[R] Nonparametric generalization of ANOVA

[Ben Bolker]

blue sky <bluesky315 <at> gmail.com> writes:


   I almost certainly shouldn't feed the trolls, but:

1. ?kruskal.test (listed in "see also" in ?wilcox.test)
2. One of the disadvantages of nonparametric tests is that it is
in general difficult to generalize them to analogues of arbitrarily
complex linear mixed models.
3. You might look for "multi-reponse permutation procedures" (MRPP).