Dear SPMers and especially PPIers,
I'm trying to run a gPPI on my data to compare to the results I obtained using regular PPI as I intend to compare three conditions (A, B, C).
If I understand correctly I should still estimate the GLM (should I create a effects of interest contrast or does the toolbox handle that?) prior to running the gPPI.
I am, however, uncertain how to contrast my conditions. My end goal would be a second level (t-test?) which yield regions showing differential connectivity to my seed regions for each possible comparison between my condition (A > B, B < A, B > C, B < C, A > C, A < C). Can I simply write 6 contrasts into the structure or will the colinearity between contrasts be an issue here?
|