On Mon, Feb 6, 2012 at 8:27 AM, Torsten Ruest <[log in to unmask]> wrote:

Hi there,

I'd like to generate some figures that indicate the change (in %) in activation before and after treatment within a given ROI. Each participant had 3 runs with the same ER task before, and 3 after treatment. These runs have been combined per session and participant with a 2nd level model. I'd now like to extract the cope values (contrast of interest is stimulus-baseline) of these combined runs, pool them into groups (3: patients, controls, no training controls), and compute the percentage change per group.

I followed discussions on similar topics using featquery and became aware of different scaling parameters. If I understood right, I could just take the copes of the combined runs for each participant (2nd level model, and thus scaled the same), average them, and divide by the contrast fix of 2 (probably not even important here). Then compare the activation for each group before and after; using excel or something.

If someone could confirm that there are no unaccounted scaling factors involved at the 2nd level, ie the combined runs are comparable between subjects (ie scaling to the same fixed maximum).

Thanks very much in advance,

Torsten