Justify Your Alpha

This project responds to the widely discussed proposal by Benjamin et al. (2018, Nature Human Behaviour) to redefine statistical significance by lowering the conventional alpha threshold to p ≤ .005. Rather than adopting a single fixed threshold, the collaborating authors argue that researchers should transparently report and justify the alpha level chosen for a given study, treating the significance threshold as a design decision that depends on context rather than a universal constant.

Sau-Chin Chen is a co-author of this large multi-author methodology project, which brought together statisticians and methodologists across many institutions to formulate a shared position on error control in study design.

The project’s significance lies in shifting the debate on statistical significance away from a search for a single “correct” threshold and toward a broader emphasis on transparency and justification in research design choices, a position that has since informed wider discussions of statistical practice.

Status: Stable — completed and archived on OSF.