Preference for and Efficacy of Accumulated and Distributed Response–Reinforcer Arrangements During Skill Acquisition