Hi Bob,
It must be option (2) rather than (3) ;) and (1). The individual subject results are also 0. Any ideas why?
The datasets were preprocessed with a self-tuned version of afni_proc.py including WM local regression and spatial blurring with 3dmerge (FWHM 6mm) and nonlinear warping to TT_N27. In other words, nothing really special I believe. I can generate reasonable FC maps with InstaCorr in the individual subject datasets.
May it be related to the fact that I am running 3dGroupInCorr in a remote computer, and run afni -niml in another terminal that was open with xterm &.
Thanks in advance and best wishes,
Cesar