JIAB-2026-YG6K
A Unified Framework for Causal Mediation Analysis with Multiple Mediators
Published
Editor
Maya Henderson
Handling editor for this submission.
Assigned editor: Maya Henderson
Update the handling editor to rebalance workload or hand off the submission.
Abstract
We extend the potential-outcomes framework for mediation analysis to settings with multiple causally ordered mediators. Identification requires a no-unmeasured-confounding assumption for each mediator pathway; we derive sharp bounds when this fails and propose a sensitivity analysis applicable to both experimental and observational designs.
Reviews
2 itemsR1 Isabelle Dubois Submitted Jul 12, 2026
The revision is substantially clearer. The identification table and applied workflow resolve my concerns.
R2 Jack Thompson Submitted Jul 15, 2026
The new simulations address the misspecification concern. I recommend acceptance.
Files
2 manuscriptsReviewers
2 of 2 receivedDrag names to change the reviewer order for people who have not been invited yet.
Isabelle Dubois
Jack Thompson
Messages
2 messagesA Author Cover letter
We hope this framework will provide applied researchers with a unified toolkit for complex mediation questions.
A Author Re: Revision requested
Dear Editors,
The revised manuscript adds the requested identification table, expands the applied example, and includes new simulations for mediator-order misspecification.
Sincerely,
Kenji Yamamoto
Timeline
5 events
Published
Oct 11, 2026
In Production
Aug 17, 2026
Waiting for Reviewers
Jun 28, 2026
Waiting for Authors
May 29, 2026
Waiting for Reviewers
Apr 29, 2026
Status override
Override status only to correct a workflow state set in error. Notification actions should use Compose.