Hi Beate,

Right, those 3dWarp commands do not affect the EPI, so it does not
matter whether you run them before or after 3dTshift.

You are saying there is a difference in the statistical maps, even
for a single paradigm, depending on the anatomy is processed?

Since 3dWarp affects only the anatomy, the only way for the EPI to
be affected is via any 3dAllineate or @auto_tlrc commands. So we
would need to know the difference in how you are using those
programs for the 2 cases. Feel free to state your 3dAllineate and
@auto_tlrc commands.

My question is not whether @auto_tlrc is before or after 3dWarp,
but whether @auto_tlrc is before or after 3dDeconvolve. Again, we
would need to see those commands.

At any rate, seeing slight differences based on how warping is
done is no surprise. But it must be based on that, somehow.

- rick