A Critical Evaluation of the FBST ev for Bayesian Hypothesis Testing
Alexander Ly, Eric‐Jan Wagenmakers · Computational Brain & Behavior · 2021
Abstract The “Full Bayesian Significance Teste-value”, henceforth FBSTev, has received increasing attention across a range of disciplines including psychology. We show that the FBSTevleads to four problems: (1) the FBSTevcannot quantify evidence in favor of a null hypothesis and therefore also cannot discriminate “evidence of absence” from “absence of evidence”; (2) the FBSTevis susceptible to sampling to a foregone conclusion; (3) the FBSTevviolates the principle of predictive irrelevance, such that it is affected by data that are equally likely to occur under the null hypothesis and the alternative hypothesis; (4) the FBSTevsuffers from the Jeffreys-Lindley paradox in that it does not include a correction for selection. These problems also plague the frequentistp-value. We conclude that although the FBSTevmay be an improvement over thep-value, it does not provide a reasonable measure of evidence against the null hypothesis.