Hello, This is an "intuition" question from a "stats-user!=expert". I hope this is the right place to post it! I understand that lm() is quite sensitive to data with severe outliers. I imagine therefore that aov(), which uses lm(), is too. I have tried to hack aov() such that it calls rlm() which is much more robust against outliers. Even though the calling conventions for lm() and rlm() are quite similar, this just doesn't work: Error in terms.default(formula, data = data) : no terms component 1) Is such a modification feasible, and does it make sense to use rlm() (or similar) in an anova? 2) How to do this, or is there already such a function? Thanks in advance, RenE J.V. Bertin College de France/LPPA 11, place Marcelin Berthelot 75005 Paris, France _________________________________________________________________ -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._