What to make of the human data?

Very low p values at long time windows     Suggests pilot and first officer average behaviors differ     Suggests simple Markov model is a poor description