Regression analysis of unmeasured confounding

DOI10.1515/EM-2019-0028MaRDI QIDQ2197498zbMATH OpenOpenAlexFDO

Authors Brian Knaeble, Braxton Osting, Mark A. Abramson

Publication date 31 August 2020

Published in Epidemiologic Methods (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1908.08596

zbMATH Keywords

optimizationcorrelationmodel uncertaintypropensity

Mathematics Subject Classification ID

General nonlinear regression (62J02) Applications of statistics to actuarial sciences and financial mathematics (62P05) General biostatistics (92B15)

Abstract: When studying the causal effect of $x$ on $y$, researchers may conduct regression and report a confidence interval for the slope coefficient $$. This common confidence interval provides an assessment of uncertainty from sampling error, but it does not assess uncertainty from confounding. An intervention on $x$ may produce a response in $y$ that is unexpected, and our misinterpretation of the slope happens when there are confounding factors $w$. When $w$ are measured we may conduct multiple regression, but when $w$ are unmeasured it is common practice to include a precautionary statement when reporting the confidence interval, warning against unwarranted causal interpretation. If the goal is robust causal interpretation then we can do something more informative. Uncertainty in the specification of three confounding parameters can be propagated through an equation to produce a confounding interval. Here we develop supporting mathematical theory and describe an example application. Our proposed methodology applies well to studies of a continuous response or rare outcome. It is a general method for quantifying error from model uncertainty. Whereas confidence intervals are used to assess uncertainty from unmeasured individuals, confounding intervals can be used to assess uncertainty from unmeasured attributes.

Recommendations

• Causal inference accounting for unobserved confounding after outcome regression and doubly robust estimation
• Assessing the Sensitivity of Regression Results to Unmeasured Confounders in Observational Studies
• Sensitivity analysis of unmeasured confounding for the causal risk ratio by applying marginal structural models
• Confounding and collapsibility in causal inference
• Sensitivity Analysis for Unmeasured Confounding in Meta-Analyses

Cites work

• scientific article; zbMATH DE number 5060482 (Why is no real title available?)
• A new criterion for confounder selection
• Causal inference in statistics: an overview
• Design of observational studies
• External validity: from do-calculus to transportability across populations
• Generalized Cornfield conditions for the risk difference
• Observational studies.
• Odds are the sign is right
• Reversals of Least-Square Estimates and Model-Invariant Estimation for Directions of Unique Effects
• The central role of the propensity score in observational studies for causal effects
• The sensitivity of linear regression coefficients' confidence limits to the omission of a confounder

Cited in

(4)

• Regression analysis of experiments with complex confounding patterns guided by the alias matrix.
• Regression analysis under link violation
• The sign of the unmeasured confounding bias under various standard populations
• An asymptotic threshold of sufficient randomness for causal inference

Describes a project that uses

Uses Software

• DOS
• GitHub
• ElemStatLearn

This page was built for publication: Regression analysis of unmeasured confounding

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2197498)