SpeakerDr. Liang Li
The University of Texas
MD Anderson Cancer Center
TitlePropensity Score Analysis with Matching Weights
AbstractPropensity score matching is widely used for studying treatment effects in observational studies. In this talk, I will introduce the method of matching weights as an analogue to the most widely used propensity score matching method in medicine: one-to-one pair matching on the propensity score without replacement. Compared with pair matching, the proposed method offers more efficient estimation, more accurate variance calculation, substantially better balance, and simpler asymptotic analysis. A statistical test for the misspecification of the propensity score model is proposed for balance checking purposes. An augmented version of the matching weight estimator is developed that has the double robust property, i.e., the estimator is consistent if either the outcome model or the propensity score model is correct. I will illustrate the use of the matching weight method in three real data examples that correspond to different degrees of overlap in the propensity score distributions between the treatment groups.