Dear Daniel,
> I'm trying to model event-related activity that is due to two stimuli
> presented sequentially with an SOA of 1.5 seconds. In looking at the
> raw data, the hemodynamic response to such an event often has a wider
> peak than does the typical HRF for a single event. I'm wondering
> whether modelling the data using a dispersion derivative (i.e., hrf +
> time + dispersion derivatives) will enable SPM to fit a canonical hrf
> to the data that has the appropriate width. In other words, does the
> dispersion derivative allow the width of the of the canonical HRF to be
> adjusted analagous to the way that the temporal derivative allows the
> onset of the canonical response to be adjusted?
Absolutely.
Karl
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|