A substantial body of empirical evidence indicates these preschool programs have prevented grade repetition and special education placements for disadvantaged children over the long term. A review of over 30 longitudinal studies by Barnett (1998) concluded that preschool programs serving disadvantaged children also produced long-term gains in achievement as measured by standardized tests. In drawing this conclusion, Barnett relied heavily on the findings of controlled experiments with sound longitudinal follow-ups that lost few study participants over time. The few studies that have examined high school graduation rates found sizable effects on these as well (Barnett, 1998).

In contrast to the findings for other outcomes, initial effects on IQ tests clearly disappear over time in the vast majority of studies. Why this occurs and how important it is are much less clear. There is considerable controversy about how well IQ measures intelligence in the way it is commonly understood by the general public (Sternberg and Detterman, 1986; Neisser et al., 1995). The lack of long-term gains in IQ, at the same time that such gains are produced in subject-matter-specific knowledge and skills and school success, raises similar questions. However, two of the most intensive programs, which began full-day, year-round educational child care in the first year of life and continued to age 5, produced very large initial IQ effects and some IQ advantage that persisted years after leaving the program (Garber, 1988; Campbell and Ramey, 1993). Even in these studies, the size of the effect on IQ declines over the years, while the improvements in achievement and school success do not (Barnett, 1998). It is also interesting that a similar program, with a primary focus on parents and relatively greater emphasis on social-emotional development, did not sustain effects on IQ even up to the end of the program (Lally et al., 1987).

The programs that researchers developed specifically to investigate the influence of preschool education on economically disadvantaged children are a useful source of information about positive influences on development. These programs have been found to be highly effective in producing immediate benefits for children and to produce longer-term effects in at least a dozen rigorous longitudinal studies. Some of the studies with the strongest outcomes were highly controlled random assignment experi-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement