>>For trials modelled by their duration, the parameter estimate reflects the response
>>*per unit time*.
>
> What is the *unit time*? Is it the shortest duration and then it is multiplied linearly, such that an
> event that lasts 0.5 seconds should have half the height of something that lasts 1 sec? or does it
> mean that the event is "happening twice" with the same thing happening the 1st half of the trial
> and then again the 2nd half of the trial? I'm a bit confused as to what this means.
The latter. The height does not change precisely linearly with duration, but that
doesn't matter: the response is simply exactly what you would expect from a
linear convolution of either a 0.5s or 1s boxcar with an HRF.
> My paradigm is complex and is actually a memory paradigm. The visual stimulus turns off when
> the subjects complete the different memory tasks: prediciting recognition performance,
> recognition, and evaluating recognition performance. I worry that the parametric modeling would
> get complicated since i'm interested in reaction time across process and within performance and
> was hoping the durations might be simpler, but i'd like to get a clearer idea of what it is doing.
Go with different durations then.
Best wishes
Rik
|