Dear Donald,
Thank you for your quick reply and good advice.
But it is still not clear for me, how to make the contrasts. In my design , what it really matters are the differences between experimental conditions.
I choose my ROI functionally based also on two distinct contrasts , let's say A>B (and its negative: B>A)
So in my PPI-GLM I create 15 EVs (not 17 because 2 conditions are instruction and response, which I probably do not need to have an interaction with the ROI).
Does it make sense to create such contrasts in PPI (like A>B in my original GLM A(1) > B(-1) ) ? :
PPI_A (ConditionA*ROi:1) > PPI_B (Condition_B*ROI:-1)
PPI_B (Condition_B*ROI:1) > PPI_A (ConditionA*ROi:-1)
Any hint is much appreciated!
Best
Elif
|