Comparing Propensity Score Matching and Coarsened Exact Matching: An Example Case Using Community College Outcomes

• In Event: Using National Data Sets to Explore Postsecondary Pathways and Outcomes

Abstract

This study compares two quasi-experimental matching methods: coarsened exact matching and propensity score matching- to evaluate the community college outcomes of high school career and technical education (CTE) completers from the Illinois Class of 2003 to their observationally equivalent peers who had completed college preparatory programs. We use the same covariates in both matching methods, but in the case of PSM the coarsened variables are used to calculate propensity scores (predicted probability of treatment group membership).

Authors

• Eric Lichtenberger, Illinois Board of Higher Education

• Cecile Dietrich, Radford University