Propensity Score Matching
Understanding the impact of interventions requires answering the critical question: “Compared to what?” Propensity-based matching is a powerful method used to identify comparison groups that closely resemble treatment groups by leveraging known characteristics of a population. This course offers a comprehensive introduction to propensity score analysis, a widely utilized tool in the evaluator’s toolbox.
Participants will learn the fundamentals of propensity score analysis, including when and why these techniques are helpful. Through a step-by-step walkthrough, the course will explore two distinct examples: one from an economics perspective and another from the criminal justice field. These examples will demonstrate the practical application of propensity score analysis, helping participants understand how to implement this methodology in their own programs.
Additionally, the course will include hands-on experience with R software, providing participants with the skills to conduct their propensity score analyses effectively.
Speakers:
Gio Circo, PhD
Gio Circo is a data scientist at Gainwell Technologies, where he is part of the Artificial Intelligence and Machine Learning team. His work primarily focuses on causal inference, Bayesian statistics, and anomaly detection. Before joining Gainwell Technologies, Gio was an Assistant Professor at the University of New Haven, where his research centered on gun violence, policing, and the fear of crime. He earned his Ph.D. in Criminal Justice from Michigan State University and holds an M.A. and a B.A. in Criminal Justice from Illinois State University.
Materials:
- Propensity Score Matching Part 1 (19:31)
- Propensity Score Matching Part 2 (30:25)
- Propensity Score Matching Part 3 (23:02)
- View slides
- Code through example