Non-parametric test equivalent to two-way anova?

Hi everyone. I have a question on non-parametric tests that replace anova tests.

In my first data set, I have scat density data grouped according to habitat type. I have used Kruskal-wallis instead of one-way anova since it is a non-parametric test which does not assume normal distribution.

However, I have another data set where scat density is also divided according to sex. In this case I would have two nominal variables (sex and habitat) so Kruskal-wallis would be unapplicable. Therefore, would there be a non-parametric test equivalent to two-way anova or any other statistical test recommendable for this case?


Super Moderator
If your design is a 2x2 ANOVA, then it would be appropriate to replace the original data points with their ranks and subsequently conduct the usual OLS parametric analysis on the ranks.