Bert Gunter <gunter.berton <at> gene.com> writes: > > This almost certainly has nothing to do with mixed effects models per > se. x1 and x2 are probably ( if what you see is not due to a few > unusual values) correlated, so the standard decomposition into main > and interaction effects does not have the usual meaningful > interpretation (e,g, that you would get from orthogonal regressors).
See also http://stats.stackexchange.com/questions/5450/ what-if-interaction-wipes-out-my-direct-effects-in-regression/ (URL broken to make gmane happy) ______________________________________________ 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.
