Discussion of "Data-driven confounder selection via Markov and Bayesian networks" by Jenny H\"aggstr\"om

TL;DR

This paper discusses the importance of accurate causal effect estimation, introduces a partially identified approach for colliders, and highlights the advantages of double robustness in high-dimensional confounding scenarios.

Contribution

It proposes a new partially identified causal inference method for colliders and emphasizes the significance of double robustness and high-dimensional confounder handling.

Findings

Partially identified approach for colliders

Double robustness offers under-appreciated advantages

High-dimensional confounder analysis is promising

Abstract

In this discussion we consider why it is important to estimate causal effect parameters well even they are not identified, propose a partially identified approach for causal inference in the presence of colliders, point out an under-appreciated advantage of double robustness, discuss the relative difficulty of independence testing versus regression, and finally commend H\"aggstr\"om for her exploration of causal inference with high-dimensional confounding, while making a call for further research in this same vein.