Hi Tim,
No, you're right. As I understood the question it was about how to
describe the activation in meaningful terms given the large extent
and that's what I tried to focus on. I think Russ's method was
aiming at the same thing: namely, a method for describing regional
activation in terms of mean percent BOLD signal change rather than
evaluating significant activation.
The issue of whether the activation *in any given area* was
significant or not is a different one. If the interpretation depends
on this second point (i.e. if the story critically relies on
significant amygdala activation), then the cluster result that's been
described doesn't fulfill this criteria. In that case, Mike would
need either a specific a priori hypothesis regarding amygdala (or
whatever) activation and an appropriate small volume correction
(typically using voxel, rather than cluster, stats) or some other
method to demonstrate significant activation in that area.
Is that your understanding as well?
Cheers,
- Joe
|