[R] repeated measure in GLM

[Thomas Lumley]

On Wed, 19 Nov 2003, John Christie wrote:

>
> 	I was recently asked to perform a GLM analysis (the person comes from
> the JMP world) on a repeated measures design.  I have found some things
> using aov but I cannot find anything with glm.  In fact, multiple
> regressions in general with repeated measures seems to be poorly
> covered in documentation.  I remember SPSS has separate commands to
> handle them.
> 	I have within variables SOA and ec and a between factor subjrt.  The
> independent variable was RT.  The ec factor cannot be construed as
> parameteric whereas SOA can.
>

If the person comes from the JMP world, GLM doesn't translate as glm().
You probably want either aov() or lme() [in the nlme package], depending
on how complicated the analysis is.

	-thomas


Thomas Lumley			Assoc. Professor, Biostatistics
tlumley at u.washington.edu	University of Washington, Seattle