Thank you for your comment, Bruce.

Regardless of technical issues, I am still troubled by the small magnitude of extracted value of pial surface area. Following group comparison and correction in QDEC, there is a significant cluster (size: 760 mm^2). In order to make a bar group of group means, I use numbers from the "cache.th20.neg.y.ocn.dat" file, but the average is less than 1. Did I use the right file to plot group mean of pial surface area, but why the magnitude is so small ?

Thank you for your help!

Karl