| Title: | Robust estimation and model selection for the controlled direct effect with unmeasured mediator-outcome confounders |
| Journal: | Annals of the Institute of Statistical Mathematics |
| Published: | 2 Jun 2026 |
| DOI: | https://doi.org/10.1007/s10463-026-00985-w |
Publication 19801
WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
Abstract
Controlled Direct Effect (CDE) is one of the causal estimands used to evaluate both exposure and mediation effects on an outcome. When there are unmeasured confounders existing between the mediator and the outcome, the ordinary identification assumption does not work. In this manuscript, we consider an identification condition to identify CDE in the presence of unmeasured confounders. The key assumptions are: 1) the random allocation of the exposure, and 2) the existence of instrumental variables directly related to the mediator. Under these conditions, we propose a novel doubly robust estimation method, which works well if either the propensity score model or the baseline outcome model is correctly specified. Additionally, we propose a Generalized Information Criterion (GIC)-type model selection criterion for CDE that ensures model selection consistency. Our proposed procedure and related methods are applied to both simulation and real datasets to confirm the performance of these methods.</p>
5 Authors
- Shunichiro Orihara
- Shinpei Imori
- Kosuke Morikawa
- Atsushi Goto
- Masataka Taguri
1 Application
| Application ID | Title |
|---|---|
| 71079 | Mendelian randomization study to investigate diabetes medications and diabetes-related phenotypes in relation to risk of cancer and mortality. |
Enabling scientific discoveries that improve human health