On Wed, 6 Jun 2001, Federico Turkheimer wrote:
> you should not use corrected thresholds for apriori hypothesised
> anatomical brain regions.
For peak/magnitude statistics, I would disagree. Unless you know the
exact voxel of the brain region of interest, you have a multiple
comparisons problem on your hands: As you search over the volume of your
apriori hypothesised regions, you are implicitly performing multiple
tests.
-Tom
-- Thomas Nichols -------------------- Department of Biostatistics
http://www.sph.umich.edu/~nichols University of Michigan
[log in to unmask] 1420 Washington Heights
-------------------------------------- Ann Arbor, MI 48109-2029
|