Simplifying Matching Methods for Causal Inference

Presentation Date: 

Monday, October 16, 2017


Harvard Health Policy and Insurance Research Seminar

Presentation Slides: 

This presentation shows how to use matching to ameliorate model dependence -- where small, indefensible changes in model specification have large impacts on our conclusions. We introduce matching methods that are simpler, more powerful, and easier to understand. We also show that the most commonly used existing method, propensity score matching, should rarely be used. Easy-to-use software is available to implement all methods discussed.