Hi Darren, yes, you'll need to put an FSGD file together. This design has two discrete factors, one with two levels and one with three. It's the three level that breaks QDEC.
doug

On 9/23/12 12:34 AM, Darren Gitelman wrote:

Dear List

I'm wondering about the proper way to set up an analysis that includes site as a regressor, and whether this is possible in qdec?

My qdec.table.dat file looks like this.

fsid group site

1 grp1 1

2 grp2 2

3 grp2 2

4 grp1 3

.... and so on

I then have .level files for group which has 2 levels, and site which has 7 levels.

The type of question I want to answer for example is: Is there a difference between grp1 and grp2 regressing out the effects of site?

Is this possible with Qdec?

When I select the group variable on the design tab the analysis results shows the question of the form: Is there a difference between grp1 and grp2, but there is no mention of regressing out the site effect. Looking at the design matrix X.mat it only includes 2 columns which isn't right.

If I select both group and site on the design page I get an error.

Do I need to put together an FSGD file instead?

Thanks

Darren

--
Darren Gitelman, MD
Northwestern University
710 N. Lake Shore Dr.
Abbott 11th Floor
Chicago, IL 60611
Ph: (312) 908-8614
Fax: (312) 908-5073
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer