Dear Sun,
a "within-subject conjunction" is going to produce invalid results, see related messages on the mailing list. You can make some descriptive "conjunction" though by running a one-sample t-test for task A and another t-test for task B, threshold at a certain level and then combine these two masks with Imcalc to get the overlap = this corresponds to voxels activated during task A and task B.
Best,
Helmut
|