I am using graph-pad prism 8, for analyzing one of my experiment which has 5 groups. The normality and Log normality test showed that data is not normally distributed, so I decided to use Kruskal-Wallis test. Further I want to perform Multiple comparison (between different group) so I used Dunn's multiple comparison test.

Is the choice of Kruskal-Wallis test right? After Kruskal-wallis test which test should be used as post-hoc test, Dunn's test or Mann-whitney test?

If someone can direct me to any reading material, article or book, would be greatly appreciated.

Thanks in advance!