TITLE:
Commentary on Grice et al., 2020: A Critical Examination of Adjusted Effect Sizes (rh and PCCh) and Comparisons across Psychology and Medicine
AUTHORS:
James W. Grice, Paul T. Barrett, Mateo Martin
KEYWORDS:
Effect Size, Correlation, Percent Correct Classifications, Epidemiology, Relative Risk, Medical Intervention Studies
JOURNAL NAME:
Open Journal of Applied Sciences,
Vol.15 No.9,
September
15,
2025
ABSTRACT: In their recent paper, Persons as Effect Sizes, Grice and colleagues advocated for a technique that adjusts statistical effect sizes upwards for intervention studies with low base rates like those found in medical and epidemiological research. This technique was developed by Ferguson in part as an aid for comparing medical effect size magnitudes to those reported in psychological studies. Herein we challenge the rationale behind this technique by particularly examining Ferguson’s proposed distinction between “hypothesis-relevant” and “hypothesis-irrelevant” cases which lies at the heart of his method. We then advocate against using this technique and instead demonstrate graphical and numerical procedures, many of which are well known, that are rooted in the unadjusted raw data and that are consistent with the person-centered approach toward evaluating effect sizes. Finally, we explore the pitfalls associated with comparing effect sizes from medical studies, which often have miniscule base rates, to those found in psychological studies and conclude that such comparisons should be avoided.