Hi Jing, concerning 1): for DCM or PPI analyses, you can extract the time-courses from those voxels only that shown an experimental effect at some threshold (often 0.05 uncorrected) - but another option could be to extract from all voxels in a fixed sphere located at the peak of the 2nd level analysis. To my knowledge, both approaches are valid. concerning 2): To my knowledge, it would be valid to apply the PPI on an event-related paradigm and it has also been done previously - but the block-design is probably the more powerfull. Best, Brian Den 12-05-2015 kl. 10:03 skrev Jing Yu: > Dear all, > I posted a question regarding the correction issue of PPI analysis a few days ago, but didn't get any reply yet. Any thought or suggestion about this issue will be highly appreciated. Thank you very much! > > Best, Jing > > > > ------------------------- > Dear SPMers, > I'm trying to do the PPI analysis. After reading the PPI section of SPM manual, I have a few questions about the correction issue. > 1) In the manual, FWE correction was applied in the extracting VOIs step, and after the PPI analysis no correction was applied to the result. Why it needs to use the corrected contrast to get the VOI. In other words, can I use the uncorrected contrast to get the VOIs, and corrected it in group analysis at the second level? > 2) I noticed that PPI analysis is usually applied to block design. If I want to apply it to event-related design, would that be OK to do it? although it will have more fasle negative results ... > > Thanks! > > Best regards, Jing