In a recent Joint Center for Poverty Research working paper, "Early Childhood Intervention Programs: What Do We Know?" Janet Currie reviews the evaluations of several early childhood intervention programs. The programs she examines are predominantly center-based programs that emphasize school readiness. She finds that well-designed, well-funded early interventions can have large and significant effects on school readiness and subsequent child outcomes. She also, however, finds a paucity of high-quality research on early intervention programs; only four of the many programs evaluated used random assignment.

Effects of Early Interventions

Currie focuses on four model programs that were evaluated using randomized trials-the Perry Preschool Project, the Carolina Abecedarian Project, the Early Training Project, and the Milwaukee Project. In contrast to these small, model programs, she also looks at results from Head Start, a large, publicly funded program.

The Early Training Project showed dramatic reductions in use of special education among youth at age 12, although there were no statistically significant differences between treatments and controls in achievement test scores, grade retention, or high school graduation. Although not significant, the outcomes were nevertheless in the right direction. This intervention was the least intensive of the four.

The Abecedarian Project found that, at age 15, those children who had received a preschool intervention scored higher on achievement tests and had lower incidence of special education and grade retention. At age 21, the children who received the preschool intervention were twice as likely to still be in school or to have attended a four-year college. The effects of interventions for school-aged children were either small or not statistically significant.

The Perry Preschool Project found positive effects on achievement tests, grades, high school graduation rates, and earnings, as well as lowering crime and welfare use, as of age 27. Only the Milwaukee Project found any long-term effect on IQ. That project found higher IQs by grade 8, but the control children enjoyed no other advantages.

Head Start has never been evaluated using randomized trials. Moreover, few studies have attempted to follow children past elementary grades. The studies that do exist, however, find that in the short-term, Head Start is contributing to school readiness by improving verbal skills and health. In short, the results are favorable, but much less conclusive than evidence from the above model programs.

How They Work: Getting Inside the Black Box of Interventions

Three questions arise when considering how the interventions effect the outcomes that they do. The first is one of targeting. Where should policymakers "draw the line" on instituting intervention programs? Do disadvantaged children benefit more from the program? Is there an optimal age for intervention? A larger question is what constitutes quality? If a program is showing better results, what factors contribute to those results: small class sizes, curriculum, teacher interaction?

A tentative answer to the first question-do programs have greater benefits for more disadvantaged children-is yes, although the answer likely depends on which aspect of disadvantage is being targeted. The investigators in the Abecedarian project found program effects to be twice as large for the most disadvantaged children. In general, the findings suggest that, where budgets are limited, it is appropriate to target early intervention to the most disadvantaged children.

What constitutes quality is a more difficult question to pin down. Typically people speak of quality as constituting a combination of structure (classroom size, e.g.) and classroom process (e.g., teacher interactions, developmentally appropriate material). There is little proof that the curriculum itself matters. What appears to matter more is how the teacher relates to and interacts with the children.

Finally, the optimal age of intervention has often been pinpointed in popular discussion as between birth and 3 years. Brain studies, however, show that older children are likely to be able to benefit from intervention as well. Hence, it may be more important to worry about the quality of the intervention than about the exact timing of it.

Policy Implications

Currie concludes that the better studies show larger and more significant effects of early intervention programs. However, there are still too few quality evaluations and disappointingly little consistent evidence of the long-term effectiveness (or lack of it) of early intervention. Also, the programs for which there are excellent evaluations are all model programs, and there is a considerable gap between such interventions and the large-scale, publicly funded interventions, such as Head Start. Nevertheless, the generally positive findings support expanded funding for Head Start.

In designing programs, the following guidelines might be kept in mind. Although it may be useful to intervene before age 3, interventions for preschool and school-age children can also be effective. Second, disadvantaged children seem to benefit more from early intervention, which provides a rationale for targeting such programs to these children. Finally, the most important aspect of quality is likely to be the nature of the interaction between teacher and child.

Background

Early interventions with children are designed to provide a secure, stimulating environment for children to learn. The Perry Preschool Project involved a half-day of preschool every weekday plus weekly 90-minute home visits over a fairly long time span. Teacher-student ratios were 1:6, and all teachers had master's degrees and training in child development. The Carolina Abecedarian, the Early Training Project, and the Milwaukee Project alter this design in various ways. For example, the Abecedarian involved full-day intervention from birth to age 5, followed up with an intervention for school-aged children. Head Start, of course, is the most famous publicly funded early intervention program. However, Head Start differs from the model programs, which typically are better funded and are subject to more intensive supervision.

Reprinted from the What We Know about Early Childhood Interventions, 2000 Joint Center for Poverty Research, Policy Brief, Vol. 2, No.10