Dear Ashley,
> I'm am writing to ask how it is I go about doing a conjunction
> analysis. What I am trying to do is that I have 2 groups and I want to
> see where both groups activate. I have 22 total subjects (11 patients
> and 11 siblings). I have a total of 126 timepoints (after discarding
> the 1st 2) and subjects perform a zero back to 2 back task where they
> alternate between the 2 tasks for 16 second epochs for a total of 256
> seconds. I am much more familiar with SPM96 than SPM99, but any advice
> than anyone can give would be greatly appreciated.
This can be addressed in a second level analysis by producing 22
contrast images (one for each subject). Enter these into a second level
analysis using a design matrix corresponding to a two-sample t test
(one column with 1s and -1s, denoting group and the other a constant
term of 1s). The conjunction of the contrasts [1 1] and [-1 1] should
give you what you want. Here you are assuming the error variance is
the same for the two groups.
With best wishes - Karl
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|