On Smallest Effect Size of Interest: a Review and a Simplified Procedure

Methods for testing the smallest effect size of interest (SESOI)—including good-enough hypothesis testing, equivalence testing, minimum-effect testing, three-sided testing, and null-region frameworks—have been developed and increasingly advocated in psychology. Using a running example, this article reviews these SESOI-based methods, demonstrates their application, and highlights their similarities and differences in terms of their philosophical foundations and statistical procedures. Moreover, the occurrence and prevalence of inconclusive decisions in conditional equivalence testing are discussed, illustrating the potential consequences of applying different SESOI-based methods without recognizing their underlying connections. To facilitate the use of SESOI-based testing in psychological science and streamline the communication of statistical results across methods, a simplified procedure based on the null-region framework is recommend with two illustrative examples.