Practically Effective Adjustment Variable Selection in Causal Inference

TL;DR

This paper introduces a practical method for selecting adjustment variables in causal inference to improve accuracy, especially when data is limited, by providing criteria and algorithms applicable to DAGs and CPDAGs.

Contribution

It proposes new criteria and algorithms for variable selection in causal effect estimation, addressing limitations of existing methods and extending applicability to CPDAGs.

Findings

The method prevents accuracy degradation in causal effect estimation.

Theoretical proof of causal effect computation in CPDAGs.

Demonstrated utility on real and artificial data.

Abstract

In the estimation of causal effects, one common method for removing the influence of confounders is to adjust the variables that satisfy the back-door criterion. However, it is not always possible to uniquely determine sets of such variables. Moreover, real-world data is almost always limited, which means it may be insufficient for statistical estimation. Therefore, we propose criteria for selecting variables from a list of candidate adjustment variables along with an algorithm to prevent accuracy degradation in causal effect estimation. We initially focus on directed acyclic graphs (DAGs) and then outlines specific steps for applying this method to completed partially directed acyclic graphs (CPDAGs). We also present and prove a theorem on causal effect computation possibility in CPDAGs. Finally, we demonstrate the practical utility of our method using both existing and artificial data.