Dear listers,
I would like to get the bootstrap estimates form my lme model.
I have an HLM (multillevel) 2-level model with the dichotomous outcome. I
used glmmPQL procedure. However I have a problem since I have a rather
unbalanced proportion (90-99% of events, i.e. ones and only 1-10% of
nonevents - zeros) although sample sizes are not that small between 500 and
1000 I get pretty weird and large SE. Nothing really surprising considering
the facts - therefore one of the solutions would be to get the bootstrap
estimates of the SE.
How could that be acheived? I checked the boot command in the package
"boot", however I'm not really sure how to tell the him that
I'm running a
glmmPQL model? Has anyone some experience with that?
I also have one concern... since the SE in the lme models are not just
ordinary SE... is there any reservation in using a regular bootstrap on
them? Are there any special bootstrap tools for that?
thanks for all the help
best
Andrej
_________
Andrej Kveder, M.A.
researcher
Institute of Medical Sciences SRS SASA; Novi trg 2, SI-1000 Ljubljana,
Slovenia
phone: +386 1 47 06 440 fax: +386 1 42 61 493
[[alternative HTML version deleted]]
