I have a scenario where I need to compare multiple groups ($N > 2$) with non-Gaussian distributions. After performing the Kruskal-Wallis test and finding a significant p-value ($p < \alpha$), I proceeded to conduct pairwise comparisons using the Dwass-Steel-Crichtlow-Fligner (DSCF) post-hoc test to identify specific group differences.
The DSCF test has generated multiple p-values for the pairwise comparisons. My question is whether I need to apply additional corrections for the familywise error rate (FWER), such as Bonferroni, Holm, or Hochberg adjustments, to these p-values.
There was a mention about this in the question Differences between Dwass-Steel-Critchlow-Fligner and Mann-Whitney U-test for a post-hoc pairwise analysis that "The Dwass-Steel-Crichtlow-Fligner test also addresses (1) and (2), but has a specific approach to controlling the familywise error rate (FWER) built in." said by Alexis, but i do not have enough reputation to comment his answer, so posted a separate question.
Could someone please clarify whether additional FWER corrections are necessary after the DSCF test, or if the test inherently controls for these errors?
Thank you!
