One of the longitudinal studies, the Perry Preschool Study, also measured social adjustment of program and control children on non-school variables. Program children were found to have less self-reported delinquent behavior, fewer contacts with the juvenile justice system, and fewer arrests, At age 19, 45 percent of the program group were supporting themselves by their own (or spouses') earnings, versus 25 percent of control group, Mean annualized welfare payments to the program group members was $633, for control group members $1509, Pregnancy rates for women were such that, if there had been 100 women in each group, program group women would have had 64 pregnancies and control group women 117 pregnancies during their teenage years (Schweinhart, et al, 1984).

The Perry Study was also unique in that it has conducted a cost-benefit analysis of investment in early childhood intervention. The analysis included costs of the program (in 1981 dollars, discounted at 3 percent annually) and costs avoided for reduced repetition, special education, welfare and crime through age 19; benefits calculated included actual earnings up to age 19, and projected lifetime earnings based on educational attainment, among other factors. The analysis found that benefits (to society and the individuals) exceeded costs by a factor of 7 to 1 for 1 year of early childhood education, 3.5 to 1 for 2 years.

A number of short-term evaluations of "typical" Head Start programs throughout the United States have followed children into primary school. Harrell (1983) conducted a meta-analysis of 71 reports of research on such programs, selected from a much larger number, based on adequacy of documentation, Using mete-analytic techniques, she examined effects on IQ, on developed. abilities at point of entry into school (school readiness) and on achievement at the end of the early grades. She found evidence of positive Head Start effects in all 3 domains, with the largest effects on school readiness measures (such as the Caldwell Preschool Inventory), the smallest on achievement at the end of grades 1, 2 or 3.

Harrell concludes that Head Start "does, indeed, enhance the cognitive development of children" (p. 4). But her own data undermine this conclusion. Harrell reports that the majority of studies employed pre-post designs with no control group, and that effect sizes found for these designs were 7 times greater than those found in treatment-control comparisons (.61 vs. .09). Harrell herself notes (p. 60): "Although controls for maturation were included in calculating the effect sizes [in studies with no control group] the magnitude of this gain relative to the others suggests either that the children gained extremely rapidly or, more plausibly, that the controls used were insufficient."

Evaluation of Title I

Complementing the investment in preschool compensatory education in the United States has been an investment in compensatory education in the primary grades under Title I of the Elementary and Secondary Education Act of 1965 (now called Chapter I), Under Title I children in grades K-6 who test below certain criterion in school readiness and later achievement receive special instruction, generally in math and reading, principally through "pullout" sessions with small instructional groups. Poor and non-poor children are eligible, although monies are targeted to schools with high percentages of poor children. The cost of Title I per year per participating student is about $500. (Head Start, a much more comprehensive program, costs an average $2700 per student annually.)

The major evaluation of Title I was the so called "Sustaining Effects" study (Carter, 1984), actually a compilation of 5 related studies, The heart of this evaluation was a longitudinal study of achievement for a sample of about 3000 Title I students and about 2500 regular students (some eligible for Title I, others ineligible). Title I participants demonstrated a significantly greater rate of gain than eligible, non-participating peers in math skills throughout the primary school years, and in reading skills until the end of grade 3, Participants' rate of gain was at least equal to that of "regular" (non-needy) students, although their absolute achievement scores remained much lower. Relatively lower-achieving Title I participants benefited less than relatively higher-achieving ones. Carter reports (1984, p, 7):”By the time students reached junior high school there was no evidence of sustained or delayed effects.”

Infant Interventions

We will touch on one further group of US studies, those examining effects of investment in prenatal and infant intervention programs. The majority of these programs work with parents in the home, and they tend to have a relatively greater focus on parent education in health, nutrition and home stimulation than do preschool programs. Evidence from reviews of these programs (Anderson, Fox, 6 Lewin, 1983; Halpern, 1984; White 6 Caste, 1984) suggests that they do have beneficial short-term effects on parent childrearing behaviors and coping skills, and somewhat less consistently on pregnancy outcomes, infant health and development, Only a few prenatal and infant intervention programs have followed children into their school years (e.g., Epstein & Weikart, 1979). The little available evidence suggests that intervention prenatally and in infancy, without continued intervention in the later early childhood years, does not affect school progress and achievement.

Summary and Generalizability of the US Evidence

Summary, The US evidence suggests that in the US. context small- scale, experimental early childhood intervention programs for poor children provide a variety of short- and long-term benefits to participants and to society, When monetarized these benefits exceed costs by a factor of at least 3 to 1. It is uncertain whether more typical early childhood service programs, notably Head Start, have the same magnitude and kinds of effects as experimental programs. Evidence from the Harrell review and the two longitudinal evaluations of service programs (Rome, Georgia Head Start; New York State Pre-kindergarten) suggests that the kinds of effects obtained are the same, but the magnitude is smaller.

With respect to an investment strategy focused on improving the quality of primary school instruction for poor children, in the US context of a generally well-developed primary education system, the marginal improvements from such extra investment appear to be modest, It does appear, though, that within the K-6 framework, extra investment “is more effective in the lower grades than in the higher grades" (Carter, 1984, p. 7).

Generalizability, A variety of factors limit the generalizability of the US evidence to the developing countries. The debilitating health and developmental effects of poverty on young children in the developing countries are more powerful and pervasive than the effects of poverty on young children in the US. (This is not to minimize the harm that growing up in poverty does to American children,) Poor children and their families in the developing countries do not have access to the variety of medical, social and income support programs that buffer the effects of poverty for families in the US, In other words, the preventive and compensatory work that an early childhood intervention program has to undertake in the developing countries is greater than that of a similar program in the US

The conditions of schooling in the developing countries-- large classes, few instructional resources, often poorly trained teachers, an inadequate number of "pieces" in each grade-- are such that the newly acquired skills that preschool participants bring to primary school may be less influential than in the US in shaping the course of children's school careers. When promotion policies are only loosely tied to children's abilities, when there is no special education to be "avoided", and when there are resources for only 10 or 20 percent of primary school participants to complete secondary school, positive long-term effects on the course of children's school careers found in the US may not be replicated in developing countries,

Nonetheless, it is reasonable to expect that developing country primary school systems are at least modestly sensitive to the kinds of skills and characteristics children bring with them; and conversely that children are differentially adaptive (due both to physical health and psychological make-up) to the demands of the school setting. In other words, the causal mechanisms at work in the US context may work in similar ways in the developing countries, albeit much more weakly, end with many more pre-conditions attached. Among these preconditions would be relatively greater attention to current health and nutritional needs of children; some minimal level of early intervention program quality (to be discussed more fully later); likelihood of enrolling in primary school soon after early childhood program participation; availability of a complete primary school, and at least minimally adequate learning conditions.

