Critical Values Robust to P-Hacking

This paper develops a model of p-hacking by researchers. It then gives critical values that correct the inflated type 1 error rates caused by p-hacking. For a two-sided z-test with significance level of 5%, the robust critical value is 2.57 (instead of 1.96).

June 2022 · Adam McCloskey, Pascal Michaillat

Persistence of False Paradigms in Low-Power Sciences

This paper develops a model of science. It shows that due to homophily in tenure decisions, false paradigms may persist when a science has low power. Low power may come from lack of evidence, or from reluctance to base tenure decisions on available evidence.

December 2018 · George Akerlof, Pascal Michaillat