Causal inference with confounders missing not at random

• 成果类型: Article
• 署名作者: Yang, S.; Wang, L.; Ding, P.
• 署名单位: North Carolina State University; University of Toronto; University of California System; University of California Berkeley
• 刊物名称: BIOMETRIKA
• ISSN/ISSBN: 0006-3444
• DOI: 10.1093/biomet/asz048
• 发表日期: 2019
• 页码: 875888
• 关键词: propensity score; imputation; DESIGN

摘要

It is important to draw causal inference from observational studies, but this becomes challenging if the confounders have missing values. Generally, causal effects are not identifiable if the confounders are missing not at random. In this article we propose a novel framework for non-parametric identification of causal effects with confounders subject to an outcome-independent missingness, which means that the missing data mechanism is independent of the outcome, given the treatment and possibly missing confounders. We then propose a nonparametric two-stage least squares estimator and a parametric estimator for causal effects.