Indiana University SPEA Edward J. Bloustein School of Planning and Public Policy University of Pennsylvania AIR American University

Panel Paper: Long-Term Impacts of Preschool Programs on Graduation Rates, Special Education Placement, and Grade Retention: A Meta-Analysis

Thursday, November 12, 2015 : 10:15 AM
Brickell South (Hyatt Regency Miami)

*Names in bold indicate Presenter

Dana Charles McCoy1, Kathleen Ziol-Guest2, Rui Yang2, Hirokazu Yoshikawa2, Holly Schindler3 and Greg J. Duncan4, (1)Harvard University, (2)New York University, (3)University of Washington, (4)University of California, Irvine
In 2013, 1.3 million children attended a state-sponsored preschool program in the United States, representing an almost 50% increase since 2003 (Barnett, Carolan, Squires, & Brown, 2013; Barnett, Robin, Hustedt, & Schulman, 2003).  Financial investments in pre-k are also on the rise, with states spending almost $5.4 billion in 2013 to support early childhood education (ECE) programs for 3- and 4-year-olds (Barnett et al., 2013).  Despite increased access to and investment in ECE, evidence is mixed regarding its utility in improving long-term educational outcomes. In the present study, we use meta-analysis to synthesize more than 40 years of high-quality research on classroom-based ECE programs to quantify their longer-term impacts on high school graduation, special education placement, and grade retention.  In particular, we build upon previous meta-analyses (e.g., Aos et al., 2004; Camilli, Vargas, Ryan, & Barnett, 2010; Gorey, 2001; Lazar et al., 1982) by 1) focusing only on studies that use either experimental or high-quality quasi-experimental designs and 2) distinguishing between effects of ECE on each individual outcome.

This study makes use of a comprehensive database of ECE program evaluations conducted in the United States between 1960 and 2007.  Included evaluations were selected via systematic review of 10,000+ articles and were required to be of high quality (e.g., to show <50% attrition, to have ≥10 subjects in each group, etc.).  A total of 17 studies comparing 25 contrasts of ECE programs (against non-ECE control groups) were included in the present analysis.  Three primary outcomes – high school graduation, special education placement, and grade retention – were selected at the latest available time point. For analysis, all outcomes were coded such that positive effect sizes (as captured using Hedges’ g) indicated more desirable outcomes (i.e., higher graduation rates, and lower rates of special education/grade retention) for ECE participants.

For analysis, we used a two-level random intercept model with effect sizes at level 1 nested in contrasts at level 2.  All models were weighted by the inverse variance of the effect size times the inverse of the number of effect sizes per contrast (Cooper & Hedges, 2009; Lipsey & Wilson, 2001).  Results of these analyses suggest an average impact of 0.35 SDs (S.E=0.07, p<.01) across all outcomes.  Specifically, the average impact of ECE on graduation was 0.32 SDs (S.E.=0.04, p<.01), on special education was 0.35 SDs (S.E=0.06, p<.01), and on grade retention was 0.29 SDs (S.E=0.07, p<.01).  Using the 76.5% of effect sizes with percentage point data available, these effect sizes translate to an 11.7 percentage point increase in graduation, 11.6 percentage point decrease in special education, and a 10.7 percentage point decrease in grade retention for children attending ECE. 

Results suggest that, on average, ECE programs have historically had moderate, positive impacts for improving students’ graduation rates, reducing special education placements, and reducing grade retention. Additional sensitivity analyses will be included in the final presentation.  Implications of this work for potential ECE-based cost-savings and policy will also be discussed.