Hi,
I am trying to test that there are no preexisting gray matter differences between four medication groups. In SPM, I created a one-way between subjects model and now have to define the F-contrast. My question is, since there are four groups, will the F-contrast matrix only have 3 rows because group 1 is the reference group to the other 3?
Right now, this is how I have defined the contrast matrix:
[1 0 0 -1
1 0 -1 0
1 -1 0 0]
Is this correct for the omnibus test? If so, would the T-contrasts look like this?
[-1, 1, 0, 0] (difference between group 1 and 2)
[-1, 0, 1, 0] (difference between group 1 and 3)
[-1, 0, 0, 1] (difference between group 1 and 4)
[0, -1, 1, 0] (difference between group 2 and 3)
[0, -1, 0, 1] (difference between group 2 and 4)
[0, 0, -1, 1] (difference between group 3 and 4)
Thanks in advance!
|