I posted these question several days ago but did not receive any replies.
Three questions regarding output from a generalized linear mixed model in
v6.2:
1. Are the fixed effect estimates produced by glme on the logit scale if
"fam = binomial" was used?
2. If the function, "intervals(obj, which = 'fixed')" was used on the above,
would the intervals likewise be on the logit scale?
3. What is (are) the best statistic(s) to use to compare nested generalized
linear mixed models?
Respectfully,
Frank Lawrence
v6.2
WinXP
|