HighScope Perry Preschool Study
The following research bulletin was originally published online by the Phi Delta Kappa Center for Evaluation, Development, and Research. Reprinted with permission.
Phi Delta Kappa Center for Evaluation, Development, and Research (June 2002, No. 32)
How the HighScope Perry Preschool Study Grew: A Researcher's Tale
By Lawrence J. Schweinhart, Ph.D.
The HighScope Perry Preschool Study, which was initiated in the early 1960s, is now widely regarded as a landmark study establishing the human and financial value of high-quality preschool education. The purpose of the study was to evaluate the HighScope model, in which teachers help children plan, carry out, and review their own educational activities.
David Weikart and his colleagues developed the model and initiated the study as a local school district project. I joined the evaluation team in 1975. This is my story of this extraordinary study.
Weikart's early childhood work began in 1962 with the Perry Preschool program in the Ypsilanti Public Schools, where he was special education director. The district's ineffective way of coping with rampant school failure was grade retention. Weikart began the preschool program in an effort to get ahead of the problem. Many students had moved to Ypsilanti from a part of the country where some poor and minority students were not beginning their education until after first grade. Weikart and his colleagues reasoned that just as it would have been better for these students to have begun school in kindergarten as their classmates did, so it would have been even better for them to have started school a year or two earlier than that.
Two federal initiatives, Head Start and the Bureau for the Education of the Handicapped, both begun in 1965, prompted a nationwide demand for early childhood education programs and teachers, particularly for children living in poverty and children with other special needs.
There were few early childhood education programs in the nation, but one of them was in the Ypsilanti Public Schools. The teaching staff, still in the first years of their experimental program, found themselves in demand throughout the nation as teacher trainers. Their numbers and budget grew until they no longer fit within the confines of a small school district.
The research team
Weikart and his colleagues established the nonprofit HighScope Educational Research Foundation in 1970 to house their proliferating work in early childhood education.1
I joined the research team in 1975, HighScope had grown; and the researchers were conducting follow-up studies of several internal programs, as well as evaluating Head Start's Developmental Continuity Project, an early effort to get schools and Head Start programs to work together. My work began with the HighScope Preschool Curriculum Comparison Study. 2 Then I turned to the HighScope Perry Preschool Study through age 15.3
The Perry study of the effects of the preschool program on children through age 10 found that, despite the diminishing effect of the program on children's intellectual performance, fewer children who had been enrolled in the program (17%) were held back a grade or placed in special education than were children who had not been enrolled in the program (38%).4 This principal finding — along with consistent, nearly significant program effects on achievement test scores from first through fifth grades — inspired us to collect data from 14- and 15-year-olds who had been enrolled in the program, hoping to find even longer-term effects. The long-term study was becoming a lifelong study.
When I first joined the research team, I was dubious about finding any longer-term program effects. After all, such long-term effects had never been found for a part-day preschool program. I had learned to be skeptical about research claims until all the evidence was in. The evidence did come in. I still remember my first look at the computer printouts revealing the substantial program effect on achievement test scores for 14-year-olds. I thought it was a mistake. Surely the weak effects found for first through fifth grades would disappear nine years after the kids had left the preschool program. But the effect was actually bigger than it was for the children in their earlier years, and it definitely was statistically significant. Perhaps this quest for finding long-term effects was not so quixotic after all.
Consortium for Longitudinal Studies
In the late 1960s and early 1970s, the research examining the effectiveness of preschool programs was inconclusive. An evaluation of the Head Start program by the Westinghouse Learning Corporation published in 1969 led policy makers and the public to believe that Head Start was a failure.5 At the same time, Urie Bronfenbrenner and his colleagues reviewed existing studies of early childhood program effects and drew the questionable conclusion that the only critical feature of effective preschool programs was that they targeted parents.6 In order to gather stronger, clearer evidence on these questions, Irving Lazar brought together researchers who had been conducting longitudinal studies of the effects of early childhood programs to form the Consortium for Longitudinal Studies. The Consortium's principal contribution was to refute the idea that preschool program effects fade with time.7 The group's work also identified clear long-term effects for children who had attended diverse early childhood programs — some focusing on parents, some on children, and some on both — indicating that fewer were placed in special education programs or retained a grade and that more graduated from high school.
The growth of findings
Since 1970, HighScope has published five comprehensive monographs on the HighScope Perry Preschool Study, one on the effects of the program at the end of children's enrollment in preschool, one at age 10, one at age 15, one at age 19, and one at age 27.8 We now are collecting data for a report on the effects of the HighScope Perry Preschool program at age 39 through 41.
