Is post-hoc power analysis redundant too in the case of a *significant* result?
Posted by Giulia, at stats.stackexchange.com,
I've read that post-hoc power analysis is useless when the result of a test is statistically insignificant, which I understand (p…