loading page

Model selection isn't causal inference
  • Suchinta Arif,
  • M. Aaron MacNeil
Suchinta Arif
Dalhousie University
Author Profile
M. Aaron MacNeil
Dalhousie University
Author Profile

Abstract

Ecologists often rely on observational data to understand causal relationships. Although observational causal inference methodologies exist, model selection based on information criterion (e.g., AIC) remains a common approach used to understand ecological relationships. However, such approaches are meant for predictive inference and is not appropriate for drawing causal conclusions. Here, we highlight the distinction between predictive and causal inference and show how model selection techniques can lead to biased causal estimates. Instead, we encourage ecologists to apply the backdoor criterion, a graphical rule that can be used to determine causal relationships across observational studies.
03 Feb 2022Submitted to Ecology Letters
04 Feb 2022Submission Checks Completed
04 Feb 2022Assigned to Editor
09 Feb 2022Reviewer(s) Assigned
14 Mar 2022Review(s) Completed, Editorial Evaluation Pending
21 Mar 2022Editorial Decision: Revise Major
04 May 20221st Revision Received
06 May 2022Submission Checks Completed
06 May 2022Assigned to Editor
07 May 2022Review(s) Completed, Editorial Evaluation Pending
08 May 2022Editorial Decision: Accept