Hi Joe
I keep coming across posts in the FSL and SPM archives saying that including the RT as a
parametric regressor will account for effects related to the RT in tasks that have events that
are effectively defined by the RT. It seems to me that this might not be correct though and
I'm interested for your take on this... Mark Jenkinson has told me a number of times that
the problem with events defined by their reaction times is that the height of each different
trial's event differs in an unpredictable way, at least as modelled in fsl. That is, it seems
that a parametric regressor with value of the duration would make an untenable assumption
that the blood flow response, as modelled by FSL, increases linearly with duration. What do
you think?
Cheers
Christopher
|