 |
 |
Hi Lydia, ah I see! You seem to be talking about the PPI GUI tool in SPM, which implements "standard" PPI (sPPI). I always conduct my PPI analyses using the generalized PPI (gPPI) toolbox: https://www.nitrc.org/projects/gppi So I'm afraid I can't answer this question and need to refer to the SPM experts. However, I can highly recommend gPPI as it generalizes PPI to experimental designs with more than 2 conditions - in these cases, sPPI is no longer valid. See: https://www.ncbi.nlm.nih.gov/pubmed/22484411 I am not sure whether sPPI is valid for your design with 2 tasks + a baseline condition. Maybe a PPI expert like Donald McLaren can help here. The VOI timecourse extraction itself should not depend on any contrast; it's simply the raw timecourse of the voxels (or more precisely, usually the first eigenvariate of the BOLD signal from the voxels), possibly adjusted for effects of interest (but this is chosen elsewhere). All the best, Phil