Dear Simone,
I am afraid not. Could you not simply model everything
with hrf and time derivative and ignore the derivatives
for condition A?
An alternative would be to use hrf for A and B but add
a third trial-type C, at B onsets + 1s. The hrfs for B and
C will cover roughly the same space as hrf, with time derivative
for B (i.e. C is a late component of B).
I hope this helps - Karl
>SPM02 assumes that the basis functions are the same for all sessions.
>However, our paradigm asks for modelling different conditions with different
>kinds of basis functions. Is there some possibility left to define different
>basis functions for different conditions?
>e.g. condition A = hrf
> condition B = hrf, with time derivative
>
>Thanks for any suggestions!
|