Authors: Verne R. Bacharach, Ph.D., Appalachian State University; Alfred A. Baumeister, Ph.D., Vanderbilt University; Jaimily A. Stoecker, M.A., C.A.S., Caldwell County NC Public School District

There are various types of non-parental care arrangements for pre-kindergarten children. In the most popular type, center care, children are sent to a location where they are cared for in a group by a paid staff (Bacharach & Baumeister, 2003). We are going to refer to this type of care arrangement as pre-K center care. Psychologists and educators have argued that high quality pre-K center care can have a positive effect on the intellectual and behavioral development of children. These effects are thought to contribute to school readiness and subsequent academic success. We are going to examine the research literature to see if there is any compelling evidence based on science to support these claims.

A number of states have initiated, or are in the process of initiating, free pre-K center care programs for children from low-income families. For example, North Carolina has a program called Smart Start; Alabama is in the process of starting a program called Kid Stuff; and Georgia is currently operating a fully-funded statewide pre-K center care program for all four-year-olds. In the case of Smart Start (Dombro, 2002) and Kid Stuff (The Governor's Early Learning Commission Report, 2001), the states estimate that when fully implemented, these programs will cost in excess of $300 million per year.

There are various ways to scientifically investigate the effects of pre-K center care on child development. One method involves the random assignment of children to groups. This type of research procedure is referred to as a randomized trial and is widely believed to be the most effective way to establish a causal link between center care and child development. Using this procedure, one group of children, the center care group, is selected for the center care program. The other group is a control group: a group that is not enrolled in the program.

The purpose of random assignment is to assure that the groups, on average, do not differ from one another before the research begins. If children in the center care group differ from children in the control group before the study begins, then subsequent differences that might be found cannot be attributed to the effects of the intervention program. Random assignment does not always work as a procedure for equating groups, so it is important to examine the data from these studies to identify instances of randomization failure.

During the past 40 years there have been five relatively large-scale randomized trials conducted to investigate the relationship between pre-K center care and subsequent developmental outcomes in school-aged children. In each case, the pre-K center care was of the highest quality. We will examine each of these studies to see if they provide support for the claim that high quality pre-K center care contributes to the intellectual, academic, and behavioral development of children.

Abecedarian Project  
The Abecedarian Project (Ramey & Cambell, 1984) investigated the effects of pre-K center care on the behavioral and intellectual development of low-income children. Approximately 100 children, most of whom were black, participated in this study. Children were enrolled in the pre-K center care program shortly after birth and remained in the program until kindergarten.

With respect to behavioral outcomes, the results showed that the center care children were more likely than control children to have behavior problems in primary school. In addition, the researchers followed the participants in the study through adolescence into early adulthood, and after examining a wide variety of crime measures, the authors of the Abecedarian study concluded that the rate of crime did not differ between the groups (Clarke & Campbell, 1998).

The authors of the Abecedarian Project report that center care improved the intellectual functioning of the center care children. This conclusion is questionable for several reasons. First, all children were given intelligence tests at six months of age before the children had extensive experience in the program. At six months, the center care children had significantly higher levels of functioning than the control-group children (Abecedarian and CARE Archive, 2002), a difference that cannot be attributed to the effects of the daycare program. This finding is important because it suggests that randomization failed and that the center care children had higher levels of intellectual functioning than the control children before the study started. Subsequent differences between the groups may simply be related to this initial difference, and not to participation in the daycare program.

Second, although the data suggest that center care children had higher levels of intelligence than the control-group children when the children were young, this difference disappeared by the time the participants reached 15 years of age. The extent of difference at 15 years was the same as the extent of the difference that existed at six months of age. (See Abecedarian and CARE Archive 2002 and Campbell, Pungello, Miller-Johnson et al., 2001.) Furthermore, whatever the effect of pre-K center care on the children's intellectual functioning might have been, the size of the effect was of no practical significance. By 21 years of age, the average IQ for both groups was in the mid- to upper-80s (Campbell et al., 2001). The average IQ for adults is 100.

Third, the published results of the Abecedarian Project are difficult to interpret because during the course of the study participants dropped out of the study; some were lost permanently while others returned (surfaced) for later assessment. The dropout rates (attrition rates) and surfacing rates were not the same for both groups (Camp-bell, et al., 2001) and the authors of the published articles do not report data based solely on those participants who completed the entire battery of tests. Without data from the children who completed the entire battery of tests, it is impossible to tell what effect the loss of data from dropouts had on the results. Losing even a few participants from this study could have had important consequences for the study outcomes.

Published data from the Abecedarian Project also indicate that the center care children had more success in school than children in the control group. The center care children had higher scores on various measures of academic achievement and were less likely than control children to drop out of school (Campbell et al., 2001). However, it is unclear if the academic advantage associated with pre-K care was the result of the daycare program or was related to the fact that the center care children were on average, at birth, smarter than the control children.

The effectiveness of the Abecedarian model of center care on the intellectual development of children was further compromised by the results of subsequent studies. Project CARE, a sister study to the Abecedarian Project, and the Infant Health and Development Project, failed to find intervention effects on children's development. The daycare programs in both of these projects were modeled on the Abecedarian daycare program. Project CARE and the Infant Health and Development Project

Project CARE (Wasik, Ramey, Bryant, & Sparling, 1990) is of special interest because it was, in most ways, a direct replication of the Abecedarian Project. The study was done at the same research center by the same researchers using the same population of children and the same daycare program. The findings of the two studies, however, were not the same. In Project CARE there was no evidence of a link between participation in the daycare program and children's intellectual development. One obvious difference between the studies probably accounted for this discrepancy. At six months of age, the control and center care children in Project CARE had the same levels of intellectual functioning. At six months of age, the center care children in the Abecedarian Project had higher IQs than the control children. In other words, children's level of intellectual functioning in kindergarten can be predicted from their level of intellectual functioning before any of them experience daycare.

The Infant Health and Development Project was a state-of-the-art, very large-scale (approximately 900 participants) study of the effects of center care on the intellectual development of low birth weight babies. The children in this study were enrolled in an Abecedarian type pre-kindergarten program for two years. The center care program had no long-term effect on the children's intellectual development or on their academic achievement in primary school. The authors of the Infant Health and Development Project (McCarton, Brooks-Gunn, et al., 1997) stated: "Although it was hypothesized that the effects of early intervention would be most evident [italics added] in the prevention of school failure, no differences were found in the percentage of children who repeated a grade or who were placed in special education."

Houston Preschool Project and the Perry Preschool Project  
The two other randomized trial studies that have been reported are seriously flawed. In the Houston Preschool Project, the attrition rate was so severe that no scientifically valid conclusions are warranted (Johnson & Walker, 1987). The Perry Preschool Project (Berrueta-Clement, Schweinhart et al., 1984) is probably the most widely cited and best-known study investigating pre-K center care effects on child development. However, the scientific validity of the study has never been established because the results of the research have never been published in a peer-reviewed forum.

Peer review is an important component of any scientific endeavor. The review process gives independent scientists an opportunity to objectively evaluate the scientific merit of a study. For example, a peer-review would probably have identified one of the most obvious weaknesses of the study: The children in the Perry Preschool study were not assigned at random to the groups in the study (Spitz, 1993). Although there were other important methodological problems, the fact that the children were not assigned at random to the study groups compromises meaningful interpretation of the results.

Despite severe methodological problems, the results of the Perry Project have been widely cited as evidence for long-term positive effects of pre-K center care on the intellectual development of children. These "long-term" effects, however, did not last very long. By the time the children in the study were in the 2nd grade, the pre-K children's IQ advantage had vanished. By 14 years of age, the mean IQs for both groups were in the lower 80s (Lazar, Darlington, et al., 1982), an outcome that is reminiscent of the results of the Abecedarian Project.

The authors of the Perry Project report dramatic effects of their program on the adult behavior of the study participants. Although there was no evidence that the pre-K program had an effect on the children's behavior in school, as adults, the intervention children were less likely than the control children to commit crimes (Schweinhart, Barnes, et al., 1993). These results, however, are at odds with the results of the Abecedarian Project, in which there was no effect of pre-K center care on adult criminal behavior (Campbell et al., 2001). Which set of results is to be believed? Despite problems with the Abecedarian Project, it was clearly - from a methodological perspective - the better study, in part because participants in the Abecedarian Project were assigned at random to both the center care and control groups.

Regardless of which set of results is more believable, it is obviously foolish, at best, for states to develop expensive pre-K programs in anticipation of some type of payoff 15 to 20 years down the road when there is no consistent scientific evidence for the efficacy of these programs.

Conclusion 
Do pre-kindergarten center care programs work? Science tells us that pre-K center care, even the highest quality care, seems to have little or no effect on children's intellectual development or school performance, and that they might have negative behavioral consequences for young children. There is no convincing evidence to support the notion that they improve, or in any way influence, school readiness.

These programs do have a purpose: they are a source of free babysitting for low-income families. However, these pro-grams are very expensive. State estimates of the cost of fully implemented state operated pre-K programs run into the hundreds of millions of dollars per year. If states want to fund daycare for children from low-income families, the most cost-effective and family-friendly method would be to create voucher programs for this purpose.

Do No Harm: An Editorial Comment 
Do pre-K center care programs hurt children? Consider this: There are limited funds for programs that help young children. If the funds are squandered on ineffective programs, they will not be available for programs that work. Pre-kindergarten care probably does not hurt children who are in the care programs, but the programs are a drain on state funds that could be used to support child development and education programs that do help children. States that fund these programs are acting irresponsibly. They are diverting funds from programs that can improve the lives of children to programs that seem to have as their primary purpose, the service of special interests groups.

References 

• Abecedarian and CARE Archive (2002). V 1.0 Data File. Chapel Hill, NC: Frank Porter Graham Center.  
  
• Bacharach, V. R., & Baumeister, A. A. (2003). Child care and severe externalizing behavior in kindergarten children. Applied Developmental Psychology, 169, 1-11. 
  
• Campbell, F. A., Pungello, E. P., Miller-Johnson, S. et al. (2001). The development of cognitive and academic abilities: Growth curves from an early childhood educational experiment. Developmental Psychology, 37, 231-242. 
  
• Clark, S. H., & Campbell, F. A. (1998). Can intervention early prevent crime later? The Abecedarian Project compared with other programs. Early Childhood Research Quarterly, 13, 319-343. 
  
• Dumbro, A. L. (2002). What is Smart Start? Raleigh, NC: North Carolina Partnership for Children and Families. 
  
• McCarton, C. M., Brooks-Gunn, J., et al. (1997). Results at age 8 years of early intervention for low-birth-weight premature infants. Journal of the American Medical Association, 227, 126-132. 
  
• Ramey, C. T., & Campbell, F. A. (1984). Preventive education for high-risk children: Cognitive consequences of the Abecedarian Project. American Journal of Mental Deficiency, 88, 515-523. 
  
• Spitz, H. H. (1993). Were children randomly assigned in the Perry Preschool Project? American Psychologist, 48, 915-916. 
  
• The Governor's Early Learning Commission Report (2001). Our Children. Our Future. Our Plan. Montgomery, AL: Alabama Department of Children's Affairs.  
  
• Wasik, B. H., Ramey, C. T., Bryant, D. M., & Sparling, J. J. (1990). A longitudinal study of two early intervention strategies: Project CARE. Child Development, 61, 1682-1696.