>Sounds like you want one constant-amplitude regressor and one modulated one? Sure - that's two separate EVs - for the latter, >demean the set of response values before using them to form the regressor.
>Cheers.
Pardon the simple follow-up, but I wanted to just make sure I was clear on the EVs (since I didn't want to run it on ~80 subjects with incorrect EVs).
For the constant-amplitude regressor (which is just when the feedback is displayed), my first column would just be when (in seconds) the feedback was displayed, the second column would be just all zeros (since I'm trying to model an event with a zero-duration), and the third column would be all ones (since it is just a constant regressor)..
For the modulated regressor, my first two columns would be the same and I would just de-mean the third? I wasn't sure if that redundancy was ok in my GLM (but I'm a relatively newbie to fMRI) or if I misunderstood something? Thanks much!
Best,
jamie.
|