John: Linear models can have different covariance structures to accommodate certain dependencies in the data. The functions for data analysis in the lme4 package are flexible and can be structured as such. For example, when random intercepts only are included, this is akin to compound symmetry, but when random slopes are also introduced in a repeated measures mlm using lmer(), then you have a more general covariance matrix.
This isn't really an R question and most of the texts on mlm deal with this issue directly. Harold -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of John Christie Sent: Monday, March 20, 2006 7:54 AM To: [email protected] Subject: [R] does lme repeated measures require sphericity? I haven't been able to find an answer on this that's direct, only implied. In several places I have read that when people asked for sphericity tests they were guided toward lme or mlm models. But, there is no direct indication that the lme method is not subject to the sphericity assumption. In fact, it seems like it should be. Its just a linear model that handles random and fixed effects. I imagine most other parametric assumptions hold true. ______________________________________________ [email protected] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html ______________________________________________ [email protected] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
