Dear FSL experts,
I notice in the FSL guide that for a two-group unpaired t-test the design matrix is set up with two columns and the contrast with two columns, for example for eight subjects:
Design:
0 1
0 1
0 1
0 1
1 0
1 0
1 0
1 0
Contrast:
1 -1
-1 1
What is the difference between this design and just using just one EV column (with contrasts of 1 and -1)?
Is the issue just getting the correct number of degrees of freedom? I was running the latter, but for 150 subjects so that if it is only a DF issue, the effect will be small and shouldn't alter the results by much.
Thanks,
Mark
|