Hi Eleftheria, If a non-parametric test doesn't sink your preferred hypothesis, it can offer some peace of mind with regard to scrupulous reviewers. If it does, it is wise to ponder the reliability of your results.
Jim On Fri, Jun 28, 2019 at 10:34 PM Eleftheria Dalmaris <edalma...@gmail.com> wrote: > > Dear all, > > I need to run the following ANOVA > > b <- aov(y ~ a +b*(factor(a)/b) + c + d + e) > > I'm checking the assumption of the normality of the residuals and then the > Levene's Test for Homogeneity of Variance. I first transform my data so the > normality assumption is met, but if I do that the Homogeneity of Variance > test is failing to be met. If I transform them again for the Levene's Test > to be OK, then the normality assumption is not. > > Is one of those two test more important than the other? Both of those > assumptions should be met? > > Is it bet to choose a kruskal.test and not an ANOVA since I'm facing those > kinds of issues? > > Thanks a lot, > Eleftheria > > [[alternative HTML version deleted]] > > ______________________________________________ > R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. ______________________________________________ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