Evidence from the Developing Countries

In this section we review evidence from the developing countries examining whether participation in early childhood intervention programs-- principally those with an educational focus-- can increase children's chances of enrolling and staying in school, and perhaps school achievement, We will discuss studies in 3 groups. The first is a group of nutrition- related interventions with relatively sophisticated designs and less assailable internal validity. The second group, largely preschool interventions, are weaker methodologically, but more generalizable as models. Within the second group are a variety of program types, ranging from small-scale demonstrations to municipal of nation-wide service program s, A third group of studies are included that we call “program notes", These represent partial reports from what appear to be promising program s, that we were not able to verify or learn more about. As noted earlier, most of the studies to be reviewed took place in Latin America.

Nutrition-Related Interventions

The three intervention studies to be reviewed here were born “from academic concerns regarding the effects of PEM (protein energy malnutrition) on behavioral development" (Pollitt, Halpern, & Eskanasy, 1978, P. 52). Two of the three--the Cali Study (McKay & McKay, 1983) and the Bogota Study (Ortiz, 1983; Herrera, et al, 1980; Super, et al, 1983) included health surveillance, nutritional supplementation and an educational intervention. The third, the INCAP Guatemala Study (Klein, 1979; Balderston, 1981), involved only the health surveillance end nutritional supplementation,

1. Cali, Colombia

In the Cali Study 333 children were selected from a population of 449 potential subjects in the city's lowest income area, on the basis of (1) lowest height and weight for age; (2) highest number of clinical signs of malnutrition, (3) lowest per capita family income. These children received either 1 year of treatment beginning age 6, 2 years beginning age 5, 3 years beginning age 4, or 4 years beginning at age 3. Treatment consisted of 6 hours a day of center-based preschool education, with nutritional supplementation, health surveillance, and nutrition/health education integrated in to the daily program. Two control groups were established, one low income, normal weight and height (N = 116), one high income (N = 38).

All four treatment groups in this study demonstrated significantly greater growth than low income controls in general cognitive ability during and immediately after the treatment periods. Magnitude of gains was related to amount (and age of onset) of treatment. Modest, albeit diminished IQ, effects were found to persist at least up to age 8.

The Cali investigators managed to follow the majority of treatment children and a minority of low income controls up to age 10. Although attrition was a problem in all treatment groups, particularly the low-income control group, the Cali investigators report that it did not change the original characteristics of each group. The McKays report that children followed up (N = 274) have attended a total of 93 schools, many switching schools more than once, Part of the treatment group had a special primary program designed for it. Many children, particularly low SES controls, have attended private "backyard" and “bench" schools, being unable to enroll or stay in the public system. In these settings children are likelier than in public settings to be "promoted," even though they may have the least academic ability. In other words, the demands of formal schooling in this study were not a constant, with which children with different entering characteristics interacted.

With all these provisos, the McKays report that treatment group children were slightly more likely than low income controls to be promoted through the first three grades. Tables 5 and 6 summarize school progress for the different groups through age 10. The Cali investigators also found a significant relationship between IQ at age 8 and grade level at age 10, independent of treatment condition. In general, the magnitude of school progress effects was very modest, in relation to the magnitude of the investment in the early intervention treatment.

2, Bogota, Colombia

In the Bogota Study 443 families from the Bogota barrios, with women in the first or second trimester of pregnancy, and previous malnourished children under age 5, were assigned randomly to 1 of 6 treatment conditions: (1) no treatment, (2) supplementary nutrition from the 6th month of pregnancy to age 6 months, (3) supplementary nutrition from the 6th month of pregnancy to age 3 years; (4) supplementary nutrition from age

Tables 5 and 6

Evidence of School Progress Effects

from the Cali Study

Table 5. School failure each academic year from 1974-77 by study group.
________________________________________________________________________

                             Proportion of children failing the year
                            ____________________________________________


Group*                       1974-75          1975-76          1976-77
___________________________________________________________________

  T4                          0.18             0.24              0.27

  T3                          0.39             0.26              0.23

  T2                          0.32             0.27              0.30

  T1                          0.26             0.37              0.24

  T0                          0.32             0.41              0.35
_______________________________________________________________________

Table 6. Average grade level at the beginning of each school year for
        study group children with continuous attendance, 1974-77.
________________________________________________________________________
                     1974-       1975-      1976-       1977-

Group                  75          76          77          78          No.
________________________________________________________________________
  T4                   1.0         1.8         2.5         3.2         45

  T3                   1.0         1.8         2.4         3.2         41

  T2                   1.0         1.7         2.5         3.1         37

  T1                   1.0         1.7         2.3         3.0         76

  TO                   1.0         1.7         2.3         2.9         37

* TO: low income control   T1: 1 yr. treatment   T2: 2 yrs. treatment
  T3: 3 yrs. treatment     T4: 4 yrs. treatment

  Source: McKay & McKay, 1983

6 months to age 3; (5) early stimulation (through maternal tutoring) from birth to age 3, and (6) supplementary nutrition from the 6th month of pregnancy to age 3, plus early stimulation from birth to age 3. All families received medical care. The food supplement was provided in store-like neighborhood feeding stations, in amounts that would meet all family members' daily needs, and assure targeting where aimed. (Substitution and re-distribution nonetheless did occur,) The early stimulation/maternal tutoring consisted of twice-weekly home visits by trained para-professionals, who focused on modeling of parent-infant interaction for the mother, suggested play activities, and engaged in direct stimulation of the infant or child. Attrition left a sample at age 3 of 339 children.

Supplementation and maternal tutoring in different combinations were associated with improved cognitive abilities in various domains at ages 18 months and 3 years, with the strongest effect found for the full supplemented and home visited group. prenatal supplementation had a very modest (60 gram) effect on birth weight. Supplementation was also associated with improved physical growth at 3 years, The home visiting program had significant positive effects on the quality of mother-infant interaction (e.g., verbal interaction, contingent responsiveness, affect) in both supplemented and unsupplemented groups. Behavioral effects were greatest for supplemented infants, who theoretically had more energy with which to respond to their mothers' new skills.

The Bogota investigators (Herrera L Super, 1983) report findings for 269 children in an age 7 follow-up that included anthropometry, administration of a school readiness test (reading readiness, math, basic knowledge), and data collection on school enrollment, progress, and performance. Attrition from age 3 to age 7 was not random: families in which there were higher levels of paternal education were systematically more likely to have been lost, The follow-up apparently focused on treatment groups (1), (3), (5), and (6).

