Using Heteroscedasticity for Causal Discovery under Hidden Confounding
Abstract
The presence of hidden confounding in observational studies poses a fundamental challenge to the identification and estimation of causal structures encoded by directed acyclic graphs (DAGs). While instrumental variables (IVs) have been widely used to address hidden confounding, most existing methods require valid instruments, which may be unavailable in practice. In this paper, we develop a novel framework for causal discovery and inference in DAGs that accommodates both hidden confounding and invalid IVs by exploiting heteroscedasticity. We establish a novel linear characterization of leaf nodes through heteroscedasticity-induced relationships between cross-covariance and self-covariance vectors, enabling recursive recovery of the causal order without requiring any candidate IV to be valid. Given the identified order, we then sequentially construct surrogate IVs to identify the causal effects and directions, thereby recovering the entire DAG. We also propose a data-adaptive procedure for estimating the DAG and causal effects. We show that the proposed method consistently recovers the underlying DAG and yields asymptotically normal estimators of causal effects. Extensive simulation studies demonstrate substantial improvements over existing methods. We further illustrate the practical utility of the proposed method through an application to gene regulatory network analysis.
Related
- ProximalFM: Amortized Proximal Causal Inference under Hidden Confounding
- Spatial causal inference in the presence of unmeasured confounding and interference
- Causal Inference with Unobserved Confounding: A Mixture Learning Perspective
- Causal Path Analysis from Perturbational and Population-Scale Single-Cell Data with Multiscale Confounding and Measurement Error
- Proximal Balancing for Causal Effect Estimation under Unmeasured Confounding
BibTeX
@misc{zou2026using,
title = {{Using Heteroscedasticity for Causal Discovery under Hidden Confounding}},
author = {Jing Zou and Wei Li and Wei Lin},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.21219},
eprint = {2610.21219},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.21219},
}
RIS
TY - UNPB
TI - Using Heteroscedasticity for Causal Discovery under Hidden Confounding
AU - Jing Zou
AU - Wei Li
AU - Wei Lin
PY - 2026
DA - 2026/10/15
PB - arXiv
DO - 10.48550/arxiv.2610.21219
UR - https://arxiv.org/abs/2610.21219
AB - The presence of hidden confounding in observational studies poses a fundamental challenge to the identification and estimation of causal structures encoded by directed acyclic graphs (DAGs). While instrumental variables (IVs) have been widely used to address hidden confounding, most existing methods require valid instruments, which may be unavailable in practice. In this paper, we develop a novel framework for causal discovery and inference in DAGs that accommodates both hidden confounding and invalid IVs by exploiting heteroscedasticity. We establish a novel linear characterization of leaf nodes through heteroscedasticity-induced relationships between cross-covariance and self-covariance vectors, enabling recursive recovery of the causal order without requiring any candidate IV to be valid. Given the identified order, we then sequentially construct surrogate IVs to identify the causal effects and directions, thereby recovering the entire DAG. We also propose a data-adaptive procedure for estimating the DAG and causal effects. We show that the proposed method consistently recovers the underlying DAG and yields asymptotically normal estimators of causal effects. Extensive simulation studies demonstrate substantial improvements over existing methods. We further illustrate the practical utility of the proposed method through an application to gene regulatory network analysis.
ER -