Hello FSL users and experts,
I have some questions concerning my GLM contrasts and the demeaning of covariates:
I have two groups and covariate which mean is different for the two groups. I enter the covariate seperatly for the two groups in my design. Demeaning was done with demeaning option from FSL which I expect to demean the covariate seperatly for the different
groups and did not demean beforehand. (I conducted t-Tests with this design for DTI data).
EV1 are patients, EV2 controls. EV3 is my continous covariate for patients, EV4 for controls. EV5 is another continous covariate of no interest (I did not use FSL GLM GUI so I didn't enter a "group" variable seperatly).
This are my contrasts:
EV1 EV2 EV3 EV4 EV5
1 -1 0 0 0
-1 1 0 0 0
0 0 1 1 0
0 0 -1 -1 0
0 0 1 -1 0
0 0 -1 1 0
First question: Line 1 and 2: I want to test for difference between EV1 and EV2. Is there a probleme that there is a mean difference for the covariates in the two groups? Is there a general problem with splitting up the covariates for the two groups?
Second question: Line 3 and 4: I want to test for linear relationship with my covariate. Do I run into a problem by splitting up the covariate into the two groups and demeaning them seperatly in FSL? Do I loose power? Does it test the association seperatly
for EV3 and EV4 and do both need to be associated or do I get a total asscociation of both groups together?
Line 5 and 6 should test for interaction. I hope this is fine.
Thank you in advance for any help!
Best
Elena
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1