The school readiness test was administered to 174 children aged 5,6 to 8.7 (scores were adjusted for age). The investigators found a small, but significant overall positive effect of nutritional supplementation on readiness test scores (with or without maternal tutoring), Effects were larger at lower levels of father's education. There were no independent effects of maternal tutoring on test scores.

There was an important effect of maternal tutoring on age of initial primary enrollment. Mean enrollment age was: 5 years for the maternal tutoring group; 5,6 years for the maternal tutoring/nutritional supplementation group; 5.9 years for the supplementation group; 6 years for the control group. There were also significant positive effects of supplementation and tutoring, alone and combined, on 1st grade repetition. Children in all 3 intervention groups repeated at about a 4 percent rate, in the control group at a 13 percent rate. There were no significant overall intervention effects on teacher-assigned grades in 1st grade.

Significant positive effects at age 7 were found for supplementation and maternal tutoring, alone and combined, on height-for-age and weight-for-age. The positive effect of maternal tutoring alone on growth at age 7 was a new finding; there had been no effects at age 3. The maternal tutoring effects on growth were greatest at the lower levels of father's education, Maternal tutoring alone and with supplementation was also associated with higher levels of protein and to a lesser extent calorie intake at age 7.

Reviewing the whole pattern of findings at age 7, the investigators speculate that nutritional supplementation has more general long-term effects on children-for example on level of activity, alertness, social cooperation; whereas maternal tutoring has more selective effects--perhaps (through effects on maternal-child interaction) on children's “familiarity with a school-like learning paradigm of interaction with adults." They reiterate that the largest effects in most domains appeared to be found among the most disadvantaged children.

3. Guatemala

While the Cali and Bogota interventions were a mixture of education, nutritional supplementation, and health surveillance, the Guatemala INCAP Study was exclusively a nutritional supplementation intervention. It is included here because it has managed to follow children into their school careers and because it represents another legitimate early intervention option. In the INCAP Study (Klein, et al, 1972; Klein, 1979) nutritional supplementation was begun during the first trimester of pregnancy, and continued for index children up to age 7, Two rural villages in Guatemala were assigned a high-protein, high calorie supplement; two villages a no-protein, modest calorie supplement. The supplement was administered daily at the feeding stations (in Bogota families took it home), and amounts ingested were recorded. Total sample size was originally 671 children; about 450 children were followed longitudinally.

The INCAP team report that high supplement intake had significant effects on birth weight; on children's physical growth (height and weight) up to age 7; and on children's cognitive development up to age 3. (There is some debate about whether the calories or protein were more important.) Cognitive effects appeared to decrease in magnitude and generality beyond age 3, High supplement intake had no significant effect on verbal development at ages 5, 6, or 7, on early school progress, or on school performance. In 3 of the 4 participating villages in which parental education levels were moderately higher amount of supplement intake was predictive of likelihood of school enrollment. (Klein, 1979; Balderston, 1981,) The apparent mechanism at work here was parents with a more positive perception of the utility of schooling, selecting their bigger, healthier children for enrollment,

Preschool Intervention Studies

The next group of studies to be summarized were non-experimental evaluations of naturally occurring service/demonstration programs, In these studies children were not assigned to treatment conditions. Rather, the investigators identified groups of children already participating in an early intervention program; attempted to match them to peers who were similar but for participation; and prospectively or retrospectively followed both groups into the early primary school years. The findings of these studies should be viewed cautiously. Frequently, participating children and comparison group peers seem to be drawn from different populations, and thus may be initially different in important characteristics. Measures used are often not identified; when identified, psychometric properties and administration procedures are often not described, Children are often dispersed non-randomly to many primary schools: thus differences in promotion criteria may bias findings. Finally; the studies tend to focus only on initial adjustment to primary school.

In spite of the design limitations this group of studies sheds light an many of the factors that influence children's early school careers in developing countries. As with the nutrition-related interventions, most of the studies to be reported here took place in Latin America, We also report results from Morocco and Althone, South Africa,

1. A Four-Country Study: Argentina, Bolivia, Chile and Colombia
Filp and Scheifelbein (1982) report on a study coordinated across Argentina, Bolivia, Chile, and Colombia, in which children who were enrolled in preschools and kindergartens were compared with 1St grade peers who had not participated. Outcomes examined included age of enrollment, readiness at start of grade 1, promotion, and reading/writing ability at end of first grade, The investigators also examined whether the “socio economic status" (SES level) of preschools and primary schools--based on students' family backgrounds--was associated with particular kinds of teaching/learning experiences; whether teaching/learning approaches were generally different in preschool (or kindergarten) and 1st grade classrooms; and what prior characteristics of children generally affected the outcomes being examined.

To construct their samples, the investigators in each country started with preschool and kindergarten classes that operated either half- or full- day; 40 classes from Argentina, 52 from Bolivia, 62 from Colombia, and 40 from Chile. This gave them an original sample of 2595 children ages 4-7, about 40 percent of whom were poor. Between the end of the preschool/kindergarten year and enrollment in 1st grade the investigators lost half of their original sample of children. In 1st grade they identified an additional sample of children who had participated in preschool, and a sample of non-participant peers, drawn from the same and other 1st grade classes.

Given this general design and the enormous attrition from preschool to grade 1, it is probable that the treatment and control samples of children differed significantly on prior characteristics, Indeed, Filp and Scheifelbein note that participants in Chile and Argentina were at slightly higher SES levels within general SES groups than non-participants. Nonetheless, this was a relatively rich and complicated study, with a number of interesting findings.

Preschool observations revealed that the socio-economic make-up of the population served in each country affected the nature of children's experiences somewhat differently. In Argentina, preschools serving wealthier children had smaller class sizes, and more active, child-centered learning, In Chile there were no class size differences, but as in Argentina, teachers assumed a more didactic role in preschools serving poor children, In Colombia, preschools serving the poor had less well-prepared teachers, In Bolivia there were few differences by children's SES. Preschool in Chile and Argentina tended to reflect rehersal for first grade. In Bolivia especially, but also in Colombia, children's health and nutritional needs, being more pressing, received relatively more attention. In general, the higher the per capita income of the country, the greater the differentiation in preschool services by students' SES.

In all 4 countries, preschool/kindergarten was more child-centered, and children received more individualized attention, than in first grade, There was greater similarity in teaching/learning processes in first grade across all 7 countries, than in preschool, Filp and Scheifelbein also report that even when located in the same building preschool/kindergarten and 1st grade classes were very different in atmosphere.

Participation in preschool/kindergarten was found to have significant positive effects on age of 1st grade enrollment across social class (but especially for lower SES children) and urban/rural populations in Argentina and Chile; negligible effects were found in Colombia and Bolivia. Participation likewise bad significant positive effects an school readiness at entry into first grade in Argentina and Chile, but not in Colombia and Bolivia. (Preschool participants in Colombia and Bolivia appeared to be significantly more likely than those in Argentina and Chile to have to wait 1-3 years to enroll in 1st grade after participation in a preschool program.) Within the lower SES groups of children, other predictors (than preschool) of school readiness included number of siblings, birth order, housing, and nutritional status.

Preschool/kindergarten participation had a significant positive effect on promotion in Argentina: among low SES urban children 12 percent with preschool versus 27 percent without were retained in 1st grade; among low SES rural children 36 percent with versus 77 percent without were retained. There was also a significant positive effect on dropout rates. Findings from Colombia were also positive: among low SES children in 3 cities an average 10 percent with preschool versus 27 percent without preschool were retained in 1st grade (Chile had a policy of automatic promotion at the time of the study. There were no Bolivian data,)

Preschool effects on reading and writing ability at end of 1st grade differed from country to country, The Argentine investigators found significant positive effects of preschool participation across social class levels. Positive effects were also found in Bolivia, except for marginal urban neighborhoods. In Colombia and Chile preschool effects on developed abilities were negligible.

In summary, participation in preschool, while it did not eliminate or even narrow social class differences, did appear to affect positively age of enrollment and promotion, within social class, with benefits being greatest for low SES rural children. There were significant effects on school readiness at beginning of 1st grade, although these were more selective. There was also selective evidence of effects on developed abilities. Important covariates of all four dependent variables were family size, parental education, and physical health of the children. Preschool effects were attenuated, not surprisingly, when children had to wait 1 or more years to enroll in 1st grade,

2. Fortaleza, Brazil
Feijo (1983) reports on a study in Fortaleza, Brazil, similar to but better designed than the 4 nation study discussed above, She followed until the end of 1st grade 57 low income children who attended part-time public kindergarten (at ages 6 and 7) for a year and 70 who tried to enroll in the same kindergartens but could not, due to lack of space. She was able retrospectively to match treatment children with control peers on sex, age, birth order, and number of siblings. Pre- and post-tests (language, reading, math) were administered to all children during the kindergarten year and in 1st grade. There was almost no attrition, but upon entry into 1st grade the children were scattered among 8 schools and 26 classes in different parts of the city. Fourteen control children were converted into a second treatment group, receiving a short preparation course before 1st grade.

No significant group differences were found on pre- or post-test scores in kindergarten or 1st grade for most measures. One test, an "Interactive Beading Assessment" yielded modest differences in favor of the kindergarten group. Significant findings were found for promotion to 2nd grade: in the kindergarten group 64 percent were promoted; in the short-training group 50 percent were promoted; in the control group 34 percent were promoted. Females seemed to benefit most from the kindergarten experience,

3. Peru: The Puno Project and Its Extension
Myers and colleagues (1985) report on a multi-faceted evaluation of a large-scale, government-sponsored preschool education program in Peru. In the program, 3 to 5 year old children are brought together for 3 hours, 4 or 5 mornings a week in centers called PRONOEI (Programas No-Formal de Education Inicial). They receive education and care from a minimally trained community volunteer as well as a snack and/or noon-time meal, Mothers prepare a daily snack and/or noontime meal. The volunteer, who is selected by the community and receives a modest stipend, is supervised by Ministry of Education staff.

The PRONOEI program, which originated in the early 1970s, was seen by the Peruvian government as a realistic means of reaching Peru's poor children with preschool services, and as a catalyst for broader community participation. PRONOEI coverage grew steadily from 1975 to 1985, with coverage currently approaching 200,000 children in 6000 small centers, served by 8000 "promotores," and 800 "coordinadores” Current coverage represents about 10 to 15 percent of the 3 to 5 year old population.

In 1984, an evaluation was carried out of approximately 2000 PRONOEI receiving assistance from USAID and UNICEF (as well as from the Peruvian government). The evaluation found a modest, but significant impact on the cognitive, social and motor development of participating children in some, though not all, of the communities studied (see below). Moderate and indirect effects were found on the nutritional status of children, There were also some community-level effects; specifically, the program served as a catalyst for creation of new community groups-- mothers' clubs, parent associations, and nutrition clubs; and also affected the topics of discussion in community meetings. However, no significant effects of PRONOEI participation were found on primary school repetition rates.

The evaluation design included: interviews with parents, volunteers and community leaders in 148 PRONOEI communities; detailed observations of daily activity in 12 PRONOEI; tests administered to 334 children in 26 matched PRONOEI and non-PRONOEI villages in 3 states, including a sample of formal preschool participants; analysis of children's PRONOEI and lst grade school records; height and weight measurements; and parent interviews for tested children. Communities were selected using a stratified random sampling procedure; stratification was based on accessibility, socio-economic characteristics, and program duration, The test used was criterion-referenced, and was specially designed to reflect PRONOEI curriculum objectives, as well as general developed abilities in the psychomotor, intellectual and social domains.

Children attending PRONOEI were compared with children of the same ages from non-PRONOEI villages in each of three Peruvian states-- Puno, Cuzco, and San Martin. An attempt to match children in terms of their social backgrounds-was successful for the Puno children; but in the latter two states parents of non-PRONOEI children were more apt to have completed primary school. Results of the PRONOEI/non-PRONOEI comparison were most promising for Puno, where the two groups were matched, and where the program is the most mature (11 years in operation) and most widespread (in over 1000 rural communities). In Puno, in both Aymara- and Quechua- speaking areas, PRONOEI children performed significantly better than non-PRONOEI children on the intellectual, social, and motor sub-scales of the test administered. These differences were found in spite of an evaluation finding that the quality of the PRONOEI program was generally low, as judged by the teaching skills of the promoters, the time devoted to educational activities, the availability and use of materials, and the quality of supervision of the promoters by Ministry of Education staff,

No effects of PRONOEI participation were found on promotion from 1st to 2nd grade, or from 2nd to 3rd grade. Effects favoring PRONOEI participants were found on age of enrollment. Repetition of 1st grade was extraordinarily high--over 50 percent--in the PRONOEI and the control villages in the sample, Moreover, 25 percent of children who repeated did so more than once. In some communities, repetition rates for 1st grade approached 100 percent. The evaluation team was not able to examine the reasons for these extraordinary repetition rates, These were largely rural regions, where repetition is usually highest, As discussed earlier, repetition rates like these are often structural--inherent in the local school situation, minimizing the role of individual characteristics or abilities in determining children's promotion.

The above findings are largely, but not wholly, consistent with results from two earlier studies in Puno (Aliaga, 1981, 1983). The first of the studies examined grades assigned by teachers to PRONOEI and non-PRONOEI children, and found no differences. The second involved administration of a special test of mathematical and language ability to primary school children who bad participated in the PRONOEI program and a control group of peers. No differences were found between the two groups.

The PRONOEI experience illustrates the importance of a coordinated investment strategy linking expansion of preprimary education to qualitative improvement of primary education. It also illustrates the complex relationship between quality and coverage in the growth of early childhood intervention programs in the developing countries. Is it appropriate, for example, to promote a national early childhood service program, even one based on significant community contributions (i.e., the volunteer teachers), when the technical resource back-up for that program is not dependable? What quality and quantity of work is reasonable to expect of the community volunteer, in any case? Could a community-based program target particular children for service, and deny others? Among the more interesting recommendations of the evaluation team: establish an intermediate, paid role--the assistant teacher/coordinator--for volunteers to aspire to; use early morning radio spots to help the volunteers plan daily activities; use experienced volunteers to help adjust the curricular guides.

4. Morocco: Modern and Quranic Preschools

Wagner and colleagues (1984) are currently conducting a longitudinal study of literacy acquisition in Morocco, for which they have followed children with different early childhood experiences into their primary school careers, Their sample consists of 378 children from 5 groups: an rural group and an urban group who participated in Quranic preschool programs; an urban group who participated in modern preschool programs; and an urban group and a rural group with no preschool program experience. In constructing their original samples the investigators controlled for social class; they report that the modern preschool group comes from families more likely to be employed in the modern sector, (The "preschool" programs involved typically serve children 5 to 7 years old, in that sense being more like kindergarten. In Morocco, about 30 percent of children 5 to 7 are enrolled in Quranic preschools, and about 15 percent in modern kindergartens.

The investigators are administering a battery of achievement and general ability tests to children at the end of each grade, as well as documenting promotion patterns. They are currently analyzing data from the second year of primary school. Analyses of 1st grade achievement data indicated that rural Quranic preschool children performed significantly better than non-preschool peers on most achievement/ability test scales, In the urban setting group differences were less clear-cut; the modern preschool group performed moderately better than the other two on most tasks, but few differences were found between Quranic preschool and non-preschool groups. Second grade analyses, currently underway, suggest some attenuation of the findings in both rural and urban areas, though the investigators warn that these findings are tentative.

With respect to promotion rates, the investigators found no group differences among the five groups. This was somewhat surprising, given differences between the two rural groups and between rural and urban groups more generally in reading readiness and other test scores, The investigators suggest that the politics of schooling and a quota system for promotions may "obscure all but the most general differences in academic achievement and skill development" (Wagner & Spratt, l984, p, 9).

5. Althone, South Africa

Short (1985) reports on a Bernard van Leer Foundation supported initiative in Althone, South Africa, The program, called the Early Learning Center (ELC), served mainly disadvantaged "colored" children (classified as mixed race), Its main component was a daily 6 hour center- based preschool program for 3 to 6 year olds, that included health surveillance and nutrition supplementation, Children participated from 1 to 3 years. At the end of 1st grade ELC participants and untreated school peers (selection strategy and number sampled not reported) were administered the Pupil Behavior Inventory (Sarri and Radin). Positive effects favoring ELC children were found on inquisitiveness, social adjustment and motivation, but not on academic-task-related behavior. By age 15, 40 percent of a sample of 60 ELC participants had reached Standard 8 (grade 10) without repeating a grade, compared with a national norm for coloreds of 25 percent. By age 17, 32 percent had entered Standard 10 without repeating, compared with the colored norm of 10 percent, Girls benefited from the ELC program more than boys, and moderately deprived children more than severely deprived children.

6. Osorno. Chile

Walker (1983) and Richards (1985) report; on an evaluation of the Parents and Children Program (Programa Padres Hijos) in Osorno, a rural area of central Chile, In the program, which has been implemented in 52 rural communities, locally chosen volunteers, acting as facilitators, brought participating parents together on a weekly basis to discuss problems related to childrearing, The problems and concerns raised by parents were then used as the basis for preparation of a series of radio dramas. Ongoing weekly meetings were keyed to radio broadcasts designed to complement discussions. Local facilitators also were provided with worksheets for parents to use with their children. The idea underlying the program was to improve child development through improved community organization and personal development of parents, as well as through parent education for childrearing.

The evaluators report that in 1st grade, teachers (who did not know the children's history) were asked to rate the reading skills (not specified) of a sample of 29 children of participating mothers and 49 peers (selection procedures not described). Nineteen percent of the treatment children were rated as excellent readers versus II percent of controls; 52 percent were rated as good readers versus 35 percent of controls; 31 percent as fair readers versus 44 percent of controls; and none as poor readers versus 12 percent of controls. Treatment children also performed better than untreated peers on a Draw-A-Figure test,

Program Notes

Thus far, we have reported evidence from evaluations whose methods as well as results could be examined. In the course of our inquiries, however, we encountered brief reports of evaluation findings for which we were not able to acquire full reports that included program descriptions and research designs (at least in time for this review), We nevertheless present these unverified reports here. Taken together they suggest the same uneven, but promising, pattern of results reported thug far.

Note 1: Haiti

A follow-up study was undertaken of children who participated in a preschool project funded by USAID and implemented by CARE-Haiti. The preschools were set up as an element of Community Integrated Nutrition and Education Centers (CINEC), themselves created in communities where the World Bank was funding the construction of new primary schools, The CINEC project team reports that 86 percent of the preschool participants have completed the 4th year of primary school, compared with a “general" completion rate of 30 percent, An earlier progress report from the same project indicated that CINEC (not differentiated from preschool) children had a promotion rate of 87 percent, compared with a “general” rate of 37 percent. Moreover, "school directors and the first grade teachers unanimously agree that the CINEC children are more alert, more eager to learn, and better assimilate the different concepts of math, reading, and writing" (CARE, Haiti, 1982). Nonetheless, one reviewer of the earlier progress report suggested that CINEC and non-CINEC children attending the same primary school had about the same repetition rate, suggesting that primary school effects may be attributable to differences in primary schools as much as differences in children.

