I'm analysing about 20 lab-results of three different groups either with ANOVA, Kruskal–Wallis test or chi-squared. I correct the resulting p-values for multiple testing using the Benjamini/Hochberg procedure. I create post-hoc tests for significant variables using Tukey, Dunn and chi-squared.
My questions are: do I have to correct the post-hoc tests for multiple testing (which is the default in most statistics software) and if so, how do I adjust in qualitative variables/chi-squared?
A statistician once told my colleague it was not appropriate to adjust p-values in post-hoc tests.