We did not set out to conduct such an extraordinarily long study. Each set of findings along the way served as the impetus for the next round of data collection. The critical findings pertained to intellectual performance at the end of the preschool and grade placement at age 10. Subsequently, findings expanded to consider the effects of the program on education performance and community behavior at age 15; on education attainment, employment, and incidence of antisocial behavior at age 19; and incidence of crime, education attainment, level of earnings, rate of home ownership, level of welfare assistance, and incidence of single parenthood at age 27.
Comparing the preschool group to the no-preschool group, we found the following significant differences through age 27:
Incidence of crime. Only 7% of adults who had participated in the Perry Preschool program had been arrested five or more times, compared with 35% of those who had not participated in a preschool program. Of those in the preschool program group, 7% had ever been arrested for drug-related offenses, compared to 25% of those in the no-program group.
Earnings and economic status. Adults in the program group were four times more likely (29%) to earn $2,000 or more per month than were adults in the no-program group (7%). Almost three times as many (36%) owned their own homes, compared to those in the no-program group (13%). More than two times as many (program 30%, no program 13%) owned a second car. As adults, 59% of those in the program group had received welfare assistance or other social services at some time, compared to 80% of those in the no-program group.
Educational attainment. Seventy-one percent of those in the program group graduated from regular or adult high schools or received General Education Development certification, compared with 54% of those in the no-program group. Earlier in the study, the preschool program group had significantly higher average achievement scores at age 14 and literacy scores at age 19.
Marriage and single parenthood. Forty percent of women in the program group were married at the time of the age-27 interview, compared to 8% of those in the no-program group; and 57% of women in the program group were single parents, compared to 83% of those in the no-program group.
A distinctive feature of this study has been its economic cost-benefit analysis, an idea developed by Weikart and Phillips Foster, then an agricultural economist at the University of Maryland. Carol Weber, one of Foster's doctoral students, carried out an early cost-benefit analysis of the program.9 Weber's research served as the basis for the analyses Steve Barnett conducted for the age-19 study and updated for the age-27 study.10 The analysis weighs the cost of the preschool program against the economic benefits resulting from the program — higher earnings, reduced incidences of special education services, welfare assistance, and crime. The age-27 analysis found that every public dollar spent on the program saved $7.16 in tax dollars.
This $7.16 figure has become the most often cited statistic from the study. It enabled us to convey the worth of these activities in terms that business people understand. The Committee for Economic Development embraced early childhood education as a cause on the basis of the findings of this study. The late Brad Butler, then president and CEO of Procter and Gamble, became a spokesperson for the cause, crisscrossing the nation with the message that high-quality early childhood programs are a worthwhile public investment. I joined him in testifying before the North Carolina Legislature on behalf of the Smart Start program, where in fewer than five minutes he managed to mention twice that he was a conservative Republican in favor of the program.
Extant public policies did not reflect the substantial and growing evidence that high-quality preschool child development programs for young children living in poverty are a wise public investment. Head Start and similar programs served only about one-fifth of the eligible children. Further, inadequate expenditures per child and weak standards for curriculum, assessment, and parent outreach made it difficult to maintain adequate staffing, quality, and effectiveness in these programs. In order to influence public policy, we wanted to let the world know about the increasingly positive findings of the HighScope Perry Preschool Study.
In 1979, with funding from the Carnegie Corporation, we established the Center for the Study of Public Policies for Young Children, later known as the Voices for Children Project. The main purpose of the center was to disseminate the findings of the HighScope Perry Preschool Study, the Consortium for Longitudinal Studies, and other preschool research to policy makers and opinion leaders. We accomplished this through publications and presentations to national associations, including the National Governors' Association, the National Conference of State Legislatures, national education associations, and state conferences in the majority of the states. We also trained cadres of speakers, and we coordinated statewide speaking efforts in Michigan, Ohio, North Carolina, and South Carolina. This work drew on and contributed to these four states' leadership in early childhood programs. Michigan and South Carolina have state preschool programs. Ohio contributes substantial funding to the state's Head Start programs. North Carolina's Smart Start program is a model for community-directed funding of programs for young children, as well as home to the TEACH program for improving caregivers' education and compensation.
Although the local newspaper ran an article about the Perry Preschool Study through age 10, we wanted to gain more widespread publicity about the study's findings for children at age 15. Perhaps if more people knew about the study, the Carter Administration could use the political capital generated by the findings to increase funding for Head Start. In 1980, with support from the Carnegie Corporation, we sent press releases throughout the country and held a well-attended press conference in New York City. The story was covered by news media throughout the United States. This initiative contributed to a long period of bipartisan support for Head Start: funding increased from $735 million in 1980 to $6.2 billion in 2001.
