Leveraging Causal Graphs for Blocking in Randomized Experiments

TL;DR

This paper introduces an efficient algorithm that leverages causal graphs to select covariates for blocking in randomized experiments, thereby reducing variance in causal effect estimates.

Contribution

It presents a novel method to identify stable covariate sets for blocking using causal graphs, improving the precision of causal inference in experiments.

Findings

Algorithm effectively minimizes variance of causal estimates.

Utilizes causal graph structure for covariate selection.

Applicable to general semi-Markovian causal models.

Abstract

Randomized experiments are often performed to study the causal effects of interest. Blocking is a technique to precisely estimate the causal effects when the experimental material is not homogeneous. It involves stratifying the available experimental material based on the covariates causing non-homogeneity and then randomizing the treatment within those strata (known as blocks). This eliminates the unwanted effect of the covariates on the causal effects of interest. We investigate the problem of finding a stable set of covariates to be used to form blocks, that minimizes the variance of the causal effect estimates. Using the underlying causal graph, we provide an efficient algorithm to obtain such a set for a general semi-Markovian causal model.