Nathaniel Street wrote:
>
>
> [questions about testing significance of random effects]
>
>
Pinheiro and Bates 2000 is the standard reference for this kind of question.
The basic answer is that the Likelihood Ratio Test (implemented by anova())
is a reasonable approach (and recommended by PB 2000), although it
is conservative because the estimated value of the parameter is on
the boundary of its feasible range (i.e., the variance cannot be less than
0).
So ... read PB 2000, and fit models mod1 and mod2 with and without
the relevant random effect, and then do anova(mod1,mod2).
cheers
Ben Bolker
--
View this message in context:
http://www.nabble.com/significance-for-a-random-effect-in-Mixed-Model-ANOVA-tf4623515.html#a13252487
Sent from the R help mailing list archive at Nabble.com.
______________________________________________
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.
