Hi spmer's Please advise me on this issue.....I have a mixed design (block+event) which takes over 45 mins with 3 runs, Now the no. of events lets say are of two types, type A/response A and type B/respone B. The constraint with my task is that the no. of anticapted events of type A/response A are maybe 20 and type B/response B are only 10 or even lesser, are these numbers too small for event related analyses (signal averaging)? The number of subjects are 20 each for controls + patients. Can this help, if average the events across the whole group? Thanks, Manish Dalwani Professional Research Assistant Dept. of Psychiatry UCHSC __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com