Impacts of the New Jersey Abbott Preschool Program through Middle and High School

• In Event: 3-044 - Evaluating Publicly-Funded Preschool Programs: Understanding Impacts, Impact Variation, and Lasting Effects

Integrative Statement

State and local investments in early childhood education (ECE) have increased rapidly over the last several years (Friedman-Krauss et al., 2018) based in part on a growing body of research that high quality, well-designed ECE programs can produce large gains in children’s learning and development at kindergarten entry (e.g., Barnett et al., 2018; Gormley et al., 2005; Weiland & Yoshikawa, 2013; Yoshikawa et al., 2013). Fewer studies have estimated impacts of ECE on children’s longer-term outcomes as they progress through school, often finding evidence of fade-out of effects (e.g., Puma et al., 2012; Reynolds et al., 2002; Schweinhart et al., 2005). Even less is known about the long-term impacts of state-funded ECE.

The current paper will estimate the long-term impacts of New Jersey’s state-funded Abbott Preschool Program through middle and high school. The Abbott Preschool Program provides full-day ECE to 3- and 4-year-olds in 31 high-poverty districts in New Jersey. The program is considered high quality based on policy standards (Friedman-Krauss et al., 2018) and observed classroom quality (Barnett & Frede, 2017). Prior research finds significant impacts of the program on predominantly disadvantaged children’s academic skills: Results of a regression discontinuity design evaluation demonstrated significant positive impacts of one year of Abbott preschool on children’s vocabulary (E.S. = 0.34), math (E.S. = 0.35), and literacy skills (E.S. =0.50) at kindergarten entry (Barnett et al., 2018). The program has also been found to have positive impacts on children’s math, language, and science achievement, special education placement, and grade retention through fourth and fifth grades, with larger impacts for children who attended Abbott Preschool for two years compared to one year (Barnett et al., 2013).

The current paper will extend these findings through middle and high school, and will also explore if long-term impacts vary by children’s race/ethnicity, gender, and/or home language. To obtain the initial sample, kindergarten classrooms were randomly selected from the 15 largest Abbott school districts. Children were then randomly sampled from those classrooms without regard for whether or not they had attended Abbott Preschool during the previous year. Children who did (n=766) and did not (n=305) attend Abbott Preschool were followed through the 2016-2017 school year when they were in eleventh grade. The current study utilizes administrative data obtained from the state including records of children’s special education placement and in-grade retention as well as state standardized test scores in language, math, and science.

Ordinary least squares regression will be used to model the impacts of attending Abbott Preschool on children’s standardized test scores and logistic regression will be used to estimate impacts on special education placement and in-grade retention. To reduce selection bias, all models will control for student race/ethnicity, free lunch status, gender, age, and home language, and parent employment, mother marital status, parent education, and school district fixed effects. Clustered standard errors will be used to adjust for the non-independence of the data, as children are nested in classrooms and schools. Discussion of the results will focus on implications for ECE policy and practice.

Authors

• Allison Friedman-Krauss, NIEER/Rutgers University
  Presenting Author

• W. Steven Barnett, NIEER/Rutgers University
  Non-Presenting Author

• Kwanghee Jung, NIEER/Rutgers University
  Non-Presenting Author

• Ran Guo, NIEER/Rutgers University
  Non-Presenting Author

• Ellen Frede, NIEER/Rutgers University
  Non-Presenting Author