Comment: Reflections on the Deconfounder

DOI10.1080/01621459.2019.1689138zbMATH Open1428.62056arXiv1910.08042OpenAlexW2998143050WikidataQ126395856 ScholiaQ126395856MaRDI QIDQ5208063FDOQ5208063

Authors: Alexander D'Amour 

Publication date: 15 January 2020

Published in: Journal of the American Statistical Association (Search for Journal in Brave)

Abstract: The aim of this comment (set to appear in a formal discussion in JASA) is to draw out some conclusions from an extended back-and-forth I have had with Wang and Blei regarding the deconfounder method proposed in "The Blessings of Multiple Causes" [arXiv:1805.06826]. I will make three points here. First, in my role as the critic in this conversation, I will summarize some arguments about the lack of causal identification in the bulk of settings where the "informal" message of the paper suggests that the deconfounder could be used. This is a point that is discussed at length in D'Amour 2019 [arXiv:1902.10286], which motivated the results concerning causal identification in Theorems 6--8 of "Blessings". Second, I will argue that adding parametric assumptions to the working model in order to obtain identification of causal parameters (a strategy followed in Theorem 6 and in the experimental examples) is a risky strategy, and should only be done when extremely strong prior information is available. Finally, I will consider the implications of the nonparametric identification results provided for a narrow, but non-trivial, set of causal estimands in Theorems 7 and 8. I will highlight that these results may be even more interesting from the perspective of detecting causal identification from observed data, under relatively weak assumptions about confounders.

Full work available at URL: https://arxiv.org/abs/1910.08042

zbMATH Keywords

causal inferenceunobserved confoundingunconfoundedness

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) Causal inference from observational studies (62D20)

Cites Work

• The Hardness of Conditional Independence Testing and the Generalised Covariance Measure
• Bayesian robust inference of sample selection using selection-\(t\) models
• Sample Selection Bias as a Specification Error
• When is a sensitivity parameter exactly that?
• Title not available (Why is that?)
• Bias attenuation results for nondifferentially mismeasured ordinal and coarsened confounders
• Identifying causal effects with proxy variables of an unmeasured confounder
• Split-door criterion: identification of causal effects through auxiliary outcomes
• Bayesian approaches for missing not at random outcome data: the role of identifying restrictions
• Flexible Sensitivity Analysis for Observational Studies Without Observable Implications
• Making Sense of Sensitivity: Extending Omitted Variable Bias

Cited In (1)

• Identifying Effects of Multiple Treatments in the Presence of Unmeasured Confounding

Uses Software

• GeneralisedCovarianceMeasure

This page was built for publication: Comment: Reflections on the Deconfounder

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5208063)