Pre-kindergarten Intermediate Impacts - Results from Randomized Clinical Trials and Cohort Studies

• In Event: Projecting Intervention Impacts in Developmental and Educational Studies

Abstract

Introduction. State pre-Kindergarten (pre-K) programs are offered in 44 of the 50 states across the United States and are discussed as an important policy lever for providing opportunities to children from low-income families (Friedman-Krauss et al., 2020). Evaluations of such programs show promising results for children’s academic achievement at the end of the pre-K year (Phillips et al., 2017; DeAngelis et al., 2020). Longer-term effects, those found after the pre-K year, are much smaller and in many instances show no differences between children that attended pre-K programs and those that did not (Phillips et al., 2017). Recent results from a randomized controlled trial of the Tennessee pre-k program indicate children that attended the program had significantly worse academic outcomes, including in achievement, and increased special education placement and school disciplinary actions into late elementary school after positive effects were found at the end of the pre-K year (Durkin et al., 2022). Reasons posited for these effects include lower quality pre-K programs or subsequent elementary school environments, peer effects, and control group experiences. Yet, it is unknown the extent to which the pattern of fadeout among the TN study resembles other recent state pre-K study findings.
In this study we investigate longer-term pre-K effects on various academic and social-emotional skills in both randomized clinical trials (RCTs) and longitudinal cohort studies and contrast them with the seminal early childhood education studies. We also explore several potential reasons for fadeout, such as the need for children to transition from pre-k to high quality school settings (“sustaining environments”), changes in counterfactual conditions over time, and changes over time in the focus in pre-k instruction to early academic skills also taught in kindergarten.

Methods. For our initial analyses, we described and contrasted the academic and social-emotional skills in kindergarten and grade 1 for pre-k attendees and nonattenders in two RCTs (Tennessee, North Carolina) and four Early Learning Network (ELN) cohort studies, using the results in published papers and technical reports. A meta-analysis of these results and those from state evaluations and of potential causes for declining longer-term effects will be conducted.

Results. Table 1 shows the descriptive statistics from the two RCTS and Table 2 shows the descriptive statistics from the four ELN studies. Both the TN and NC RCTs showed initial impacts that disappeared and became nonsignificantly negative in NC. The initial differences between pre-k attendees and nonattenders also declined over time in the ELN studies, despite differences in study design and who was included in the nonattender group.

Discussion. These results suggest that the TN results of initial pre-K impacts that fade over time to become negative (albeit not reliably in first grade) are found in at least one other study. Future analyses and research should consider both study design and whether the nonattender group includes children who attended other early learning settings (NC, TN) or not (ELN studies).

Authors

• Ana Whitaker, University of Delaware
  Presenting Author

• Margaret Burchinal, University of Virginia - Charlottesville

• Drew Bailey, University of California - Irvine

• Jade V Marcus Jenkins, University of California - Irvine

• Tyler Watts, Teachers College, Columbia University

• Greg J Duncan, University of California, Irvine