Dear Jason,
> First what I have. Two sessions, lets call them A and B. Hypothesis is
> that A and B contain the same fundamental cognitive components but B
> has another cognitive component of interest. Both sessions are 32
> second blocks of
>
> off on off on off
>
> A conjunction of the two main effects ( contrasts being: [1 0] and [0
> 1]) would show activation the two have in common, correct?
Using SPM99b it would show voxels in which the activations were
significant in both contrasts (i.e. common activations).
> How would I see activation that occurs in B and NOT in A?
> Is that the contrast [-1 1]?
No this is the differential activation B > A. Anecdotally one could
use exclusive masking to display the SPM{T} for c = [0 1] and
exclusively mask with [1 0] (i.e. exclude voxels that were
significantly activated in A. BUT note that failing to show an
activation is significant does not mean it is not there. In other
words you cannot confirm the null hypothesis. Generally one should
formulate questions (and present results) with this in mind.
I hope this helps - Karl
|