Translational risks when the smallest effect size of interest leaves the lab for legal decision-making

The smallest effect size of interest (SESOI) makes stakes explicit, guides study design, and supports minimum-effect and equivalence tests. In this commentary, I treat SESOI as a context-dependent benchmark for research (design, analysis, interpretation) and outline possible risks that could arise when research benchmarks leave the lab for real-world legal decision-making. Three specific concerns follow. First, model estimates ≠ case decisions: estimates summarize patterns in data and remain probabilistic; they cannot decide specific cases. Second, SESOI is a value-laden choice: its value depends on construct, measure, population, and stakes, so different views can yield different thresholds, creating room for misplaced confidence when used in practice. Third, benchmarks risk turning into cutoffs: once formalized, SESOI can be treated as a decision boundary, create an illusion of certainty and a return to dichotomized thinking. I conclude that SESOI is especially well suited to research questions about average differences and population-level decisions, but legal adjudication concerns the evaluation of evidence in particular cases under burden of proof and asymmetric error costs. Responsible translation therefore requires clear communication of scope, remaining uncertainty, and the value judgments embedded in any threshold.