouraged in a discrete fashion by the T1axcalan section of the powerful Teacher's union to keep enrollment high at each public Table 2. Mean Ability and Cognitive Achievement Scores by State State Sample Ability Spanish Math Percent (%) M SD M SD M SD M.City* 43.9 4.0 2.8 13.4 3.7 7.6 3.3 Aguasc.* 3.5 2.4 2.1 9.5 5.8 5.6 3.9 B.Calif.* 1.0 2.7 2.5 11.4 6.0 6.6 4.0 B.Calif.S.* 1.0 3.1 2.6 11.6 5.5 6.5 3.8 Durango** 2.9 2.7 2.2 10.7 5.0 6.5 3.5 Guanaj.** 2.7 2.5 2.0 10.4 5.0 5.9 3.4 Hidalgo** 1.2 2.5 2.1 10.6 5.2 6.4 3.6 Jalisc.1* 11.7 3.5 2.5 13.0 3.8 7.7 3.6 E.Mexico* 25.3 3.2 2.3 12.6 3.6 7.3 3.1 Queret.** 0.8 3.5 2.6 12.7 4.4 7.2 3.4 S.L.Pot.** 2.1 2.5 2.0 11.1 5.3 6.1 3.6 Sonora* 1.2 2.7 1.9 11.7 3.9 6.7 3.0 Tabasco** 1.6 2.3 2.0 10.3 5.2 5.5 3.4 Tlaxcala* 0.4 1.8 2.0 6.5 6.2 3.8 3.8 Zacatecas** 0.6 2.5 1.6 11.6 4.2 7.1 3.1 Ed.Devl.Stat. 88.0 3.6 (30%) 2.6 12.8 (53%) 4.1 7.4 (31%) 3.5 Ed.Dvlp.Stat. 12.0 2.5 (21%) 2.0 10.7 (45%) 5.1 6.2 (26%) 3.4 Notes: 1. Metropolitan area of Guadalajara * Educational developed State ** Educational developing State M Mean value SD Standard deviation Tlaxcala is considered an educationally developed State because it registers low primary school repetition and dropout rates and some of the highest primary completion rates in the country. -7- school in that State to avoid classroom merging, kept the State's aggregate repetition and dropout rates low, and the primary completion high, yet the cognitive achievement of the State public primary school's graduates remained extremely low. Table 3 presents the distribution of the ability and the cognitive scores by several student characteristics. Although there is only a small difference in the ability test in favor of males, there is a more considerable difference in academic achievement. This finding, as explained below, is consistent with existing research developed in other countries (Tobias, 1978). Also as expected, students coming from the public primary morning shift schools performed better than the ones graduating from the public primary afternoon shift schools; students of parents with more schooling are higher achievers; students with more preschool background did better in the standardized test; students living with both parents performed better than those living with one parent; older students fared worse; students coming from crowded households also fared worse; and finally, students from States with better education indicators performed better in the ability and cognitive tests. The following sections provide some explanation of these findings. THE DETERMINANTS OF.ACHIEVEMENT An educational production function was used as the main analytical tool to examine the characteristics that better explain student achievement in the Mexican case. Table 4 presents a series of production functions attempting to identify the main determinants of Spanish and Math cognitive achievement among public primary school graduates in Mexico. Table 4 displays regression equations for scores on Raven's ability and for Spanish and Math with family and student background, including school experience and school shift. Several scholars have inserted an ability variable in the achievement functions to control for innate characteristics in the determinants of individual cognitive achievement (Hauser, 1971; Bryant et al., 1974; Wilson and Portes, 1975). Almost all the literature on this issue refers entirely to advanced industrial countries; little has been done elsewhere. With this article we explore the achievement-ability nexus in studying the effect of some school experience variables controlling for family characteristics among graduates from public primary schools in Mexico. The models in Table 4 present a modest capacity to explain the variance, with coefficients of determination between .087 and .218. Column I presents the results for ability. The unstandardized regression coefficients reflect that the only exogenous characteristic not affecting ability is the region where the students live'. When controlling for other characteristics, males have a significant advantage in ability; older students at time of primary completion present lower ability; although parents education is positively associated with ability, mother's education ' For the definition of region see footnote 4. -8- Table 3. Mean Scores: Ability and Cognitive Output Characteristic Ability Cognitive Spanish Math Sex Male 3.56 12.80 7.39 Female 3.38 12.37 7.12 School shift Day 3.85 13.44 7.93 Afternoon 3.03 11.95 6.58 Father's education None 2.80 11.23 6.59 At most primary 3.18 12.05 6.98 At most second. 3.71 13.11 7.47 At most higher. 