On Fri, 8 Aug 2008 13:33:13 -0700, Joe Andreano <[log in to unmask]> wrote:
>Hello all,
>
> I'm having a bit of difficulty specifying a first level model,
>which hopefully someone can help with. The experiment is a within-subjects
>block design, in which participants are scanned three times during the first
>visit, then another three times in the second visit. Since I'm looking at
>sustained effects, I don't want to concatenate the individual sessions
>within each visit, so I've modeled my experiment as six sessions. The
>problem is that I want to add a regressor which will take into account which
>of the two visits was first (they're a bit different, and presented in
>counterbalanced order).
>
> I tried adding a regressor to each session to specify this, (by
>entering a string of 1s or 0s equal to the number of scans per session) but
>in the resulting design matrix the regressors for each session extend all
>the way through the matrix, beyond each individual session, and the bars
>seem to be the same color for every session, even though half are coded 1,
>and half 0, so I'm pretty sure this is the wrong way to go about it.
>
> Does anyone know of a way that I can model the session effect here
>without concatenating?
It's fMRI, not PET, right? In which case I would assume the main effect of
session is already modeled by SPM.
>
>
>Thanks,
>
>Joe
>
|