English

Sensitivity Analysis to Unobserved Confounding with Copula-based Normalizing Flows

Methodology 2025-08-13 v1 Machine Learning Machine Learning

Abstract

We propose a novel method for sensitivity analysis to unobserved confounding in causal inference. The method builds on a copula-based causal graphical normalizing flow that we term ρ\rho-GNF, where ρ[1,+1]\rho \in [-1,+1] is the sensitivity parameter. The parameter represents the non-causal association between exposure and outcome due to unobserved confounding, which is modeled as a Gaussian copula. In other words, the ρ\rho-GNF enables scholars to estimate the average causal effect (ACE) as a function of ρ\rho, accounting for various confounding strengths. The output of the ρ\rho-GNF is what we term the ρcurve\rho_{curve}, which provides the bounds for the ACE given an interval of assumed ρ\rho values. The ρcurve\rho_{curve} also enables scholars to identify the confounding strength required to nullify the ACE. We also propose a Bayesian version of our sensitivity analysis method. Assuming a prior over the sensitivity parameter ρ\rho enables us to derive the posterior distribution over the ACE, which enables us to derive credible intervals. Finally, leveraging on experiments from simulated and real-world data, we show the benefits of our sensitivity analysis method.

Keywords

Cite

@article{arxiv.2508.08752,
  title  = {Sensitivity Analysis to Unobserved Confounding with Copula-based Normalizing Flows},
  author = {Sourabh Balgi and Marc Braun and Jose M. Peña and Adel Daoud},
  journal= {arXiv preprint arXiv:2508.08752},
  year   = {2025}
}
R2 v1 2026-07-01T04:45:45.716Z