4.49 14.43 8.36 Mother's education None 2.89 11.48 6.65 At most primary 3.24 12.18 7.03 At most second. 4.02 13.65 7.81 At most higher. 4.79 14.72 8.60 Student living with Both parents 3.51 12.63 7.31 One parent 3.39 12.65 7.11 Family size* -0.12 - 0.16 -0.08 Student's preschooling None 2.86 11.18 6.44 One year 3.45 12.43 7.22 Two years 3.77 13.03 7.51 Three years 3.81 13.01 7.44 Student's age* -0.12 - 0.17 -0.14 Region Ed.Devl.Stat. 3.55 12.77 7.36 Ed.Dv1p.Stat. 2.53 10.68 6.18 Notes: * Pearson correlation -9- is the only significant one; students living in a two-parent household attain a better ability score; students from large families get a lower score; school experience indicators, like preschool attendance and day school shift, are positively associated with ability; finally, region, is positively but insignificantly related with ability. Columns 2 and 4 of Table 4 present the unstandardized regression coefficients for the models explaining Spanish and Math respectively. Student's characteristics such as gender and age, had a significant impact on the two cognitive outputs. There is a positive correlation between being male and scoring higher achievement scores. This finding is consistent with available research in developed countries, systematically showing girls having lower scores (Maccoby, 1966, 1974). According to this evidence, boys and girls may be born alike in Math ability, but certain sex differences in performance emerge as early as such evidence can be gathered and remain through adulthood. One of the striking differences is that girls learn to hate Math sooner than boys for different, mostly cultural, reasons. One model (Eder and Parker, 1987) explains girls lower achievement in math in terms of less competitiveness demanded from females in those areas designated by society as outside their sex-role domain. Under this assumption, girls feel less pressured to compete in math, and thus have an excuse to avoid the painful struggle with this cognitive subject area. Boys on the other hand, are expected by society to be creative and independent. Thus males, in general, are more able in taking a math problem apart; they are more persistent and better able to tolerate ambiguity, and pay more careful attention to mathematical detail. The above cultural fact reproduces gender differences, limiting in many cases the content and quality of women's education. Cultural norms, inside and outside of the school, often promote sex segregation. However, there is evidence that significant changes are taking place in developing countries in terms of gradually changing the cultural patterns regarding family values toward girls education that could improve girls cognitive achievement. Currently in Latin America, for example, female enrollment in primary and secondary now exceeds that of males. Recent studies (Kostakis, 1987 for the case of Greece; and Raudenbush et al, 1991, for the case of Thailand) show that, when the cultural barriers explained above ease, girls tend to outscore boys in many subjects, including Math. This is clearly not the case yet in Mexico. There is a negative correlation between age and cognitive achievement. As older children tend to be repeaters, the negative effect indicates that primary school repetition, multiple re-entry, or late entry may have a toll in learning, as has been found frequently (Haddad, 1979; Psacharopoulos, Rojas and Velez, 1992). Family background characteristics, like educational schooling of both parents, family size and two-parents households are correlated with student achievement. Parents' educational background, our only indicator of socioeconomic level, is positively correlated with their children's cognitive achievement. This result, well documented in the literature, is an indication that increased parental schooling may imply more educational stimulation in the household, in terms of availability and use of reading materials, higher level vocabulary, and more outside cultural stimulation. It may also be an indication of higher socio-economic status, - 10 - and thus, more availability of educational materials at home. Table 4. Explaining Cognitive Output Independent variable Ability Spanish Math Male .321* .586* .457* .442* .298* Age -.100* -.303* -.262* -.158* -.113* Father's education .104* .213* .170* .124* .077* Mother's education .136* .199* .144* .126* .066* Live with both parents .201** -.207 .288* .180 .089 Family size -.058* -.117* -.094* -.026** -.001 Day school shift .853* 1.330* .985* 1.323* .941* Preschool experience .076* .121* .090* .034 -.001 Set of developed States .441 .648* .469** .730* .533* Ability .405* .449* Constant 4.777 16.086 14.153 10.382 8.238 R .106 .145 .218 .087 .203 * Statistical significance at 1% level or better ** Statistical significance at 5% level or better Family size is inversely correlated with student achievement indicating, perhaps, that in crowded households there is less time and exposure to parental stimulation, and probably there is also less money for cultural and educational stimulating materials and events. Research shows that students who live in one-parent households are disadvantaged (Mulkey, Crain, and Harrington, 1992). Some have argued that the negative effects are explained by the low income of one-parent-absent families (Hetherington et al., 1983), while other argued that the reason is found in that children from incomplete families are harmed by psychological stress and by incomplete socialization (Keith and Finlay, 1988). Consistent with these findings, we attest in the Mexican case a positive correlation between students living in a two-parent household and cognitive achievement. The only two variables related with students educational experience (education inputs) that are available in this study are the primary school shift and preschool experience of the sampled students. The,public primary day school shift is positively correlated with student cognitive achievement. In the Mexican case this is generated by an existing policy that streams better primary students into the day shift, and also because some afternoon primary school shift - 11 - students may be engaged in working activities, mainly in the non-formal sector of the economy, specially at the latter years of the primary six grade cycle. Both factors are conducive to low achievement. There is a positive correlation between preschool background and cognitive achievement. With the accumulation of empirical evidence that has been produced in the last decade, the effects of preschool experience on performance has been clarified. Myers (1992), in a review of 14 studies reporting on academic performance in non-industrialized countries, find that 8 out of 14 studies indicated that children with early education experience performed better; 5 found a negligible difference, and in one positive effects were found in rural, but not in urban contexts. A recent study in Togo (Jarousse and Mingat, 1991) even found a negative effect of preschool on academic performance. Halpern (1986) also reports these mixed results. . The available preschool empirical evidence and research suggests that: (a) positive results on cognitive achievement due to early stimulation programs are stronger during the first two years of primary school than later on; (b) effects on reading and writing are better than effects on math; and (c) children from disadvantaged families can benefit, and some times, even more than their more privileged peers (Flip, et al., 1983, and Flip and Schiefelbein, 1982). In the Mexican case, we find evidence to support these two last assertions. The effect of preschool education is positive on achievement, particularly for Spanish. According to Table 5 there is a significant marginal increase in cognitive achievement among those primary graduate students with preschool education coming from lower socio- economic households where parental schooling is minimal. The lower the socio-economic status of the student, the higher the marginal benefit the student gets for each additional year of preschool education. For example, there is a 19% increase in Math achievement score due to preschool education for a student with an illiterate father, whereas there is only a 1% gain in Math if the student's father have at the most higher education level. Table 5. Marginal Increase in Achievement by Parent Schooling and Preschool Experience (in %) Parental Schooling Father Mother Spanish Math Spanish Math None 7.2 19.0 5.3 8.8 At most primary 9.6 4.6 9.0 6.1 At most secondary 8.7 8.0 8.0 9.7 At most higher 9.1 1.0 0.0 0.0 - 12 - Furthermore, evidence gathered from a 1977-1990 time series and cross-sectional data on preschool coverage and primary school completion in Mexico shows that, other things being equal, every 10 percentage point increase in preschool coverage translates to a nearly 4 percentage point increase in primary education completion (Prawda and Psacharopoulos, 1991). There is a positive correlation between the level of educational development of a particular State and the cognitive results of their primary school graduates. This can be an indication that in the States with good education indicators, schools are better managed ; primary school teachers dedicate more time on task during the school year; there are less teacher strikes and thus more school days per calendar year; schools are better