In this case you would have six regressors:
1. control-offset
2. hiv-offset
3. control-age
4. hiv-age
5. control-ed
6. hiv-ed
If you want to test group thickness "regressing out" the effect of age
and ed, then you would just set the contrast to:
1 -1 0 0 0 0 0
Note that this will test it at age=0 and ed=0, which you may or may
not want to do. Ideally, you would test them at some age/ed in the
range of your sample. A lot of people simply demean each covariate,
but this can bias the results depending upon who you add or
remove. Better to choose an age/ed level independent of your sample.
doug
On Sat, 13 Jun 2009, Jeff Sadino wrote:
Hello Freesurfers,
My apologies for this very basic question more about statistics than
freesurfer, but I have not found an answer anywhere on the wikis, or the web
(maybe it is too basic :-) ) I am interested in age-related effects of hiv and
I wanted to covary for age and education in mri_glmfit. For example, if my
fsgd file is:
class control
class hiv
variables age ed
input subj1 control 40 13
input subj2 hiv 35 12
...
...
and I use dods, is my contrast matrix the place where I would tell it to
covary? Something like:
0 0-1 1 1 11 1
(y-int) (age slope) (covary for age) (covary for ed)
As always, thank you for your help,
Jeff
_
Hotmail® has ever-growing storage! Don?t worry about storage limits.
http://windowslive.com/Tutorial/Hotmail/Storage?ocid=TXT_TAGLM_WL_HM_Tutorial_Storage_062009
--
Douglas N. Greve, Ph.D.
MGH-NMR Center
gr...@nmr.mgh.harvard.edu
Phone Number: 617-724-2358
Fax: 617-726-7422
In order to help us help you, please follow the steps in:
surfer.nmr.mgh.harvard.edu/fswiki/BugReporting
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer