Primary Submission Category: Sensitivity Analysis
A Calibrated Sensitivity Analysis for Weighted Disparity Decompositions
Authors: Andy Shen, Samuel Pimentel,
Presenting Author: Andy Shen*
Disparities in health or well-being experienced by racial and sexual minority groups can be difficult to study using the traditional exposure-outcome paradigm in causal inference, since potential outcomes in variables such as race or sexual minority status are challenging to interpret. Decomposition analysis addresses this gap by considering causal impacts on a disparity under interventions to other, intervenable exposures (e.g. socioeconomic factors) that may play a mediating role in the disparity. While invoking weaker assumptions than causal mediation approaches, decomposition analyses are often conducted in observational settings and require uncheckable assumptions that rule out unmeasured confounders. Leveraging weighting estimators for disparity decomposition, we develop a sensitivity analysis for unobserved confounders in studies of disparities using the marginal sensitivity model. We use the percentile bootstrap to construct valid confidence intervals for disparities and causal effects on disparities under given levels of confounding under mild conditions. We also explore amplifications that give insight into multiple confounding mechanisms. We illustrate our framework on a study examining disparities in youth suicide rates among sexual minorities using the Adolescent Brain Cognitive Development Study (ABCD). Supported by the National Science Foundation under Grant No. 2142146.