Re: [Marsbar-users] building contrasts from signal change

[Matthew B. <mat...@gm...>]

Hi,

On Mon, Apr 15, 2013 at 9:54 AM, durden10 <dur...@gm...> wrote:
>
> Dear Marsbar experts,
>
> As an avid user of marsbar, I was wondering if it is possible to recreate
> contrasts of events during the extraction of signal changes. I have N
> subjects, the design of each subject contains two regressors, lets say event
> A and event B. I know how to extract the signal change of a certain ROI for
> event A and B, but since my 1-level analysis inlcudes a contrast A-B, I was
> wondering if this contrast can be re-created in marsbar using linear
> combinations of events... Otherwise I would extract signal changes for event
> A and event B and then later combine the two (simply A-B). I know this is
> probably a trivial question, but I was wondering if someone could give me
> some information about this.
>

Sorry to be slow to reply.

No - sorry - there isn't any machinery to do this automatically in
marsbar.   I guess what the interface would have to do would be to
have an 'event contrast' specification rather than a regressor
contrast as in SPM - but I haven't written that.  If it were me, I'd
probably write a small batch script to do that.

Cheers,

Matthew