hi eugene,
right. i re-ran the analyses using PEs 1, 3, 7 for conditions A, B, C respectively (i had mistakenly scripted PEs 1, 2, 3).
i still find that PE = COPE for condition A (EV 1). in contrast, the PE for condition C (EV 3) equals the COPE for condition B (EV 2), for every subject, for every run.
the new wrinkle is that i've now reproduced this swapping effect for a new set of PEs and COPEs.
any thoughts?
thanks!
libby
|