Four years later, in 1984, we sought the involvement of the Reagan Administration, securing the support of Dorcas Hardy, assistant secretary for Human Development Services, who was responsible for overseeing Head Start. A Washington, D.C., consulting firm helped us arrange a press conference in the nation's capital. As a result, David Stockman, then director of the federal Office of Management and Budget and a former Michigan Congressman, made Head Start one of the Reagan Administration's seven social safety net programs, sparing it from a cut in government funding.
I believe the HighScope Perry Preschool Study has three strengths that merit the extraordinary attention it has received. First, its design involved random assignment of poor children either to a preschool program group or a no-preschool program group. Technically, the assignment procedure had elements of accommodation to reality (for example, younger siblings were assigned to the same groups as their older siblings to prevent the spread of program effects into the no-program group). But it was close enough to true random assignment to inspire confidence that group differences are due to the effects of the preschool program. Second, longitudinal follow-ups through age 27 had very little missing data — an average of only 5% per measure, minimizing attrition as a potential source of design contamination. Third, the pattern of findings is internally consistent and plausibly related to the preschool program.
The simple message of this study — and a growing number of studies like it — has gotten through to a great many people. Despite political turns to the right and left in national leadership, Congress continues to provide solid bipartisan support for Head Start. The majority of states now offer state preschool programs, most aimed at averting academic and social problems for at-risk children and an increasing number at providing access to all children whose parents want them. Both Head Start and state preschool programs have their own style of commitment to program quality, shaped by the myriad decisions of program planners and practitioners.
Most important, this study and others like it offer us hope, hope that we can make a difference in the lives of children, especially those living in poverty, and hope that focused government action can be a part of the solution. Hope is a fragile thing, easily undermined by strong doses of unfortunate, seemingly intractable realities. But it is only with hope that we can make a lasting contribution to our society. And when objective research reinforces that hope, it allows researchers to merge our personal values and professional activities toward positive, well-grounded actions.
For further information, see HighScope's website at www.highscope.org; write to HighScope Educational Research Foundation, 600 North River Street, Ypsilanti, MI 48198-2898; e-mail email@example.com, or call 734.485.2000.
David P. Weikart et al., Ypsilanti Preschool Curriculum Demonstration Project: Preschool Years and Longitudinal Results Through Fourth Grade (Ypsilanti, Mich.: HighScope Press, 1978).
Lawrence J. Schweinhart and David P. Weikart, Young Children Grow Up: Effects of the Perry Preschool Program on Youths Through Age 15 (Ypsilanti, Mich.: HighScope Press, 1980).
David P. Weikart, James T. Bond, and J. T. McNeil, Ypsilanti Perry Preschool Project: Preschool Years and Longitudinal Results Through Fourth Grade (Ypsilanti, Mich.: HighScope Press, 1978).
Westinghouse Learning Corporation, Impact of Head Start: Evaluation of the Effects of Head Start on Children's Cognitive and Affective Development, 2 vols. (Washington, D.C.: Clearinghouse for Federal, Scientific, and Technical Information, 1969).
Urie Bronfenbrenner, Is Early Intervention Effective? vol. 2 of Report on Longitudinal Evaluations of Preschool Programs (Washington, D.C.: U.S. Office of Child Development, 1974).
Irving Lazar and Richard Darlington, Lasting Effects of Early Education: Report from the Consortium for Longitudinal Studies (Chicago: University of Chicago Press for the Society for Research in Child Development, 1982); Consortium for Longitudinal Studies, As the Twig Is Bent — Lasting Effects of Preschool Programs (Hillsdale, N.J.: Erlbaum, 1983).
David P. Weikart et al., Longitudinal Results of the Ypsilanti Perry Preschool Project (Ypsilanti, Mich: HighScope Press, 1970); Weikart, Bond, and McNeil, Ypsilanti Perry Preschool Project; Schweinhart and Weikart, Young Children Grow Up; John R. Berrueta-Clement et al., Changed Lives: Effects of the Perry Preschool Program on Youths Through Age 19 (Ypsilanti, Mich.: HighScope Press, 1984); Lawrence Schweinhart, Helen V. Barnes, and David P. Weikart, Significant Benefits: The HighScope Perry Preschool Study Through Age 27 (Ypsilanti, Mich.: HighScope Press, 1993).
Carol U. Weber, Phillips W. Foster, and David P. Weikart, Economic Analysis of the Ypsilanti Perry Preschool Project (Ypsilanti, Mich.: HighScope Press, 1978).
W. Steven Barnett, Lives in the Balance: Age-27 Benefit-Cost Analysis of the HighScope Perry Preschool Program (Ypsilanti, Mich: HighScope Press, 1996).
Lawrence J. Schweinhart is President at HighScope Educational Research Foundation in Ypsilanti, Michigan.