Contrasting Identifying Assumptions of Average Causal Effects: Robustness and Semiparametric Efficiency

11/30/2021
by   Tetiana Gorbach, et al.
0

Semiparametric inference about average causal effects from observational data is based on assumptions yielding identification of the effects. In practice, several distinct identifying assumptions may be plausible; an analyst has to make a delicate choice between these models. In this paper, we study three identifying assumptions based on the potential outcome framework: the back-door assumption, which uses pre-treatment covariates, the front-door assumption, which uses mediators, and the two-door assumption using pre-treatment covariates and mediators simultaneously. We derive the efficient influence functions and the corresponding semiparametric efficiency bounds that hold under these assumptions, and their combinations. We compare the bounds and give conditions under which some bounds are lower than others. We also propose semiparametric estimators, quantify their efficiency and study their robustness to misspecification of the nuisance models. The theory is complemented with simulation experiments on the finite sample behavior of the estimators. The results obtained are relevant for an analyst facing a choice between several plausible identifying assumptions and corresponding estimators. Here, the choice is a trade-off between efficiency and robustness to misspecification of the nuisance models.

READ FULL TEXT
research
12/03/2020

Multiply robust estimation of causal effects under principal ignorability

Causal inference concerns not only the average effect of the treatment o...
research
11/28/2022

Causal Inference with Confounders MNAR under Treatment-independent Missingness Assumption

Causal inference in observational studies can be challenging when confou...
research
01/30/2023

Bagging Provides Assumption-free Stability

Bagging is an important technique for stabilizing machine learning model...
research
11/07/2017

Overlap in Observational Studies with High-Dimensional Covariates

Causal inference in observational settings typically rests on a pair of ...
research
04/20/2023

Identification and multiply robust estimation in causal mediation analysis with treatment noncompliance

In experimental and observational studies, there is often interest in un...
research
01/28/2023

Covariate-assisted bounds on causal effects with instrumental variables

When an exposure of interest is confounded by unmeasured factors, an ins...
research
06/26/2019

User Sentiment as a Success Metric: Persistent Biases Under Full Randomization

We study user sentiment (reported via optional surveys) as a metric for ...

Please sign up or login with your details

Forgot password? Click here to reset