On Smallest Effect Size of Interest: Its Development and a Simplified Procedure

Methods for testing the smallest effect size of interest (SESOI)—such as good-enough hypothesis-testing and equivalence tests—have been developed and increasingly advocated in psychology. This article reviews these frequentist methods from philosophical and statistical perspectives, highlighting their conceptual foundations and procedural similarities. Although these approaches have gained popularity in the recent decade, the importance of SESOI was already emphasized by Neyman and Pearson within their original framework of hypothesis testing. When null hypothesis significance testing (NHST) is conducted following the Neyman–Pearson framework, these tests yield identical results as their underlying logic is fundamentally the same once simplified. Building on this observation, the article proposes a simplified procedure involving that defines the null hypothesis as a negligible effect, the alternative as a non-negligible effect, and sets error rates according to theoretical expectations. This simplification helps avoid unnecessary computational complexity and prevents inconsistent statistical decisions that may arise from conditional equivalence testing, where an equivalence test is conducted only after a nonsignificant result from NHST.