equipped; teachers better trained; and school books and other teaching and learning materials arrive, by and large, on time (Prawda and Psacharopoulos, 1991). Finally, columns 3 and 5 of Table 4 present the extended model, incorporating reasoning ability to the original models, causing the R2 to substantially increase and lowering the effects of the original variables. The positive effect between ability and achievement is expected given the high correlation between them. The important result of these extended models, however, is to the extent that the Raven matrices reflect differences in the innate ability of students, our results does not discredit the implications of most original predictor; they remained with a significant effect. The exception are preschool experience and family size for the Math test. CONCLUSIONS The following recommendations are based on the analysis of the evidence gathered from Mexican public primary school graduates: a) Supply of formal and non-conventional modalities of early stimulation programs should be increased targeting 4 and 5-year-olds living in: (i) poor, mostly, rural areas; and (ii) in the neighborhood of a public primary school showing low achievement scores in Spanish and Math; b) Parents with few years of schooling whose children are low academic achievers should be provided with non-conventional programs and materials to kindle their interest in reading and stimulate their parental school counseling activities with their children; c) Implement proven policies7 and enhance the allocation of proven educational inputs8 7 Available research has found that children who repeat one or more grades in their early years of schooling are likely to drop-out from the educational system later. Furthermore, repeating a grade does not necessarily help learning. Thus a recommended policy consists in setting automatic promotion for the first two or three grades of the primary cycle, and institute the pass-fail evaluation procedure thereafter (Haddad, 1979). - 13 - to: (a) reduce repetition and drop-out rates in targeted public primary schools, specially in the first two grades, and in the afternoon shifts; and (b) reduce late primary school entry through early stimulation programs and dissemination campaigns; and d) Strengthen the sectoral managerial capacity in the developing Mexican States (Prawda and Psacharopoulos, 1991; Prawda, 1992). REFERENCES Boissiere, M., J.B. Knight, and R.H. Sabot (1985), "Earnings, Schooling, Ability, and Cognitive Skills," American Economic Review, 75, 1016-1030. Bryant, E.C., E. Glaser, and M.H. Hansen (1974), Associations Between Educational Outcomes and Background Variables: A Review of Selected Literature, A Monograph of the National Assessment of Educational Progress, Denver. CEPAL (Comisi6n Econ6mica para Amdrica Latina) (1991), Qua aprenden y quidnes aprenden en las escuelas de Uruguay, Montevideo, Uruguay. Coleman, J.S., E.Q. Campbell, C.J. Hobson, J.J. McPartland, A.M. Mood, F.D. Weinfeld and R.L. York, (1966) The Equality of Educational Opportunity, USGPO, Washington D.C. Coleman, J.S., T. Hoffer, and S. Kilgora (1982), High School Achievement: Public, Catholic and Private Schools Compared, New York, Basic Books. Eder, D., and S. Parker, (1987) "The Cultural Production and Reproduction of Gender: The Effect of Extracurricular Activities on Peer-Group Culture," Sociology of Education, 60, 200-213. Filp, J., S. Donoso, C. Cardemil, E. Dieguez, J. Torres, and E. Schiefelbein, (1983) "Relationship Between Preprimary and Grade One Primary Education in State Schools in Chile," in King and R. Myers (eds), Preventing School Failure: The Relationship Between Preschool and Primary Education, Ottawa, Interantional Development Research Center. Filp, J. and E. Schiefelbein, (1982) "Efecto de la educacidn ereescolar en el rendimiento de primer grado de primaria: el estudio UMBRAL en Argentina, Bolivia, Colombia y Chile" Revista Latinoamericana de Estudios Educativos, 12, 1: 9-41. * Textbooks, educational learning and teaching materials, reading resource centers, targeted in-service teaching training, etc (Lockheed and Verspoor, 1991). - 14- Fuller, B., (1987) "What Factors Raise Achievement in the Third World" Review of Educational Research 57, 3: 255-292. Guevara, G., (1991) "Mexico. Un pais de reprobados?" Nexos. Haddad, W., (1979) "Educational and Economic Effects of Promotion and Repetition. Practices" World Bank Staff Working Paper 319. Washington DC. Halpern, R. (1986) "Effects of Early Childhood Education on Primary School Progress in Latin America," Comparative Education Review, 30, 2: 193-215. Hauser, R.M., (1971) Socioeconomic Background and Educational