On Test Selection, Prioritization, Bisection, and Guiding Bisection with
Risk Models