Meeri Kim wrote: > I have a quick question about removing a global signal as a > covariate of no interest across all voxels. Does the global signal > have to be of the right order in a numerical sense, or does it just > use the general trend? The shape matters, but the scale doesn't. dan