[R] "r squared" and anova for linear mixed-effects model

[Ben Bolker]

florencia bonatto <mfbonatto <at> yahoo.com.ar> writes:

> 
> 
> I have a linear
> mixed-effects model (from the package 
> nlme) with a random effect; Is there something like
> an "r squared" for the whole 
> model which I can state? I´d like to kown: How would I do anova for a linear
> mixed-effects model?
> 

  Unfortunately, this is not as simple as you would like it to be
(although depending on the simplicity of your model there *may* be
 a simple analogue).
  I have added a section to http://glmm.wikidot.com/faq that discusses
it and points you to various previous threads on the mailing list.