R help - May 2004 - mixed models

Hi folks

Can anyone direct me to a guide to analysing mixed models in R that is 
understandable by mere mortals (or worse, mere biologists)? I have read 
MASS and also Mick Crawley's book on the subject, have thrashed my way 
through the various help files from the NLME package and I'm afraid I 
remain baffled :-(. Particularly confusing is exactly what groupedData 
does, how I should use it to define the structure of my dataset and how 
I then use that to define a model in LME. So pretty much everything 
then.

Thanks for any help

Rob Knell

> Hi folks
>
> Can anyone direct me to a guide to analysing mixed models in R that is
> understandable by mere mortals (or worse, mere biologists)? I have read
> MASS and also Mick Crawley's book on the subject, have thrashed my way
> through the various help files from the NLME package and I'm afraid I
> remain baffled :-(. Particularly confusing is exactly what groupedData
> does, how I should use it to define the structure of my dataset and how
> I then use that to define a model in LME. So pretty much everything
> then.
The canonical reference is "Mixed Effects Models in S and S-PLUS", by
Pinhero and Bates, and it is thorough and readable.  As a secondary source
to get you started while you find the first, the S-PLUS statistical manual
has a pretty good chapter on it (including a demo of groupedData).

http://www.insightful.com/support/splus60unix/statman1.pdf

See chapter 14.

Cheers

Jason