If you want to evaluate the effects of noncontingent reinforcement on repetitive behavior, and you only have one client who engages in this behavior, what design would be best?