Examining Differential Attrition in a State Pre-Kindergarten Impact Study Using Regression Discontinuity Design

• In Event: Poster Session 3
  In Poster Session: Head Start and Pre-Kindergarten Issues

Abstract

The proposed paper examines differential attrition in a prekindergarten impact study using Regression Discontinuity Design with age as a cut-off variable. While RDD has been increasingly used to evaluate impact of public preschools very few has fully discussed the impact that attrition may have on their findings. The proposed presentation discusses five types of attrition examined in a recent impact study using RDD to examine Michigan’s state prekindergarten program -- Great Start Readiness Program (GSRP). Through this work, we hope to illuminate the potential bias that these differential attritions may create, suggest possible reasons (including Michigan-specific reasons) for this phenomenon, and propose methodological adjustments for data collection and analysis.

Authors

• Tomoko Wakabayashi, HighScope Educational Research Foundation

• Zongping Xiang, HighScope Educational Research Foundation