EcoSta 2022: Start Registration
View Submission - EcoSta2022
A0997
Title: Identification and estimation of natural mediation effect in the presence of treatment induced confounding Authors:  Fan Xia - University of Washington (United States)
Kwun Chuen Gary Chan - University of Washington (United States) [presenting]
Abstract: Natural mediation effects are desirable estimands for studying causal mechanisms in a population, but complications arise in estimating natural indirect effects in the presence of treatment-induced confounding. For instance, the usual sequential ignorability assumption no longer guarantees identification. We will consider additional assumptions needed to guarantee identifiability and its relationship to interventional effects. Moment-type and locally efficient estimators will be presented.