Is there any non parametricnonparametric version of ANOVA able to deal with multiple factors AND evaluate their interaction(s).?
If the interactions are really impossible to evaluate, is there any generalization of the Friedman test to more thatthan two factors?