Hi...did someone post any suggestions on this and I missed them? Just
wanted to check.
On Nov 3, 2005, at 3:25 PM, Richard Albistegui-DuBois wrote:
> I have data from several subjects, and I want to make sure I'm
> planning my multilevel design properly.
>
> Each subject was scanned twice on a simple motor task, with a
> behavioral measure of interest acquired at each timepoint. What I'm
> interested in is knowing whether there are any group changes in
> activity in motor cortex which are associated with the difference
> in the two behavioral measures.
>
> My current plan is to do a first-level analysis of each timepoint
> separately, with a second-level analysis within each subject,
> comparing the two timepoints. For example, with timepoints A and B:
>
> Level 1: Timepoint A
> Level 1: Timepoint B
> Level 2: A.feat-B.feat
>
> Then, at the third level, take the .gfeat from each single subject
> A-B comparison, and combine them all with two EVs: EV1 is one for
> all subjects, EV2 is the difference in behavioral score between A
> and B for that subject. I would have three contrasts: EV1=1, EV2=0;
> EV1=0, EV2=1; EV1=0, EV2=-1
>
> Does this sound correct to y'all? Have I explained it comprehensibly?
>
> Many thanks.
>
> Richard Albistegui-DuBois
> UCLA NeuroRehab, lab of Bruce Dobkin, MD.
> Office: 1-132 Reed
> Phone: 310-825-4016
> Mobile: 310-774-1305
> Fax: 310-794-9486
> AIM: dubistegui
> email: [log in to unmask]
>
|