Hi,
My experiment has one anticipatory period immediately followed by a stimulation period. I plan to set 4 EVs for each period, and each one of the 8 EV has a trial-by-trial behavioral measure as parametric modulator. I want to perform parametric modulation analysis to see which brain region covaries with behavioral measures in both periods.
I think if I simply want to see activation, this model would have a multicollinearity problem. But if I am only interested in parametric modulation results, does multicollinearity problem still matter?
Thank you
Manuel
|