[R-sig-ME] Bug?: var estimates on exp scale for gaussian(link=log)

Douglas Bates bates at stat.wisc.edu
Mon Jan 7 23:33:53 CET 2008


On Jan 7, 2008 4:26 PM, Gregor Gorjanc <gregor.gorjanc at bfro.uni-lj.si> wrote:
> Gorjanc Gregor <Gregor.Gorjanc at ...> writes:
> > Hi!
> >
> > I would like to fit log-normal fixed model with lmer. Fixed effects estimates
> > are returned on log scale, while variance components are returned on exp scale.
> > Why is this the case?
>
> I would just like to add that I expected to get variance estimates on link (log)
> scale as it is done for poisson (log link) or binomial (logit or probit link)
> families.

I'd need to look at the code more closely but I suspect that the
results you are getting are spurious.




More information about the R-sig-mixed-models mailing list