Thanks, I see, now. I misunderstood you.
So you are treating the 9 Time 1 coefficients as a single series of
numbers, ditto the Time2 coeffs?
I think 3d Tcorrelate should work for that, although you might
have to fool it into thinking your 9 stat subbriks are time steps
using...er...prolly 3dTcat or 3drefit.
I'm also just rolling the idea of correlation of numbers where the
sequence is defined but not really meaningful, like your 9 subjects
probably are. I don't think there's a problem with that, I just hadn't
thought of it before.
rich