Dear colleague working with product-moment correlation, that is, dear all of you!
Transforming properly an estimate of product-moment correlation (r) to the scale of Cohen’s d is important when willing to qualitatively evaluate the magnitude of r in terms of “very small”, “small”, “medium”, “large/high” “very large/high”, and “”huge” because the traditional thresholds are based on the scale of d. As far as I know, we have missed the exact formulae for the polytomous and continuous settings (if there are, I'm happy to know...) even if we have one for the dichotomous settings in Cohen's classic book. Good news: I derived two general forms of the transformation that give exact correspondence also in polytomous and continuous settings, and I also studied two short-cut formulae which give quite robust approximation. It would be appreciated if you'd spend some minutes with my preprint (Preprint R effect size and Generalized Cohen d 2024 with name
) and give some feedback.- Badges
- Science topic
- Similar topics
- Psychology