Our new R package makes it easy to report the percentage of effect sizes that a given effect is larger than for a given research field. So instead of reporting a 'medium' effect, you can say something like, "we found an effect that was larger than 42% of effect sizes reported in the field"
Reposted from
PsyArXivBot
A tutorial for calculating field-specific effect size distributions: http://osf.io/h368x/
Comments
(If 100 papers were published in 2023, each with effect sizes larger than 50% of papers published in 2022, then the field specific effect size distribution perhaps has a higher mean than we thought)
For the effect size guide (https://mgto.org/effectsizeguide), might be good to add that as an alternative/supplement approach. I know, I know, issues/debates about that whole benchmark thing, but people are going to do these things, might as well try to guide them to do better.