Hi,
You can model the duration explicitly, i.e. use the three column format for
making the regressor where the second column equals RT.  Then you can bring
this regressor up to group level without doing anything special.

jack

On Fri, 6 Nov 2009 14:39:24 +0000, Shal Hat <[log in to unmask]> wrote:

>Hi,
>
>My subjects underwent a reaction time stimuli, where the stimulus duration was
>judged by their response time. I was wondering how I could do a group
analysis with
>that, since the stimuli duration will be different for each subject in the
group.
>
>Thanks !