Dear Jerry,
Jerry Allison wrote:
> Search as I might, I can't find a succint answer to this rather simple question.
>
> I have two identical fMRI sessions in a single subject.
>
> The paradigm is a simple AB for each session.
>
> I prescribe the number of sessions as two in the SPM99 design matrix.
>
> This results in a design matrix having 4 columns. The first two columns contain the condition effect matrix for the two sessions,
So, I infer that you entered the # of conditions as one. This is probably fine, but I just wanted to mention this because
it does affect the structure of the contrasts.
> offset in time. The last two columns contain white bars representing session effects for each session, again offset in time.
>
> In oder to compute a t contrast for the specified condition, do I specify the contast as [1 1 0 0]?
This contrast will test the (null) hypothesis "the average of the effects across the two sessions is zero".
>
>
> I have processed this set of data two ways. When I analyze the data as two sessions, I get what appears to be reasonable activation for the n-back task that we are using. When I process the data collectively as one session, at the same level of significance, there seems to be global activation. The activation patterns are probably similar, but when I lump the data into one session, I have to dramatically lower the p value to see the pattern. What would explain this behavior.
Hmmm, interesting, as I would have guessed that if anything the opposite pattern of results would have been observed.
Did you "model intrinsic autocorrelations with AR(1)"? If so, maybe the lack of a second session effect messed this step up a bit (and if so, perhaps providing invalid p-values for the one session model). For the case you described, the two session model would be at least as valid as the one session model, so I'd go with the two session model.
>
>
> When I realigned and coregistered the data, I treated it as one subject, two sessions, thus all of the data is essentially realigned to the first scan of the first session. When setting up the models, data were all "scaled", thus I would expect the data to be globally scaled to the same value (100?).
That is what Iwould think, but I'd wait for an answer from someone with more expertise on that aspect of SPM99.
Sincerely,
Eric
>
>
> Jerry Allison, Ph.D.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|