Hi Eugene,
Thanks for the quick response! I think your suggestions make sense, but let me check my understanding.
For 1 & 3, at the second level, I should feed in the runs from my subject's first and second scans. In my design matrix, I'll have two EVs - one for the first scan and one for the second scan. The runs will be assigned to the correct EVs. I'll have [1 -1] and [-1 1] for my between-scan contrasts.
Then, at third level, I should use the cope.feats from my difference contrasts. For the design matrix here, I should model the mean and (demeaned) change in CAPS score.
Is that right?
Thanks again,
Liz
|