Causal Inference in Randomized Encouragement Design Studies with Missing-Data
In this talk I consider a problem in causal inference where we want to estimate the local complier average causal effect (CACE) parameter in the setting of a randomized clinical trial with a binary outcome, cross-over non-compliance, and missing data. I first discuss the moment and ML estimators that require the assumption of latent ignorability and then discuss the development of a moment estimator that relaxes the assumption of latent ignorability and incorporates sensitivity parameters that represent the relationship between potential outcomes and associated potential response indicators.
Seminar Date:
April 10, 2007