On Tue, Nov 29, 2011 at 10:22 AM, Marjorie Dole
<[log in to unmask]> wrote:
Dear SPM users,
I have a question concerning second level analysis:
I have an experimental design with three conditions, A, B, C, and a baseline. I want to evaluate fMRI activations specifically for each condition versus the two others in my whole group, by testing the contrasts : A-(B+C), B-(A+C) and C-(A+B).
So I did for my first level analysis the following three contrasts: A-baseline, B-baseline and C-baseline.
At the second level I used a full factorial design with one factor (Condition; three levels: A, B, C) and then I contrasted each condition versus the two others using the following contrasts : 2 -1 -1 ; -1 2 -1; -1 -1 2.
Is this approach valid?
Do you have any suggestion?
Thanks for your help,
Marjorie