[R-sig-ME] Generalized mixed models for poisson distributions

Douglas Bates bates at stat.wisc.edu
Fri Oct 24 17:42:19 CEST 2008


On Fri, Oct 24, 2008 at 8:36 AM, Page E. Van Meter <vanmete7 at msu.edu> wrote:
> Hello,
> Reading through this forum has been very informative, but I am afraid I have
> a rather basic question. I am new to both R and generlized models (I have
> experience running linear mixed models with SPSS).

> I have been playing with the glm function in R (using Faraway as my guide),
> but this forum seems to focus on glmer and lmer. What is the difference
> between running mixed effects models using glm and glmer?

I think you are confusing generalized linear models, which are models
used when the response is binary or binomial or a count, and
mixed-effects models, which are models where some of the coefficients
in the predictor are fixed-effects parameters, which apply to an
entire population or well-defined subgroups of the population (females
vs males, say), and other coefficients are random effects, which apply
to the particular observational or experimental units being
considered.

The glm function fits generalized linear models with fixed-effects
parameters only.  The glmer function fits generalized linear mixed
models (the 'me' in the name is from 'mixed-effects').

I hope this helps.




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