Bayesian Inference and Partial Identification in Multi-Treatment Causal Inference with Unobserved Confounding

Jiajing Zheng, Alexander D’Amour, Alexander Franks
Proceedings of The 25th International Conference on Artificial Intelligence and Statistics, PMLR 151:3608-3626, 2022.

Abstract

In causal estimation problems, the parameter of interest is often only partially identified, implying that the parameter cannot be recovered exactly, even with infinite data. Here, we study Bayesian inference for partially identified treatment effects in multi-treatment causal inference problems with unobserved confounding. In principle, inferring the partially identified treatment effects is natural under the Bayesian paradigm, but the results can be highly sensitive to parameterization and prior specification, often in surprising ways. It is thus essential to understand which aspects of the conclusions about treatment effects are driven entirely by the prior specification. We use a so-called transparent parameterization to contextualize the effects of more interpretable scientifically motivated prior specifications on the multiple effects. We demonstrate our analysis in an example quantifying the effects of gene expression levels on mouse obesity.

Cite this Paper

BibTeX
@InProceedings{pmlr-v151-zheng22a, title = { Bayesian Inference and Partial Identification in Multi-Treatment Causal Inference with Unobserved Confounding }, author = {Zheng, Jiajing and D'Amour, Alexander and Franks, Alexander}, booktitle = {Proceedings of The 25th International Conference on Artificial Intelligence and Statistics}, pages = {3608--3626}, year = {2022}, editor = {Camps-Valls, Gustau and Ruiz, Francisco J. R. and Valera, Isabel}, volume = {151}, series = {Proceedings of Machine Learning Research}, month = {28--30 Mar}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v151/zheng22a/zheng22a.pdf}, url = {https://proceedings.mlr.press/v151/zheng22a.html}, abstract = { In causal estimation problems, the parameter of interest is often only partially identified, implying that the parameter cannot be recovered exactly, even with infinite data. Here, we study Bayesian inference for partially identified treatment effects in multi-treatment causal inference problems with unobserved confounding. In principle, inferring the partially identified treatment effects is natural under the Bayesian paradigm, but the results can be highly sensitive to parameterization and prior specification, often in surprising ways. It is thus essential to understand which aspects of the conclusions about treatment effects are driven entirely by the prior specification. We use a so-called transparent parameterization to contextualize the effects of more interpretable scientifically motivated prior specifications on the multiple effects. We demonstrate our analysis in an example quantifying the effects of gene expression levels on mouse obesity. } }
Endnote
%0 Conference Paper %T Bayesian Inference and Partial Identification in Multi-Treatment Causal Inference with Unobserved Confounding %A Jiajing Zheng %A Alexander D’Amour %A Alexander Franks %B Proceedings of The 25th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2022 %E Gustau Camps-Valls %E Francisco J. R. Ruiz %E Isabel Valera %F pmlr-v151-zheng22a %I PMLR %P 3608--3626 %U https://proceedings.mlr.press/v151/zheng22a.html %V 151 %X In causal estimation problems, the parameter of interest is often only partially identified, implying that the parameter cannot be recovered exactly, even with infinite data. Here, we study Bayesian inference for partially identified treatment effects in multi-treatment causal inference problems with unobserved confounding. In principle, inferring the partially identified treatment effects is natural under the Bayesian paradigm, but the results can be highly sensitive to parameterization and prior specification, often in surprising ways. It is thus essential to understand which aspects of the conclusions about treatment effects are driven entirely by the prior specification. We use a so-called transparent parameterization to contextualize the effects of more interpretable scientifically motivated prior specifications on the multiple effects. We demonstrate our analysis in an example quantifying the effects of gene expression levels on mouse obesity.
APA
Zheng, J., D’Amour, A. & Franks, A.. (2022). Bayesian Inference and Partial Identification in Multi-Treatment Causal Inference with Unobserved Confounding . Proceedings of The 25th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 151:3608-3626 Available from https://proceedings.mlr.press/v151/zheng22a.html.

Related Material