Did you do both contrasts a minus b and b minus a? The stats on tbss, at least the old version, were one sided so you just have to include both contrasts in your matrix.
I.E. 1. -1 on one line and -1 1 on the next line
Dg
------Original Message------
From: Ruth Carper
Sender: FSL - FMRIB's Software Library
To: [log in to unmask]
ReplyTo: FSL - FMRIB's Software Library
Subject: [FSL] TBSS - single subject vs control group
Sent: Aug 27, 2008 5:19 PM
Hi,
I'm interested in comparing DTI data from a single subject to a
multi-subject control group using TBSS. Is this possible? I tried running
-randomize- with a 1 vs. many design, but results were produced only for the
comparison where the Control group has greater FA than the single subject,
but failed on the comparison where the Control group has lower FA.
--Ruth
Sent from my Verizon Wireless BlackBerry
|