nlme model specification

I wanted to point out that BIC doesn't need to be though of in a Bayesian
context and there is no need for the user to explicitly specify a prior to
use BIC -- it is simply -2*(loglik) + k*log(n), with k being the number of
estimated parameters and n the sample size.

Yep, my mistake. BIC doesn't need priors, as stated on the first page of the
article I pointed out! (It had been a while, OK?)

But, as you mention below, with random effects models it's tough to
determine what values you plug in for n and k. Same goes for AICc. I think
this leads back to considering a different overall strategy for model
selection in a random effects context, such as Bayes factors. I am not aware
of the R packages that support Bayesian model selection, but I bet they're
out there.

-----
David Hewitt
Research Fishery Biologist
USGS Klamath Falls Field Station (USA)

View this message in context: http://www.nabble.com/nlme-model-specification-tp17375109p17442783.html
Sent from the r-sig-ecology mailing list archive at Nabble.com.