 |
 |
Hi Steve, Thanks for the reply. As you suggested, I looked at the cope1.feat/stats/zstat image. The unthresholded zstat image does show a fair amount of activation. So, it looks reasonable enough.... I also followed your second suggestion, and looked at the timeseries for the filtered functional data that were fed into the higher-level analysis. I basically selected random voxels and observed their timecourse of activation. I'm not sure how much useful information I'm getting from doing this. Is there something else I should be looking at in the individual subjects' filtered functional data? Should I look specifically at the maxZstat voxel? Thanks, Mona On Tue, Jul 28, 2009 at 11:45 AM, Steve Smith<[log in to unmask]> wrote: > Hi - have you looked at the unthrehsolded group-level zstat image that is in > the <something>.gfeat/cope??.feat/stats directory? Check that it contains > something somewhat reasonable...if your model was correct at second-level > and all subjects activated consistently then yes you should get a strong > group result. You can load in the data that was input to the final > group-level analysis (filtered_func_data in that cope??.feat directory) into > FSLView and turn on timeseries viewing to see what the different subjects' > responses were that fed into the groupstats. > > Cheers. > > > On 27 Jul 2009, at 19:26, Mona Lisa Chanda wrote: > >> Hi, >> >> I just completed a mixed effects OLS analysis for the group mean of 18 >> first- >> level FEAT directory inputs. When I looked at the output, the higher-level >> COPE map was empty. I'm not sure what this means. There was a MaxZStat >> of 2.6, but no blobs of activation are visible on the map (presumably >> because >> so few voxels passed threshold). I'm puzzled about this because (a) most >> of >> the lower level COPES do show activation (b) I did not apply any masks to >> the >> data (c) my supervisor predicted that we'd see tons of activation :) >> >> I understand that this result may simply reflect the statistics that were >> applied at the higher-level (e.g. different degrees of freedom). But, I'm >> also >> concerned that something might have gone wrong with the analysis. >> >> Any thoughts on this? >> >> >> Thanks, >> Mona >> > > > --------------------------------------------------------------------------- > 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 > --------------------------------------------------------------------------- > -- Mona Lisa Chanda Northwestern University Department of Physiology 303 E. Chicago Morton 5-660 Chicago, IL 60611 Cell: (312) 972-3173 Office: (312) 503-1703 Fax:(312) 503-5101