> I have an event-related design with 64 trials, each lasting for 8 3 sec TRs.
Each stimulus lasts 8 TR's? I'm confused: Would it be a block design?
> I have a stimulus function that lasts for 513 volumes, and each
> trial lasts 8 TRs. So I would like -iresp to output an impulse
> response function for each voxel that is 8 TRs long.
>
> e.g. 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 ...
How are you modeling the HRF's? Can you paste the 3dDeconvolve script?
> In the manual it says that -iresp will output a timeseries that is
> p+1 in length. So I would think that I would need to set
> -stim_npr 8. But all of my stimuli are TR-locked in their
> presentation and I do no have subthreshold TR timepoints.
-stim_nptr is useful only if you have event onset times locked with fractions of a TR.
>Do I need to use stim_nptr and then 3dTshift ? MUCH CONFUSION! Please help!
Doesn't sound to me you need such a setup.
Gang