Causal Effect Estimation Methods

TL;DR

This paper explores the connection between causal graphical models and potential outcome models, demonstrating how common estimators can be derived from the graphical framework, with extensions to continuous variables.

Contribution

It establishes a link between two major causal inference frameworks and shows how popular estimators can be derived from causal graphical models.

Findings

Inverse probability weighting estimator can be derived from causal graphical models.

Doubly robust estimator can be obtained using causal graphical models.

Results can be extended to continuous variables.

Abstract

Relationship between two popular modeling frameworks of causal inference from observational data, namely, causal graphical model and potential outcome causal model is discussed. How some popular causal effect estimators found in applications of the potential outcome causal model, such as inverse probability of treatment weighted estimator and doubly robust estimator can be obtained by using the causal graphical model is shown. We confine to the simple case of binary outcome and treatment variables with discrete confounders and it is shown how to generalize results to cases of continuous variables.