Hi Gang

OK I'll try the AM2 procedure.

1- But how do I set the stimfiles. I have 20 trials for each 8 blocks for each subject. Does each trial get a number related to the covariate (reaction time, RT) I want to account for ? would it be: time_of_trial * RT_for_that_trial
0*1.1
10*0.77
20*0.55
30*0.87
40*0.23
50*1.08...and so on for 20 trials??

So eah trial would get a different RT?

2- And what about testing for a quadratic effect also? Or should I just run 3ddeconvolve to acount for RT and then test for quadratic effect at group level?

3- The output will be 2 beta and 2 t-stat for each of the regressor, the mean and the mean*AM? The 1st one is the same as if I had used the -stim_times option?

pat