Note 2: Brazil

An audio-visual presentation from the UNICEF office in Brazil dealing with preschool education includes figures from a study carried out in Rio de Janeiro involving approximately 1000 children. The researchers in this study apparently found that 28 percent of children with preschool program experience failed first grade versus 60 percent without preschool experience. Again, we do not know how these two groups compare in background characteristics.

Note 3: Brazil

Since 1977, a portion of a World Bank nutrition loan to Brazil has been used to support early childhood education and supplementary feeding for 4 to 6 year olds, These activities have been carried out through PROAPE (Programa de Alimentacao do Pre-escolar), with the goal of enhancing school participation rates and school achievement, The following comes from a report on the lessons learned from this World Bank project: "A careful evaluation of the CEAPE program (the prototype for PROAPE) in the Municipio de Leme in Sao Paulo) suggested that....school performance scores were better and repetition rates were lower among CEAPE children" (p.13),

Further: "The percentage of former PROAPE participants who passed in the first and second years of elementary school was 73.5 percent compared with only 59.5 percent in the control group. In terms of academic performance, children with two years exposure to PROAPE performed consistently better than control groups."

Note 4: Malaysia

A study carried out in Malaysia by the Ministry of Education (with funds from UNICEF)to examine the effects of preschool attendance on primary school achievement “,..seems to have established that children attending kindergartens perform better than other children in the early stages of primary schooling. How much this is due to what they learn in the Kindergarten and how much can be attributed to the fact that they are a selected group is not clear at present" (Crellin, , p.2).

Note 5: Bangladesh

The following comes from a USAID project identification document discussing females' access to education: “The Meher Universal Primary Education Project in Bangladesh expanded educational opportunities by establishing small-scale feeder schools serving as preschools and early primary centers, employing local women teachers thus lessening the burden on the crowded first grade of the established primary school, increasing female participation rates and reducing wastage rates" (USAID, 1984, P. ).

Summary of Evidence from the Developing Country Studies
As we noted at the outset, this review is a first attempt at drawing defendable generalizations about the effectiveness of early childhood interventions in enhancing children’s school careers in the developing countries. We have pointed out many of the limitations of available evidence, and will not reiterate them here, except to note that these limitations imply a need for international donors to support a careful program of research and development in this general area. Limitations not withstanding, we are able to draw some "first order" conclusions.

The available evidence of the effects of early childhood intervention yields a picture of modest positive effects on initial adjustment to the demands of the primary school. The particular mechanisms enhancing this adjustment appear to reflect some combination of earlier age of enrollment, improved school readiness and more selectively, improved health and energy level (thus, presumably, attendance as well). Changes in parental knowledge and attitudes are hinted at in a few studies, but their contribution to children's initial adjustment is largely undocumented. Only the INCAP study reports positive effects on absolute likelihood of school enrollment; most studies had no data on this variable.

The most striking finding suggested by the present review is that structural features of the primary school systems (promotion quotas, teacher attitudes, low quality instruction, resource inadequacies, and so forth) seemed rapidly to overwhelm any early childhood intervention effects. Children's individual abilities and physical integrity still played a role in influencing early school progress in studies reviewed, but within a much narrower and absolutely lower range of variance in possible outcomes.

The evidence is not sufficiently broad or detailed at this point to draw conclusions regarding the kind of early childhood intervention most likely to benefit children (holding Quality constant) under the circumstances found in most developing countries. A daily educational program, with nutritional supplementation and health surveillance, were the most common elements of programs reviewed, Since individual studies addressed different questions, the findings across studies were not cumulative with respect to preferred combinations of inputs and procedures.

There is suggestive evidence in a few studies that relatively more deprived children benefited more cognitively and socially than their less deprived peers, Unfortunately, these same children were also more likely to attend the poorest quality primary schools, making these effects short-lived.

Policy Dilemmas:
Toward a Reasonable Program of Investment

The evidence discussed in this review reconfirms that for many children in the developing countries the transition to primary school is a difficult one, Both the personal resources that children bring to school and the conditions of primary schools themselves present obstacles to a successful adjustment, The personal and social costs of this poor adjustment include continued high primary school repitition and dropout rates, and consequently another generation of functional illiterates, an expanding base of minimally flexible and adaptive human resources, and a reproduction of the survival-oriented parenting that the current generation received, This likely scenario leads us to suggest that improving poor children's adjustment to primary school should be a central policy goal of developing country governments and international donors during the remainder of this decade at least.

Within such a framework, early childhood intervention programs have a number of modest, but crucial, roles to play, The first and most obvious is to improve poor children's skills and energy to cope with the cognitive and social demands of formal schooling as a developmental setting; as in the US case, to assure a more positive initial school experience, A positive initial school experience can boost a child's psychological capacity to cope with subsequent school demands, including those created by school system dysfunctionalities.

A second role for early childhood interventions is to provide pedagogical models for early primary practice; to extend the early childhood practitioner's view of children's needs, and consequently the role of the teacher implied by such needs, into the first two years of the primary school.

In a more ambitious vein, it is possible to imagine an integrated early childhood-early primary program that has as a central objective hooking children in to the formal educational process. Such a basic reorientation would require a research and development program-- funded by international donors in most cases-- that that began with an examination of what happens to children currently during the transitional years (5 to 8); what teachers do, why they do it, and the effects of their behavior on children.

An integrated early childhood-primary program could serve as a vehicle for linking family and community interests and strengths to the formal schooling system; for example, infusing the values and content of the local culture into the curriculum, at first at the preprimary then at the primary level; serving as a focus for community development energies that, once organized, could also be focused on other issues; involving teachers in the solution of community problems linked to child development but not obviously part of the formal school system's mandate; reaching parents and their children with educational services before the perceived costs of children's participation in a process with ill-defined returns becomes too high.

Beyond envisioning an integrated early childhood-primary program lie a number of thorny problems: Should the early childhood program fall under the jurisdiction of the primary school establishment (i.e. the Ministry of Education), running the risk of becoming simply a downward extension of current primary school practices? Should experienced child development specialists and early childhood educators be placed in positions of authority within the primary system's bureaucracy, and be asked to re-think and re-design an early primary program from the ground up? Should private early childhood programs be integrated into the formal educational system? Should two separate systems be maintained, with mechanisms such as a common curriculum and common teacher training courses being used to encourage continuity?

We believe that it would be premature to urge an integration of early childhood intervention programs into Ministries of Education; there are too many dangers in such an approach. But we feel strongly that the first two years of primary school present too radical a departure from prior experience for many children, Creating a transitional level of formal education that served children in the 4 through 7 age range would only work if an autonomous unit linked to but with some independence from the Ministry of Education could be established. Perhaps structures currently used in England-- for example the Infant Schools-- could serve as a model.

