Confounding Equivalence in Causal Inference
Methodology
2012-03-19 v1 Artificial Intelligence
Abstract
The paper provides a simple test for deciding, from a given causal diagram, whether two sets of variables have the same bias-reducing potential under adjustment. The test requires that one of the following two conditions holds: either (1) both sets are admissible (i.e., satisfy the back-door criterion) or (2) the Markov boundaries surrounding the manipulated variable(s) are identical in both sets. Applications to covariate selection and model testing are discussed.
Cite
@article{arxiv.1203.3505,
title = {Confounding Equivalence in Causal Inference},
author = {Judea Pearl and Azaria Paz},
journal= {arXiv preprint arXiv:1203.3505},
year = {2012}
}
Comments
Appears in Proceedings of the Twenty-Sixth Conference on Uncertainty in Artificial Intelligence (UAI2010)