afni.nimh.nih.gov

Hi Simon,

It would not be so appropriate for 3dTcorr1D to output p-values in any case, since ther
e are surly degrees of freedom lost that it does not know about. One degree of freedom
should be subtracted for every regressor that has been projected out, including for ba
ndpassing (2 regressors per frequency). Consider this older post.

- rick