[R-sig-ME] (no subject)

Douglas Bates bates at stat.wisc.edu
Mon Apr 2 22:17:15 CEST 2007

The values returned by fixef(fm) are the maximum likelihood (ML) or
restricted maximum likelihood (REML) estimates when fm is a linear
mixed-effects model.  For a generalized linear mixed model the
fixef(fm) values are those values that maximize the Laplacian
approximation to the likelihood.  That is, they are the ML estimates
up to approximation of likelihood function.

The purpose of mcmcsamp is to provide a sample from the posterior
distribution of the model parameters.  These samples are typically
used to assess precision of parameter estimates or to assess the
significance of certain terms in the model.

The summary of the result returned by mcmcsamp does give raw means of
the sample from the posterior distribution but I wouldn't quote those
as parameter estimate.  I would use the ML or REML estimates of the
parameters as the "best guess" at the parameter values.

On 4/2/07, Hallstrom, Wayne (Calgary) <Wayne.Hallstrom at worleyparsons.com> wrote:
> Hi,
> I am wondering what the difference between fixed effect estimates by the
> fixef() and mcmcsamp() routines? I noticed that these two seem to be
> aimed at determining that same fixed effect coefficients, but that the
> resutls differ. Is one prefereable to the other? I have been assuming
> tha the mcmcsamp() routie ints the one to go with based on reading
> previous discussions about it and lmer.
> Thnak you,
> Wayne Hallstrom
> "This email is confidential.  If you are not the intended recipient, you must not disclose  or  use the  information  contained in it.  If you have received this email in error,  please notify us immediately by return email and delete the email and any attachments. Any personal views or opinions expressed by the writer may not necessarily reflect the views or opinions of any company in the WorleyParsons Group of Companies."
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

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