If you would expect sustained activity to occur for the duration of
the target box, then entering this longer duration (your second
analysis) is the correct approach. So it's nice that the more
reasonable approach also seems to give you more robust activation! :)
On Fri, May 15, 2009 at 9:08 PM, Jiansong Xu User <[log in to unmask]> wrote:
> Dear All:
> We have acquired fMRI data with an ER design task. The task displays a cue
> on the screen for 1 s indicating the participants may win money if he/she
> can press a button while a target box is displayed on the screen. The target
> box will be displayed about 3 - 5 s after the cue disappeared. We predict
> that the brain regions associated with motivation should be activation by
> the cues.
> At first, I analyzed fMRI data using ER design, i.e., using 0 as duration of
> cue. The result is consistent with our prediction, i.e., activation in the
> ventral striatum.
> But, we believe that the the brain regions associated with motivation should
> be activated during the entire period between onsets of cue and target. So,
> I analyzed the fMRI data again using the interval between onsets of cues and
> targets (3 - 5 s) as duration of cues. The result showed much greater and
> more extensive activation than the first analysis.
> I'm wondering if my second analysis is reasonable?