Chapter 64. Schedules of Reinforcement

1. Real life does not provide continuous reinforcement for learned behaviors. Instead, behaviors are typically rewarded on a partial reinforcement schedule, in which responses are sometimes reinforced, and sometimes not. The principles of operant conditioning predict that the type of reinforcement schedule will influence the rate of learning and the frequency of the behavior.

2. Partial reinforcement, also called intermittent reinforcement, generally produces slower initial learning (also called acquisition), but the behavior is more resistant to extinction.

3. In real life and in the laboratory, partial reinforcement is administered according to either a ratio schedule (which counts the number of behaviors) or an interval schedule (which counts the passage of time).

4. Fixed-ratio schedules reinforce behavior after a set number of responses, while variable-ratio schedules provide reinforcers after an unpredictable number of responses. In this illustration, each response is shown in gray, and the reinforcers are shown as red squares.

5. With fixed-interval schedules, reinforcement is given for the first response that occurs after a fixed period of time has elapsed, while variable-interval schedules reinforce the first response after varying time intervals. In this illustration, each response is shown in gray, and the reinforcers are shown as red squares.

In the laboratory and in real life, behavior is usually reinforced intermittently, according to the four schedules listed above. The rat in this operant chamber (also called a "Skinner box") is being trained to press a bar; the training is reinforced with food pellets.

On the next pages, you will see four simulations illustrating how each of the schedules of reinforcement influences the rat’s behavior.

This rat is being reinforced with 1 food pellet for every 5 bar presses. That means the first time the rat presses the lever, nothing will happen—and the same is true for presses 2, 3, and 4. But on press 5, a pellet of food will slide down the dispenser. Then, the counter starts again, and the apparatus rewards the rat for every fifth press.

This rat is being reinforced with 1 food pellet for an average of 3 bar presses. Sometimes, the food pellet will appear the very first time the rat presses the lever, but at other times, the rat will be reinforced after press 2, 3, 4, or 5. Over the course of several minutes, the rewards will average out to one for every three presses—but the timing is unpredictable.

This rat is being reinforced with 1 food pellet for the first bar press during every 4-minute time period. The first time the rat presses the lever, a food pellet appears. But, no matter how many times the rat presses after that, the next reward will not appear until 4 minutes have passed. Then, the timer resets, and the cycle repeats.

This rat is being reinforced with 1 food pellet for the first bar press during a variable time period that averages 3 minutes. Sometimes, the interval will be 1 or 2 minutes, but at other times, it will be 4 or 5 minutes. The interval is unpredictable, but the rat must press the bar at least once during the interval in order to get the reward.