Dear Dr. Brett, In marsbar I defined a significant cluster from the group analysis and saved it in a file. I then selected that file while in marsbar for single subject analyses and asked for plot of the mean signal. Of course, the voxels in that cluster may or may not be significantly active in individual subjects, but I presumed that the signal present in that cluster would be plotted. When I saw the email about signal from non-significant voxels not being plotted, I thought that could also be the case in my analysis so I emailed the list. Thanks, Warren Darling At 11:47 AM 11/5/2004 -0800, Matthew Brett wrote: >Hi, > > > What we have seen is that despite setting a specific ROI in Marsbar, if > > this falls within a non-signifincant area the time course returned is from > > the nearest significant cluster (a threshold previously defined in your > > model). This is the same behavior seen in SPM (as Marsbard apprears to > > simply evoke the plotting function from SPM). > >The marsbar interface allows you to define spheres, boxes etc, but >also clusters from SPM. Obviously, to define a cluster from SPM, the >cluster has to be above the threshold you set - by definition. Is it >possible to explain a bit more about the problem? > >Thanks a lot, > >Matthew