I deal a lot of with non-parametric data. I know that TukeyHSD and Duncan test are suggested for ANOVA. What kind of post-hoc tests are appropriate for K-W and Friedman tests? I already use Wilcoxon–Mann–Whitney test for Kruskal-Wallis but it couldn't been applied for a Friedman test.
Are there other post-hoc test I may use? I know there is a Bonferrini correction, but it is criticized as too conservative.