Hi, let's reconfirm what's in this higher-level analysis: In this particular analysis, all entries correspond to ML-MH The first 12 entries (modelled by EV1) are groupA condition C The next 12 (EV2) are groupA condition P Then group B condition C Then group B condition P Is this not correct? I wasn't 100% sure from your original email what was in this higher-level analysis. Cheers. On 21 Jun 2007, at 18:57, Remya Nair wrote: > Hi Steve- > > Thanks for your help. > I just want to confirm one thing though... > > I understand with the way you set up the EVs, but to compare groupA > vs groupB ML-MH for condition C, shouldn't the contrast be [1 -1 -1 > 1]? > and not [1 0 -1 0] as you wrote. > My understanding is that would be a comparison of groupA cope1 vs > groupB cope1, right? > > Thanks again. > R > > On Jun 17, 2007, at 6:08 PM, Steve Smith wrote: > >> Hi, >> >> On 15 Jun 2007, at 17:14, Remya Nair wrote: >> >>> hi- >>> >>> i was hoping someone could help me set up an anova. >>> >>> let me describe the experiment: >>> >>> there are 2 groups with 12 subjects in each group. >>> each subject did 8 runs (4 run for condition C and 4 runs for >>> condition P, with 3 levels for each condition (ML/MH/SH)). >>> >>> i ran the first-level analysis with 3EVs (ML/MH/SH) and then >>> combined >>> the 4 runs of a single condition. >>> >>> now i have 2 .gfeat directories for each subject: 1 for condition C >>> and one for condition P (each has 4 cope.feat dirs - ML, MH, SH, >>> MH-SH). >>> >>> i want to look at the group differences across different conditions >>> e.g., groupA vs groupB, looking at ML-MH for condition C. >>> >>> so i would have 48 inputs in my higher level analysis: >>> 12 groupA cope1 >>> 12 groupA cope2 >>> 12 groupB cope1 >>> 12 groupB cope2 >>> >>> but i am not sure how to set up my EVs and contrasts. >> >> Something like: >> >> 1 0 0 0 >> 1 0 0 0 >> 1 0 0 0 >> 1 0 0 0 >> 1 0 0 0 >> 1 0 0 0 >> 1 0 0 0 >> 1 0 0 0 >> 1 0 0 0 >> 1 0 0 0 >> 1 0 0 0 >> 1 0 0 0 >> 0 1 0 0 >> 0 1 0 0 >> 0 1 0 0 >> 0 1 0 0 >> 0 1 0 0 >> 0 1 0 0 >> 0 1 0 0 >> 0 1 0 0 >> 0 1 0 0 >> 0 1 0 0 >> 0 1 0 0 >> 0 1 0 0 >> 0 0 1 0 >> 0 0 1 0 >> 0 0 1 0 >> 0 0 1 0 >> 0 0 1 0 >> 0 0 1 0 >> 0 0 1 0 >> 0 0 1 0 >> 0 0 1 0 >> 0 0 1 0 >> 0 0 1 0 >> 0 0 1 0 >> 0 0 0 1 >> 0 0 0 1 >> 0 0 0 1 >> 0 0 0 1 >> 0 0 0 1 >> 0 0 0 1 >> 0 0 0 1 >> 0 0 0 1 >> 0 0 0 1 >> 0 0 0 1 >> 0 0 0 1 >> 0 0 0 1 >> >> and for the question: >>> e.g., groupA vs groupB, looking at ML-MH for condition C. >> use the contrast [ 1 0 -1 0] etc. >> >>> also, what if i want to look at ML vs. (MH-SH)? >> Probably the best thing is to achieve this contrast at first- >> level making the higher-level analyses easy. At first level this >> would be: >> [1 -1 1] though I'm not sure whether you do exactly mean ML - >> (MH-SH) >> >>> likewise, is it possible to look at group differences across >>> conditions C+P? >> >> Sure - in a similar manner to the above. >> >> Cheers, Steve. >> >> >> >>> >>> thank you in advance! >>> >>> R >> >> >> --------------------------------------------------------------------- >> ------ >> Stephen M. Smith, Professor of Biomedical Engineering >> Associate Director, Oxford University FMRIB Centre >> >> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >> +44 (0) 1865 222726 (fax 222717) >> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >> --------------------------------------------------------------------- >> ------ >> ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ ---