We recently carried out an MRI study and (to keep it simple) we compared two conditions A and B. We found two regions that preferred A to B, region 1 and 2, respectively. I was keen to explore the time series a little further. So what I did was take the A-B difference for each subject in each of the regions and compute the 95%CI for each 2 sec interval from the start of the stimulus presentation. We get a difference time series for each region and for each point we have a 95%CI. I have plotted the CIs for each point on each function (one for region 1 and one for region 2). Now I note at which time point the CIs reflect a sig diff (i.e., dont cross the x axis) and what I find is that the critical difference occurs earlier for region 1 than 2. At last a window on the time course of psychological functioning? I, of course, now want to claim my Nobel Prize but suspect this is completely flawed. Please advise. Philip. ******************************************************************** Philip Quinlan E-Mail: ptq1(a)york.ac.uk Department of Psychology FAX: (01904) 433181 The University of York Tel: (01904) 430000 Ext. 3135 Heslington Direct : (01904) 433135 York YO10 5DD U.K. ********************************************************************