6/26/84, updated 6/29/01

Controlling Behavior Through Reinforcement

David A. Gershaw, Ph.D.

Once people do what you want them to do, how can you influence them to repeat their actions? The answer is simple reinforce them. A reinforcement can either be primary satisfying biological needs like hunger, thirst, etceteras or secondary learned factors such as receiving money, approval or various forms of attention.

However, regardless of the type of reinforcement used, when you give the reinforcement influences the response of the person. Reinforcement is most effective, when it occurs immediately after the desired response. With humans, the reward can be delayed to some extent, as long as the person perceives it as related to the desired behavior.

If behavior is not reinforced, it goes to extinction. This means that if the behavior does not continue to "pay off" it will not be continued. For example, if your boss stops paying you, eventually you will quit working.

How often do you have to reinforce behavior to keep extinction from occurring? Behavior does not have to be reinforced continuously every time the desired behavior occurs to keep it from going to extinction. However, responses are learned more quickly with continuous reinforcement. Unfortunately, continuous reinforcement leads to very rapid extinction, when the "payoff" stops.

In contrast, partial reinforcement getting rewarded for only some of the desired responses leads to slower learning, but the responses also take longer to go to extinction. There are at least three different schedules of partial reinforcement.

The first is a fixed ratio schedule of reinforcement. Although not reinforced for every response, this means that the organism is rewarded according to the number of desired responses that it performs. For example, a pigeon may be reinforced for every fifth peck at a dot. Once this response has been learned, it will be continued at a relatively high, consistent rate.

In humans fixed ratio reinforcement is related to working on commission or piecework. Like pigeons, humans will tend to show a high, consistent response rate and are relatively slow to go to extinction. When I was a small child, my father offered me a penny for every five weeds I dug out of our lawn. Not only did I do our lawn, but I also "stole" weeds from neighbors' lawns!

In contrast to fixed ratio, a fixed interval schedule causes a much different type of reaction. Fixed interval reinforcement is getting rewarded only after a specific time interval for at least a minimal response. With the pigeon, it would be like rewarding it at the end of each minute, if it pecked at least once during that time. In contrast to fixed ratio, the number of responses is not important after the organism has reached the minimum level. In this case, if the pigeon pecks once, 10 times, 50 times or even 1000 times, it still gets rewarded only at the end of that minute.

With humans, fixed interval is like working for a salary, whether you get paid by the hour, weekly or monthly. As long as you do the minimum required work, you get paid at the end of the interval. Doing extra work during the interval does not lead to extra pay.

In contrast to fixed ratio reinforcement,
a fixed interval schedule leads to
a slower response rate and procrastination.

In both pigeons and humans, a fixed interval schedule leads to a much slower response rate. Any increase in work (if at all) occurs near the end of the interval. For example, some parents tell their children, "If you are good all year long, Santa will bring you lots of goodies." This helps very little from January through November. However, for the last weeks in December, they typically have a "little angel" on their hands!

Similarly, both school exams and term papers are on a fixed interval schedule. Notice how many students wait until the last minute and then "cram" for their exams. Likewise, students do relatively little work on a term paper during most of the semester. They tend to rush through most of it in the week before it is due. Fixed interval reinforcement is a prime cause of procrastination. Of course, not all students respond this way, because exams and term papers are also rewarded on a fixed ratio schedule. A certain number of desired responses are needed to get a specific grade.

Next week, we will discuss random reinforcement the third schedule of partial reinforcement and its effects on gambling and superstition.

Go back to listing of additional articles.

Go back to "A Line on Life" main page.