Posted by Hackjin Kim on November 19, 2002 at 02:07:25:
Doug,
I am analyzing a blocked fMRI data (with each stimulus block turned on for 8 TRs) using 3dDeconvolve. I created 2 TR lags for each block regressor but generated linear contrast (LC) map only for the second lags (h2) taken from different stimulus blocks. I am wondering if we can still trust this LC map without concerning about multicollinearity problem that may arise from overlapping between lags. If not, would you please suggest me what could be the optimal way to create LC maps in case of blocked design analysis.
Thanks,
Hackjin.