Dear colleague working with effect sizes!
For your use (and entertainment, maybe) I developed formulae for r effect size, Cohen's d and Cohen's f that transform the estimate to the metrics of a common language estimator of effect size. That is, if you have a correlation or an estimate by Cohen's d, you can turn this value to a probability that a case from a higher-ranked category (such as a person with a higher educational level or SES level) scores higher in a metric variable (such as in an attitude scale or achievement test). Consequently, the correlation means something practical. Referring to McGrath's and Wongs (2002) classic example: a probability of 0.92 means that in any random pairing of young adult males and females, in 92 out of 100 blind dates, the male will be taller than the female. Now, we can transform r effect size to this kind of probability linked with common language effect size.
Be free to glance at or dive in the preprint below. Any comments are warmly welcomed.
Preprint Common language intrerpretation of r effect size, Cohen's d,...
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts