> the stim file that I have is a list of time points in milliseconds (e.g. 997.7) Is
> this the right format? Should it be in whole index values as displayed by AFNI?
There are two options for feeding stimulus regressors. One is -stim_times with which the user is supposed to provide the stimulus onset times in seconds (not milliseconds) AND the basis function(s) for modeling the hemodynamic response. The other is -stim_file that expects the user to provide the regressor directly.
> when using the stimtime file what rtype would be the best to use for
> correlational analysis?
I don't have the big picture of what you're trying to do and what kind of experiment you have, so I can't say much about this.
Gang