Dear experts,
For a 2nd level analysis in SPM5 I use a flexible factorial design including a subject factor and 3 subject groups.
I am working with 2 factors: one factor (group) with 3 levels and the second factor with 2 levels.
The design matrix consists of (numSubjects + numLevelsFactor1*numLevelsFactor2) columns (e.g. 16 columns for 10 subjects).
It is no problem to set the t-contrasts but I am struggling to test the difference between group1 vs. group2 for the level1 of factor 2 using an F-contrast.
This is the matrix I try to use in the contrast manager:
[ones(1,numSubjGroup1)/numSubjGroup1*-1 ones(1,numSubjGroup2)/numSubjGroup2 zeros(1,numSubjGroup3) -1/2 0 1/2 0 0 0;
ones(1,numSubjGroup1)/numSubjGroup1 ones(1,numSubjGroup2)/numSubjGroup2*-1 zeros(1,numSubjGroup3) 1/2 0 -1/2 0 0 0]
Unfortunately, I am getting the message "Invalid contrast".
Can anyone tell me whats wrong in the contrast definition?
Thanks a lot!
Maurice
|