References

Aliaga, J. 1981. Informe final de la evaluacion del proyecto piloto experimental de Educacion Inicial No-Escolarizada. Lima UNICEF.

Aliaga, J. 1983. Programas comunitariaos pre-escolares de intervencion temprano y desarrollo infantil en contextos diferenciados rurales de Puno. Lima: UNICEF.

Anderson, M., Fox, H., & Lewin, L. 1983. Reducing infant death and disability: Effective interventions and state strategies for implementation. Washington, DC: Lewin & Associates,

Balderston, J., Wilson, A., Freire, M,, & Simonen, S. 1981, Malnourished children of the rural poor. Boston: Auburn House.

Beirn, R., Kinsey, D. , & McGinn, N, 1972. Antecedents and consequences of early school leaving. Cambridge: Harvard Graduate School of Education, Occasional Papers in Education and Development, No. 8.

Berg, A. 1981. Malnourished people: A policy view. Washington, DC: The World Bank.

Bernard van Leer Foundation. 1984, Moving to scale, Early childhood education project in Kenya. Prepared for the Second Inter-Agency Meeting on Community-Based Approaches to Child Development. The Hague, Netherlands,

Birdsall, N. 1980. Population and poverty in the developing world. Washington, DC. World Bank Staff Working Paper No. 404.

Brimer, M., & Pauli, L. 1971. Wastage in education: A world problem. Paris: UNESCO/IBE.

CARE/Haiti, November 1982, personal communication.

Carter, L. 1994. Sustaining effects of compensatory and elementary education. Educational Researcher. 13,7, 4-13.

Cicirelli, V., et al. 1969. The impact of Head Start: An evaluation of Head Start on children's cognitive and affective development. Athens, Ohio: Westinghouse Learning Corporation.

Clement, J., et al. l984. Changed lives: The effects of the Perry Preschool Program on youths through age 19. Ypsilanti High/Scope Educational Research Foundation, Monograph No. 8.

Colclough, C. 1980, Primary schooling and economic development: A review of the evidence. Washington, DC: World Bank Staff Working Paper No. 399.

Crellin, C. 1985. Staff mission report, Quantan, Malaysia, December 1984.

Epstein, A. & Weikart, D. 1979, The Ypsilanti-Carnegie infant education project. Ypsilanti: High/Scope Foundation Research Monograph.

Feijo, M. 1982. Early childhood education programs and children's subsequent learning: A Brazilian case. Unpublished Ph.D., Dissertation, Stanford University.

Filp, J., et al. 1983. Relationship between preprimary and grade one primary education in state schools in Chile. In K. King & R. Myers (Eds.), The relationship between preschool and primary education. Ottawa: International Development Research Centre.

Filp, J., & Schiefelbein, E, 1982. Effecto de la educacion preescolar en el rendemiento de primer grade de primaria: el estudio UMBRAL en Argentina, Bolivia, Colombia y Chile. Revista Latinoamericana de Est. Educ.. 12,1, 9-41.

Fujimoto, J., & Villanueva, Y. 1984. Going to scale: The Peruvian experience. A Paper prepared for the Second Inter-Agency Meeting on Community-Based Child Development. New York UNICEF,

Gordon, I., & Jester, R, 1980. Middle school performance as a function of early intervention. Final Report to the Child Welfare Research and Demonstration Grants Program, Administration for Children, Youth and Families.

Grant, J. 1985. The state of the world's children 1985. New York: UNICEF.

Grawe, R. 1979, Ability in preschoolers, earnings and home environment. Washington, DC: World Bank Staff Working Paper No, 322.

Gray, S., Ramsey, B., & Klaus, R. 1982. From 3 to 20 - The early training project. Baltimore: University Park Press.

Haddad, W. 1979, Educational and economic effects of promotion and repetition practices. Washington, DC: World Bank Staff Working Paper No, 319.

Haglund, E. 1982. The problem of the match--cognitive transition between early childhood and primary school: Nigeria. Journal of Developing Areas. 17, 77-92,

Halpern, R. 1980, Early childhood programs in Latin America. Harvard Educational Review. 50,4, 481-496.

Halpern, R. 1984. Lack of effects for home-based early intervention? Some possible explanations. American Journal of Orthopsychiatry. 54,1, 33-42.

Harrell, A. 1983. The effects of the Head Start program on children's cognitive development. Preliminary Report. Washington, DC: US Department of Health and Human Services, Administration for Children, Youth and Families,

Herrera, M., et al, 1980, Effects of nutritional supplementation and early education on physical and cognitive development, In R. Turner and F. Reese (Eds.), Life span developmental psychology: Intervention. New York: Academic Press,

Herrera, M., & Super, C, 1983. School performance and physical growth of underprivileged children: Results of the Bogota Project at seven years. Report to the World Bank. Cambridge: Harvard School of Public Health,

Heyneman, S, Jamison, D,, & Montenegro, X. 1984. Textbooks in the Philippines: Evaluation of the pedagogical impact of a nationwide investment. Educational Evaluation and Policy Analysis. 6,2, 139-150,

Heyneman, S., & Loxley, W. 1983. The impact of primary school quality on academic achievement across 29 high and low income countries. American Journal of Sociology. 88, 1162-1194.

International Review of Education, 1983. Special issue: The universalization of primary education. 29,2.

Irvine, D. 1982. Evaluation of the New York State experimental pre-kindergarten program. Paper presented at the annual meeting of the American Educational Research Association, New York City.

Irwin, M., Engle, P., Yarbrough, C., Klein, R., & Townsend, J. 1978. The relationship of prior ability and family characteristics to school attendance and school achievement in rural Guatemala. Child Development. 49, 415.

King, K., & Myers, R. 1983. Preventing school failure: The relationship between preschool and primary education. Ottawa: International Development Research Centre.

Klein, R., et al. 1972. Is big smart? The relation of growth to cognition. Journal of Health and Social Behavior. 13, 219-225,

Klein, R. 1979. Malnutrition and human behavior: A backward glance at an ongoing longitudinal study. In D. Levitsky (Ed.), Malnutrition, environment and behavior. Ithaca: Cornell University Press,

La Belie, T., & Verhine, R. 1975. Nonformal education and occupational stratification: implications for Latin America. Harvard Educational Review. 45, 2, 160-190,

Laosa, L. 1982. School, occupation, culture and family: The impact of parental schooling on the parent-child relationship, Journal of Educational Psychology. 74,6, 791-827.

