Estimating the Treatment Effect of City Connects Using Difference-in-Differences Analysis

• In Event: Poster Session 3
  In Poster Session: Research Methods and Approaches to Examining Academic and Behavioral Outcomes

Abstract

In educational research, random assignment into treatment and control groups is not always possible due to practical, ethical, and political reasons. Thus, it becomes challenging to argue that observed effects are casual in quasi-experimental studies. However, literature suggests design elements and statistical controls may improve the causal interpretation of the estimates. Also, Fisher and Rosenbaum recommend conducting several observational studies to examine the pattern across findings. Following these approaches, we used a multiple baseline interrupted time series design with a comparison group and estimated the treatment effects of City Connects, a student support intervention implemented in high-poverty elementary schools, on report card scores using a Difference-in-Differences approach. Results indicated positive treatment effects on math report card scores for City Connects.

Authors

• Bercem Akbayin, Boston College

• Terrence J. Lee-St. John, Boston College

• Eric Dearing, Boston College

• Anastasia Raczek, Boston College

• Erin Sibley, Boston College

• Mary Elizabeth Walsh