Dear Christy,
> I have a question that is related the the discussion of SVCs. If one
> were interested in a general region of the brain, say for the sake of
> argument, posterior parietal cortex, is it valid to choose a voxel that
> is roughly in the center of posterior parietal cortex and create a
> sphere with a rather large radius that covers most of the lobe?
This is perfectly OK and is exactly what the SVC was designed for.
Mathew's point was that the center you choose should not be influenced
by the data you are trying to make inferences about. The sphere can be
as small or as large as you want. If it is very small (i.e.
approximates the volume of a Resel) this corresponds to an uncorrected
p value. If it is very large (i.e. approximates the search volume)
then the p values will be correspond to the corrected p values given in
the tables.
With best wishes - Karl
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|