Performance, American Sociological Association, Washington D.C. Hetherington, E., K. Camara, and D. Featherman, (1983) "Achievement and Intellectual Functioning of Children in One-Parent Household" in Achievement and Achievement Motives, edited by J. Spence, San Francisco: W. H. Freeman. Heyneman, S., and W. Loxley, (1988) "The Effects of Primary School Quality on Academic Achievement Across Twenty-nine High and Low Income Countries" American Journal of Sociology, 88, 1162-1194. Jarousse J.P., and A. Mingat, (1991) "L'ecole Primaire en Afrique, Functionnement, Qualite, Produits: Le Cas du Togo,* Processed, IREDU, Dijon. Keith, V., and B. Finlay, (1988) 'The Impact of Parental Divorce on Children's Educational Attainment, Marital Timing, and Likelihood of Divorce" Journal of Marriage and the Family, 50, 797-809. King and R. Myers (Eds.), (1983) Preventing School Failure: The Relationship Between Preschool and Primary Education. Ottawa: International Development Research Centre. Kostakis. A. (1987) "Differences Among School Outputs and Educational Production Functions" Sociology of Education, 60, 232-241. Lockheed, M., and A. Verspoor, (1991) Improving Primary Education in Developing Countries, Oxford, Oxford University Press. Macobe, E. E., (1966) Development of Sex Diferences, Stanford University Press. Macobe, E. E., and C. N. Jacklin (1974) Psychology of Sex Differences, Stanford University Press. -15- McKay, A., and H. McKay, (1983) "Primary School Progress After Preschool Experience: Troublesome Issues in the Conduct of Follow up Research and Findings from the Cali, Colombia Study," in King, K., and R. Myers (Eds.) Preventing School Failure: The Relationship Between Preschool and Primary Education, Ottawa, IDRC. Ministerio de Educaci6n Pdblica de Costa Rica (1988). Memoria de 1987, San Jos6, Costa Rica. Mulkey, L.M., R.L. Crain, and A.J.C. Harrington, (1992) "One-Parent Households and Achievement: Economic and Behavioral Explanations of a Small Effect," Sociology of Education, 65, 48-65. Myers, R.G. (1992) "Early Childhood Development Programs in Latin America: Toward Definition of an Investment Strategy," A View from LATHR, 32, The World Bank, Technical Department, Latin America and the Caribbean Region, Washington D.C., August. Pozner, P. (1983) "Relationship Between Preschool Education and First Grade in Argentina" in King, K., and R. Myers (Eds.) Preventing School Failure: The Relationship Between Preschool and Primary Education, Ottawa, IDRC. Prawda, J. and Psacharopoulos, G. (1992) "Educational Development and Costing in Mexico, 1977-1990. A Cross-State Time Series Analysis" forthcoming International Journal of Educational Development. Prawda, J. (1993) "Educational Decentralization in Latin America. Lessons Learned" forthcoming International Journal of Educational Development. Psacharopoulos, G., and E. Velez, (1992) "Schooling, Ability, and Earnings in Colombia, 1988," Economic Development and Cultural Change, 40, 629-643. Psacharopoulos, G., C. Rojas, and E. Velez, (1992) "Achievement Evaluation of Colombia's Escuela Nueva: Is Multigrade the Answer?," Paper presented at the Annual Conference of the Comparative and International Education Society, Annapolis, March. Raven, J.C., (1956) Guide to the Colored Progressive Matrices (sets A, Ab, B), London: Lewis. Raudenbush, S.W., S. Kidchanapanish, and S.J. Kang, (1991) "The Effects of Preprimary Access and Quality on Educational Achievement in Thailand," Comparative Education Review, 35, 255-273. Schiefelbein, E., (1991) In Search of the School of the XXI Century. Is the Colombian Escuela Nueva the Right Pathfinder?, Santiago, Unesco-Unicef. - 16- Schiefelbein, E., and Wolff, L., (1992) "Repetition and Inadequate Achievement in Latin America's Primary Schools: A Review of Magnitudes, Causes, Relationships and Strategies". A View from LA7HR, 31, Human Resource Division, Technical Department, Latin America and the Caribbean Region, The World Bank. Secretaria de Educaci6n Piblica -SEP, (1992), Evaluacidn integral de la educacidn prescolar, primaria y secundaria, Fase 11. Direcci6n General de Evaluaci6n e Incorporaci6n. M6xico. Sinha, U., (1968) "The Use of Raven's Progressive Matrices Test in India," Indian Educational Review, 3, 118-132. Tobias, S (1978) Overcoming Math Anxiety. Houghton Mifflin Company. UNESCO-OREALC, 1990) "Situaci6n educativa de Am6rica Latina y el Caribe, 1980-1987". Proyecto Principal de Educaci6n. Santiago, Chile. Wilson, K.L., and A. Portes (1975) "The Educational Attainment Process: Results from a National Sample", American Journal of Sociology 81, t43-363. World Bank, (1990) Basic Education in Mexico. Trends, Issues and Policy