Experimental designs provide some of the strongest evidence of causality, primarily in response to the removal of confounding through randomization. But observational data abounds in the social and behavioral sciences, limiting causal inference.This workshop will introduce the potential outcomes framework for causal inference and the role of matching methods in estimating treatment effects in non-experimental data. We’ll focus on propensity score matching, one among many matching approaches, and walk through the implementation and diagnostics of matching. The workshop is intended for participants who are comfortable with multiple regression and familiar with limited dependent variables. We’ll use R to illustrate the approach.