Dear Xavier,
>For a given stimulus type, I would like to look for regions that are
commonly activated across the two groups. I believe this >is done using
a conjunction analysis. I have got Tom Nichols' revised conjunction
programs, but am not sure how to >proceed. I would be very grateful for
any (ideally step-by-step) help with what I suspect is an extremely simple
Just a word to say that if you use SPM2, the procedure is rather
straightforward.
First, compute your stimulus effect for each subject (1st level).
Second, use these contrasts into a single matrix (so that you can test
[10] [01] or [1-1] 2nd level). Then select the [10] (gpA) and [01] (gpB)
contrasts (left click + ctrl) ... SPM will ask you what kind of
conjunction you want to perform "Conjunction / Global"
Global indicates that one or more of the conjoined contrasts show an
effect whereas Conjunction indicates that all conjoined contrasts show
an effect.
Hope it helps
- cyril
|