See the nlme library and accompanying book by Pinheiro and Bates. The lme function is appropriate for linear mixed models with normal errors, and the 'weights' argument provides flexible methods for modeling error heterogeneity.
hth, Kingsford On Mon, Nov 2, 2009 at 12:10 PM, jinyan fan <fanjin...@yahoo.com> wrote: > Dear Colleagues, > > I am running a three-way mixed-design ANOVA, with one manipulated IV, a group > membership IV, and pre-/post- within subject factor. I am interested in the > three-way interaction effect. The regular ANVOA is problematic, because (a) > the sample sizes are very unbalanced, due to the group IV, (b) the > homogeneity of variance is violated, and (c) the homogeneity of covariance is > also violated. I wonder if R has a procedure to address this issue, either by > some sort of Robust ANOVA procedure, or by some alternative tests? Thanks a > lot. > > Jinyan Fan > > Assistant Professor > Department of Psychology > Hofstra University > Hempstead, NY 11550 > Work: 516-463-6349 > > > > [[alternative HTML version deleted]] > > > ______________________________________________ > R-help@r-project.org mailing list > 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 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.
