From p-Values to Practical Relevance: An Introduction to Effect Sizes Through a Legal Psychological Example

Null-hypothesis significance testing and p-values are frequently criticized for their focus on detecting non-zero differences and their inability to provide evidence for the null hypothesis. In this article, we highlight how effect sizes, when meaningfully interpreted, can address these issues. Specifically, we argue that researchers should consider the smallest effect size of interest (SESOI) - the smallest effect size that yields practical or theoretical relevance. We propose several methods for estimating the SESOI and present a consensus study among Indonesian professionals which can be used to estimate the SESOI for child eyewitness testimony research. Results suggest that most Indonesian professionals consider one to two memory errors sufficient to take action, such as deeming testimony unreliable. We then showed how the SESOI, combined with confidence intervals, can be used data and power analyses (e.g., minimum-effect testing, equivalence testing). Finally, we emphasize that the practical relevance of an effect size should be carefully evaluated before making policy recommendations.