glmmPQL: Confidence intervals for predicted values

Tue, Nov 1, 2011 2:26 AM

Dear all,

Is there a straightforward way to get confidence intervals for the predicted values in a glmmPQL model?

I?ve tried the following example:

library(MASS)
s1=glmmPQL(y ~ trt + I(week > 2), random = ~ 1 | ID,
                family = binomial, data = bacteria)


predict(s1,data.frame(ID="X01",trt="placebo",week=4),se.fit=T)  #se.fit doesn?t work

confint(s1) # produces output but it?s not clear what is actually returned

I also thought of trying boot.ci from the boot library, but I really would need help on that.

Many thanks in advance for any suggestions!

Best wishes,
Christoph

##
(running R 2.13.2 on Windows 7 64-bit).

Dr. Christoph Scherber
Georg-August-University Goettingen
Department of Crop Science
Agroecology
Grisebachstrasse 6
37077 Goettingen
tel +49 (0)551 39 8807
fax +49 (0)551 39 8806
www.gwdg.de/~cscherb1