For each subject, create a contrast that compares A and B (for
example, A > B might be a 1 -1 contrast). This will create a
corresponding con* img file.
To compare men and women, enter those con* images in a second level
two-sample t-test (e.g., Group 1 scans = men, group 2 scans = women).
Your design matrix will have two columns, one for each group (ideally
you will have approximately the same number of people in each group).
So, if you enter men for group 1, and women for group 2, a t contrast
will give you voxels in which the (A vs. B) contrast is significantly
bigger for men than women.
Hope this helps!
On Mon, Jul 6, 2009 at 4:01 PM, Marco
Hubert<[log in to unmask]> wrote:
> i have no idea if this is the correct way but i will try it:
> i have the following problem (maybe it looks to simple for the advanced people
> here :)):
> I have Stimuli A and B as well as men and women. Now i want to contrast A vs
> B and compare this contrast between men and women.
> How can i manage this in SPM5 within a 2nd level analysis.
> Again: A vs B by Comparing men vs women.
> Thanks in advance!