Hi,
Consider fitting a generalised linear model to a binary response with group fitted as a factor. Suppose further than group has 4 levels but the response is
either all zeros or all ones for some of the levels...
For this (albeit simplified) case, the group kindly pointed me in the direction of the function brglm in R, which has worked really well.
However, brglm doesn't allow random effect terms to be included in the model.
Is anyone out there developing a bias reduced generalised linear mixed effects regression model (brglmer)?
Kind regards,
Gary Walker.
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|