Dear SPM users,
I have a question concerning a correct way to compare different basis
functions on a single-subject level in fMRI.
I have two basis functions (A and B) and I would like to know which one
models the brain function better. So to test this I created a design matrix
(using SPM5) with two sessions and entered the same data in both sessions.
Then in first session I used the basis function A, and in the second
session, the basis function B. To test the difference I then created two
T-contrasts to see the difference in both directions (A>B and B>A).
My concern here is that I'm not sure if this is statistically valid
comparison? After all, I have the same data there twice. If it's not valid,
any ideas how to compare these two basis functions? I have done ROI analysis
and checked this on voxel time-series level, but I would like to test this
on statistical map level as well.
Any thoughts on this are appreciated!
Dept. of Clinical Neurophysiology
Kuopio University Hospital