Dear all,
I am trying to get the emperical bayes estimates together with their
standard errors out of lme4. Up to now I have used MLwiN to get these
estimates. I have fitted the following - very simple - model, just to
find out how this works.
test<-lmer(y~(1|subject),data,REML=F)
ranef(test,postVar=T)
str(ranef(test,postVar=T)
If I use the formulation of the emperical bayes estimates and their
standard errors from Longford (1993), this comes down to using a
shrinkage factor based on the level 1 and 2 variance and the sample size
of the subject at hand (lambda=var(u)/{var(u)+var(e)/n}). Using the
formulas of Longford (1993), I get exactly what R is getting also. If I
fit the same model in MLwiN, the EB-estimates are the same, the
se(EB-estimates) are, however, different. Does anyone know where this
difference comes from?
all the best,
Maria Schipper
[[alternative HTML version deleted]]
______________________________________________
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.
