Dear David and Gilles,
Currently, Anova() in the car package doesn't handle lmer objects. I'm
working on a default method for Anova that will work with objects that
respond to coef() and vcov() and produce Wald-like tests. This should
almost work for lmer objects, replacing coef() with fixef(), bu
The answer to your first question is "yes, the order does make a
difference." I have not worked with lmer, but the standard anova applied
to lm() will provide what are called Type I sums of squares. Each effect
is adjusted for all prior effects.
Look at John Fox's car package. I don't know if i
Dear R user
I have 2 problems with lmer.
The statistical consultance service of my university has recomended to me to
expose those problems here.
Sorry for this quite long message.
Your help will be greatly appreciated...
Gilles San Martin
1) anova()
I fit a first model :
model1 <- lmer(eclw
3 matches
Mail list logo
