I'm a first-year BSc Biology student working on my end-of-module stats project, so that's the expected level of my work.
I need to find associations between hindlimb length and frog locomotor modes (e.g. tree jumping). I have an overall sample size of 164, but when divided into the different locomotor modes, my smallest sample size is 9.
Some of the hindlimb lengths data, such as the lengths associated with the burrower-walker-hopper group, have a very significant skew (shapiro wilk: W = 0.90326, p-value = 0.00648). This is fine for performing Kruskal-Wallis, but I can't find a post-hoc test that works for nominal data. Is there one that I'm not considering? Or alternatively, can I apply the central limit theorem to this scenario?
I am working in R Studio and Excel.
Thanks!
