Dear FSL experts,
I am running an unpaired t-test to compare 2 groups (patients vs controls) with 2 regressors of NO interest: age and gender.
I am using the following design matrix (design.mat),
/NumWaves 2
/NumPoints 11
/PPheights 1 1
/Matrix
1 0 65 2
1 0 41 1
1 0 63 1
1 0 51 1
1 0 52 1
1 0 59 2
0 1 56 1
0 1 47 1
0 1 40 2
0 1 48 2
0 1 46 2
where the 3rd column represents age and 4th column, gender (1:Female and 2: Male)) and the following contrast matrix (design.con):
NumWaves 2
/NumContrasts 2
/PPheights 1 1
/Matrix
1 -1 0 0
-1 1 0 0
Then, when I run randomise by command line, I got the following error:
"ERROR: Program failed
An exception has been thrown
Error: This (f)contrast appears to be rank defficient, please check your design"
I have been reading trough the post and fslwiki and I thought that maybe I had to demean the 2 covariates of no interest (although I am a little bit confused about it). The new design matrix that I have created for this case is:
/NumWaves 2
/NumPoints 11
/PPheights 1 1
/Matrix
1 0 14.6207 0.620689655
1 0 -9.3793 -0.37931034
1 0 12.6207 -0.37931034
1 0 0.6207 -0.37931034
1 0 1.6207 -0.37931034
1 0 8.6207 0.62068965
0 1 5.6207 -0.37931034
0 1 -3.3793 -0.37931034
0 1 -10.3793 0.620689655
0 1 -2.3793 0.620689655
0 1 -4.3793 0.620689655
But I am still having the same error.
Could someone please check the matrices and give some advise?
Thank you so much!
Best regards,
Elena
|