Hi again,
I did what Vladimir said and it worked fine. I used the baseline time window (-100
0) in the Paired t-test, and I got smaller clusters of activation with a the t
contrast.
Though for one of the conditions FWE .05 yielded good localized results, for
another condition I had to use Uncorrected .0005 instead to get better clusters.
Any comments on that?
I also performed a Region of Interest using Marsbar and got significant results in
the regions I was interested, while regions without data were not significant.
Thanks again
kind regards
Eduardo
|