Fixed-interval and fixed-ratio reinforcement schedules with human subjects