You have 16 groups so you need 16 items in the contrast vector (assuming no continuous variables), so you need something like -1 +1 repeated 8 times to give you 16

On 11/17/2021 7:23 AM, Thalhammer, Melissa wrote:

External Email - Use Caution

Dear Freesurfers,

I am struggling with the group analysis (FT vs. PT) to investigate cortical thickness while correcting for scanner (4 sites) and sex. I have posted a question already regarding the handling of discrete variables and I got to know that all discrete variables should be treated as classes rather than variables in the FSGD file. However, I have difficulties implementing the correct Contrast file right now. My classes are the following:

PTMaleScanner1 FTMaleScanner1 PTMaleScanner2 FTMaleScanner2 PTMaleScanner3 FTMaleScanner3 PTMaleScanner4 FTMaleScanner4 PTFemaleScanner1 FTFemaleScanner1 PTFemaleScanner2 FTFemaleScanner2 PTFemaleScanner3 FTFemaleScanner3 PTFemaleScanner4 FTFemaleScanner4

where the two groups FT vs. PT should be compared while regressing out scanner and sex. As far as I understood, we need "1 -1" for the group difference as well as "1 -1" for the gender difference. But how should the complete contrast matrix look like? Thank you in advance!

Best, Melissa

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer