Marginal generalized raking for missing or mismeasured data
Date:
I discuss an extension of generalized raking based on influence functions to estimate marginal parameters (e.g., the average treatment effect), which is more robust to model misspecification than prior methods.