Hello,
I believed I have figured out the problem. There was an error in the script such that the SEED region files that were catenated then included in 3dDeconvolve as -stim_files had not been "wavered". So I fixed the script so that the wavered SEED files were catenated then included as -stim_files in the 3dDecon. This appears to fix the problem.
Is there a way of performing the context dependent connectivity analysis (as outlined in the script I provided in the earlier message) that would not involve the use of waver (i.e., the waver function would be performed during the 3dDecon step).
-Steve