Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 56 - Causal Inference
Type: Contributed
Date/Time: Sunday, August 8, 2021 : 3:30 PM to 5:20 PM
Sponsor: Section on Statistics in Epidemiology
Abstract #318367
Title: Semiparametric Estimation of Treatment Effects When Outcome Is Missing Not at Random
Author(s): Shintaro Yoneyama* and Mihoko Minami
Companies: Keio University, Graduate School of Science and Technology and Keio University, Department of Mathematics
Keywords: missing not at random; causal inference; observational study; counterfactual framework; instrumental variable; estimating equation approach

When investigating treatment effects in the observational study in epidemiology and social science, we often face two problems: confounding and missing data. In order to properly estimate the treatment effects with missing data, these two problems must be dealt with simultaneously. When the outcome is missing not at random (MNAR), existing estimation methods for treatment effects require some modeling of the outcome. However, identifying the proper model for the variable with MNAR is quite difficult. In this presentation, we propose a method to estimate the average treatment effect (ATE), the average treatment effect on the treated (ATT), and the average treatment effect on the untreated (ATU), without modeling the outcome when the outcome is MNAR by using an instrumental variable. We show that the proposed estimators are consistent and asymptotically normal. Through the simulation results, we also show that the proposed estimators perform better in terms of RMSE than other commonly used methods for adjustment of missing.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2021 program