Recommendations, Report 8930-ME, Latin America and the Caribbean Regional Office, The World Bank, Washington, DC. World Bank, (1991) Chile. Primary Education Improvement Project. Staff Appraisal Report. Report No. 9769-CH. Latin America and the Caribbean Region, Country Operations Department IV, Washington, DC. - 17 - - Continued from back Page No. 19 'Investment in Science Research and Training: The Case of Brazil and Implications for Other Countries' by Laurence Wolff, with contributions from George Psacharopoulos, Aron Kuppermann, Charles Blitzer, Geoffrey Shepherd, Carlos Primo Braga and Aicyone Saliba, September 1991 No. 20 'Prenatal and Perinatal Health Care: A Diagnostic Instrument' by Francisco Mardones, September 1991 No. 21 'Maternal Anthropometry in Prenatal Care. A New Maternal Weight Gain ChartI by Pedro Rosso, September 1991 No. 22 'Povery and Inequality in Latin America and the Caribbean During the 70s and 80s: An Overview of the Evidence' by Dominique van de Walle, September 1991 No. 23 wSocial Indicators in Latin America and the Caribbean* A Compilation of Statistics from 1970 to the Present' by George Psacharopoulos and Bill Wood, October 1991 No. 24 'ICETEX - A Student Loan Success Story in Colombia' by Samuel Carlson, October 1991 No. 25 'Educational Development and Costing in Mcico, 1977-1990: A Cross-State Tune-Series Analysis' by Juan Prawda and George Psacharopoulos, November 1991 No. 26 'A Cost-Beneft Analysis of Educational Investment in Venezuela, 1989- by Ariel Fiszbein and George Psacharopoulos, November 1991 No. 27 'Educational Decentralization in Latin America: Lessons Learned' by Juan Prawda, March 1992 No. 28 'Education and the Labor Market in Uruguaya by George Psacharopoulos and Eduardo Velez, June 1992 No. 29 "Some Operational Guidelines and Issues for the Evaluation of investments in Primary Health Care in the LAC Region' by Francisco Mardones and Barton Burkhalter, June 1992 No. 30 'The Economics of Higher Education in Brazil" by Jean-Jacques Paul and Laurence Wolff, July 1992 No. 31 'Repetition and Inadequate Achievement in Latin America's Primary Schools: A Review of Magnitudes, Causes, Relationships and Strategies' by Ernesto Schiefelbein with Laurence Wolff, August 1992 No. 32 'Early Childhood Dewelopment Programs in Latin America: Toward Definition of an Investment Strategy' by Robert G. Myers, August 1992 No. 33 "Primary School Quality in Meico' by Juan Carlos Palafox, Juan Prawda and Eduardo Velez, November 1992 Views from LATHR No. 0 "The Magnitude of Poverty in Latin America in the 1980s" September, 1990 No. 1 "An Ounce of Prevention is Worth How Much Cure? Thinking about the Allocation of Health Care Spending" by Philip Musgrove, September 1990. No. 2 "Decentralization and Educational Bureaucracies" by Juan Prawda, November, 1990 No. 3 "What Should Social Funds Finance?: Portfolio Mix, Targeting, and Efficiency Criteria" by Margaret E. Grosh, December 1990 No. 4 "Financial Balance in Chile: The ISAPRES (instituciones de Salad Previsional) Health Care System and the Public Sector" by Philip Musgrove, January, 1991 No. 5 "Population, Health and Nutrition issues in the Latin American and Caribbean Region and the Agenda for the 90's" by Oscar Echeverri, January, 1991 No. 6 wPopulation and Family Planning in the 1990's: Reconciling Macro and Micro Issues' by Bruce D. Carlson, February, 1991 No. 7 "The Feasibility of Student Loanm in Latin America: A Simulation" by Samuel Carlson and Guozhong Xie, March, 1991 No. 8 "Transforming the Vicious Circle - The Costs and Savings of School Inefficiency in Mexico" by Samuel Carlson, April 1991 No. 9 "Colombia's "Escuela Nueva': An Education Innovation" by Eduardo Velez, May 1991 No. 10 "Health Technology Development and Assessment: Do LAC Countries Have a Choice?" by Oscar Echeverri, June 1991 No. 11 "The Recurrent Cost Factor in the PHR Sector" by Jacob van Lutsenburg Maas, July 1991 No. 12 "The Burden of Death at Dierent Ages: Assumptions, Parameters and Values" by Phillip Musgrove, August 1991 No. 13 "Government Expenditure on Social Secrors in Latin America and the Caribbean: Statistical Trends" by Hongyu Yang, August 1991 No. 14 "From Manpower Planning to Labor Market Analysis" by George Psacharopoulos, September 1991 No. 15 "An Update on Cholera in the Americas" by Francisco Mardones, August 1991 No. 16 "What do we think about Health Care Finance in Latin America and the Caribbean?" by Philip Musgrove, September 1991 No. 17 "Population Growth, Exernalities and Poverty" by Nancy Birdsall and Charles Griffin, September 1991 No. 18 "Wage Trends in Latin America" by Alejandra Cox Edwards, September 1991 - Continued on inside Page