You can run randomise with a group size of 1 (individual), and of course the
control group will be size N.
There is a limitation though because you can only do N-1 permuations and
therefore the distribution for your test statistic will be sparse, and
importantly it will limit the maximal alpha that can be detected. For
instance if you have 100 controls than the maximal alpha you can detect is p
= .01.
-----Original Message-----
From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf
Of Karim Virani2
Sent: Friday, September 17, 2010 9:38 AM
To: [log in to unmask]
Subject: [FSL] vbm - single result compared to group
Hello all-
Once VBM has run its course, is there a way to compare individual results
vs. the group?
Thanks
Karim
|