dear matthew brett.
thanks for your reply about my problem with the plot function. even i do
not really understand all the background and the differences between the
different approaches to get an impression about the time series of the
signal in one roi, i decided to stick to the marsbar method(the second
figure of my former question), whether it is correct or not. i will see
what the reviewers will say about that.
thanks for your (and liangs') help.