 |
 |
This is a multi-part message in MIME format. Hi all, I have a question about conjunction analysis on the mixed effects in SPM. I have a 2x2 design. If I want to perform a conjunction analysis to see which areas show activity for the second AND the fourth condition (VS baseline), I would try it like this: 1: Create two new contrasts: [0 1 0 0], [0 0 0 1] (SPM does not give them automatically but that is not the problem). 2: Perform a conjunction analysis by selecting both contrasts: [0 1 0 0] & [0 0 0 1], selecting conjunction when prompted, etcetera. I came across a presentation on analyses in SPM (see attachment). It seems they suggest a different approach. I do not understand their line of thought. I do see that my conjunction analysis differs in two ways: 1) I do not have a 2x3 but 2x2 design; 2) in the example, they do not contrast against baseline but one condition against the other. Still, they seem to suggest an analysis different from mine. They seem to suggest I should perform the analysis as such: [0 1 0 1] (both rhythm temp grouping and verbal temporal grouping VS baseline) and mask (is that what is meant by 'given'?) that with the conjunction [0 1 0 0] & [0 0 0 1]. I hope I understood the example -- having a presentation without presenter proves difficult at times. I wonder how the two approaches differ and why one should choose one over the other. Does anyone have an idea? Cheers, Dirk. -- ======================================= Dirk Schuit MRC Cognition and Brain Sciences Unit 15 Chaucer Road Cambridge CB2 7EF UK +44 (0)1223 355294 (extension 218) [log in to unmask] =======================================