Lazar, I., & Darlington, R., et al. 1982. Lasting effects of early education: A report from the Consortium for Longitudinal Studies, Monographs of the Society for Research in Child Development. No. 195.

Leslie, J, 1984. Design and implementation of nutrition projects. Lessons learned from the World Bank experience in Brazil, Indonesia, Colombia end India. Washington, D.C.: The World Bank.

Levenstein, P., O'Hara, J., $ Madden, J. 1983. The Mother-Child Home Program of the Verbal Interaction Project. In Consortium for Longitudinal Studies (Ed.), As the twig is bent...lasting effects of preschool programs. Hillsdale, NJ: Lawrence Earlbaum,

LeVine, R, 1980, Influences of women's schooling on maternal behavior in the third world. Comparative Education Review, 24,2, 78-105 (Pt. 2).

Lockheed, M., Jamison, D., & Lau, L. 1979. Farmer education and farm efficiency: A survey. Economic Development and Cultural Change..

Mauras, M., Latorre, C., L Filp, J. 1979. Alternatives de atencion al preescolar in America Latina y el Caribe. Santiago, Chile: UNICEF.

McKay, A., 1982 Longitudinal syudy of the long-term effects of the duration of early childhood intervention on cognitive ability and primary school performance. Unpublished Ph.D. Dissertation, Northwestern University, Evanston, Illinois.

McKay, A., & McKay, H. 1983. Primary school progress after preschool experience: Troublesome issues in the conduct of follow-up research and findings from the Cali, Colombia study. In K. King & R. Myers (Eds.), Preventing school failure: The relationship between preschool and primary education. Ottawa: international Development Research Centre.

Monroe, E., & McDonald, M. 1981. Follow-up study of the 1966 Head Start program. Rome City Schools, Rome, Georgia.-(Mimeo)

Moock, P., L Leslie, J. (Forthcoming). Childhood malnutrition and schooling in the Terai region of Nepal. Journal of Development Economics.

Myers, R,, et al, 1985, Preschool education as a catalyst for community development. Prepared for US Agency for International Development, Lima, Peru. (Mimeo)

Ortiz, N, 1983. Psychosocial stimulation and complementary nourishment during the first three years: Its repercussions on scholastic achievement. In K. King & R. Myers (Eds.), Preventing school failure: The relationship between preschool and primary education. Ottawa: International Development Research Centre,

Palmer, F. 1983. The Harlem study: Effects by type of training, age of training and social class, In Consortium far Longitudinal Studies (Eds.), As the twig is bent...lasting effects of preschool programming. Hillsdale: Lawrence Earlbaum.

Pollitt, E. 1984. Child Development Reference Document: (I) Risk factors in the mental development of young children in the developing countries. (II) Early childhood intervention programs for the young child in the developing world. Prepared for UNICEF. University of Texas Health Science Center. (Mimeo)

Pollitt, E. & Thompson, C. 1977. Protein-calorie malnutrition and behavior: A view from psychology. In Nutrition and the brain, vol 2., edited by R. Wortman & J, Wortman. New York: Raven/Press,

Pollitt, E., Halpern, R,, A Eskanasy, P, 1980. Poverty and malnutrition in Latin America: Early childhood intervention program. New York: Praeger.

Popkin, B,, & Lim-Ybanez, M, 1982. Nutrition and school achievement. Social Science and Medicine. 16, 53-61.

Pozner, P, 1983. Relationship between preschool education and first grade in Argentina. In K. King & R. Myers (Eds.), Preventing school failure: The relationship between preschool and primary education. Ottawa: International Development Research Centre.

Prakasha, V. 1984. Battling costs for quality and quantity: emerging responses to early childhood care and education. Paris: UNESCO.

Raizen, S., & Bobrow, S, 1974, Design for a national evaluation of social competence in Head Start children. Santa Monica: The Rand Corporation.

Rhode, J. 1983. Why the other half dies: The science and politics of child mortality in the third world. Assignment Children. 61-62, 35-67.

Richards, H. 1985. The evaluation of cultural action. London: McMillan.

Rogoff, B. 1980. Schooling and the development of cognitive skills, In H. Triandis & A. Heron (Eds.), Handbook of Cross-Cultural Psychology. vol, 4. Boston: Allyn-Bacon,

Scheifelbein, E, 6 Farrell, J. Causas de la desercion en la ensenanza media. Revista de Educacion. 68, 43.

Short, A. (Forthcoming). Seeking change: Education for the disadvantaged in South Africa. Ypsilanti: High/Scope Press.

Super, C., et al. 1983. Infant and caretaker behavior as mediators of nutritional and social intervention in the barrios of Bogota. In T. Field (Ed.), Culture and early interaction. Hillsdale, NJ: Lawrence Earlbaum.

Toro, J., & de Rosa, I. 1983. Papito, yo porque tengo que repitir el ano?. Toronto: Ontario Institute for Studies in Education, (mimeo)

Triandis, H. 1980, Reflections on trends in Cross-Cultural research. Journal of Cross-Cultural Psychology. 11,1, 35-58.

UNESCO. 1984. The drop-out problem in primary education: Some case studies. Bangkok. Regional Office for Education.

UNICEF/Brazil. 1984, Annual report, Brasilia.

UNICEF/India. 1984. ICDS. New Delhi.

United States Agency for International Development. 1982. A.I.D Policy Paper: Basic education and technical training. Washington, D.C., Bureau or Program and Policy Coordination.

United States Agency for International Development/Nepal. Project identification document. Girls’ access to education project.

Wagner, D., & Spratt, J. 1984. Modern versus Quranic preschools in Morocco: Who attends and who achieves?. Paper presented at the American Anthropological Association Meetings, Denver.

Walker, H. (Forthcoming). La evaluacion iluminativa del programa Padres e Hijos. Mimeo.

Walters, P. 1981. Educational change and national economic development. Harvard Educational Review. 51,1, 94-106.

Wax, M,, & Wax, R, 1971, Great tradition, little tradition, and formal education. In. M. Wax, S, Diamond, & F. Gearing (Eds.), Anthropological perspectives on education. New York: Basic Books.

White, K,, L Caste, G, 1984. An integrative review of early intervention efficacy studies with at-risk children: Implications for the handicapped. Logan, Utah: Early Intervention Research Institute, Utah State University.

Woodhead, M. 1985. Preschool education has long-term benefits-- but can they be generalized? Oxford Review of Education. forthcoming,

News    What's On    Policies    Guidance  Quality   Links    Info on Bristol    Parents Page

98 Plan    Key Issues for 99 Plan    Emerging 99 Plan    The Bristol Partnership