see for example https://daniellakens.blogspot.com/2014/12/observed-power-and-what-to-do-if-your.html

Best Matthias

Can you tell what the reviewer means by post hoc power? The answer you seek will likely depend on the answer.

Hello Frost,

The concept of post hoc power (e.g., what would be the statistical power to detect an effect equal in size to that observed in the data set, with the same sample size and same alpha level) is not very productive. In the case of having found non-significant results, you know (of course) that you had insufficient power for those results (given their magnitude). In the case of having significant results, you know you had power of at least .50 (and it was sufficient for statistical significance). The real question is, what is the magnitude of the effect, and is it large enough to be considered noteworthy on its own? As well, the likelihood that the effect size observed in the data set is exactly that of the population is nil.

Having said all that, I understand that authors usually want to honor reviewer recommendations so as to enhance the likelihood of manuscript acceptance and publication. However, you don't have to abide by recommendations that are questionable.

Check these sources for a fuller explanation of why post hoc power is more often a bad, not good, idea:

Levine, M., & Ensom, M. H. H. (2002). Post hoc power: An idea whose time has passed? Pharmacotherapy, 21(4), 405-409. doi:

10.1592/phco.21.5.405.34503

Yuan, K. & Maxwell, S. (2005). On the post hoc power in testing mean differences. Journal of Educational and Behavioral Statistics, 30(2), 141-167. doi: 10.3102/10769986030002141

Gelman, A. (2018, July 9). Don't calculate post-hoc power using observed estimate of effect size. Annals of Surgery. doi: 10.1097/SLA.0000000000002908

https://journals.lww.com/annalsofsurgery/Citation/publishahead/Don_t_Calculate_Post_hoc_Power_Using_Observed.95527.aspx

Good luck with your work!

thx everyone!