OK, perhaps I wasn't clear the first time. I've already run 3dDeconvolve (using waver, etc).
What I want to do now, in order for the purpose of graphically depicting the differences in beta intensities, is to average all of my data for each voxel from a particular event that I've included as a regressor in the model--I'll later re-run regression analysis on this averaged dataset.
The tricky thing is, I don't quite know how to specify which time points to average given that the stimulus is not locked to the TR.
Jane