Performance Management: Changing Behavior That Drives Organizational Effectiveness (part 25)

Schedules of Reinforcement and Extinction

by

Charles Lamson

Just as performance differs while each schedule is in effect, so it does during extinction. When you decide to use extinction, you are making a clear decision to withhold reinforcement and to choose this path for behavior change rather than a more direct route of correction or clear feedback and reinforcement for other behavior. It is a value decision you make about the relative choices you have in arranging conditions for employee improvement. Sometimes, extinction is the most benign and appropriate. Sometimes it is not. Consider that. Then, once you have decided to use extinction, as described below. When you stop reinforcement completely, you are using extinction.

What happens to performance under extinction depends on the behavior's history of reinforcement on the prior schedule. For example, there are rather unique distinctions between between using extinction with behavior under the control of positive reinforcement (R+) as compared to negative reinforcement (R-). While we cannot cover all that there is to say about the differences, do keep in mind that if a behavior has been under control of negative reinforcement, it is harder to know if extinction has occurred, because in the absence of aversive events, the behavior will not occur. Just because a person does not exhibit phobic behavior during some period of time does not mean he has been cured of the phobia. Rather he has not been in the presence of the feared situation. The classic characteristics of extinction do not occur with negative reinforcement. With that said, four things often happen when a pattern of behavior undergoes extinction.

Effects of Extinction

1. Initially, the behavior will increase. This is technically known as extinction burst.
2. Negative emotional behaviors, such as verbal abuse, tantrum-like behavior, and even aggression, occur.
3. The rate at which the desirable or undesirable behavior decreases will vary widely depending on the schedule, but be prepared for a gradual decrease.
4. Once the behavior has stopped, it will likely recur later, even though it has not been reinforced after extinction began.

The first three characteristics describe quite well what happens during extinction on a continuous schedule (FR 1 - The number in FR 1 refers to the number of behaviors required for reinforcement. FR 2 would mean every other behavior would be reinforced; FR 4 would mean every fourth behavior would be reinforced.). We have all heard people make an assessment of someone with a behavioral problem such as, "He will have to get worse before he can get better." They are referring to an extinction burst. An extinction burst occurs when, early in extinction, there is a dramatic increase in the behavior of concern in an attempt to get the usual reinforcement. Unfortunately, if the attempt is successful (gets reinforced), the rate of the behavior will be more difficult to extinguish again. The interval between the first attempt at extinction and the subsequent reinforcement, in effect, signals a delay. This makes the likelihood that new efforts to extinguish behavior will have to be applied for longer periods of time.

For example, in trying to extinguish behaviors like temper tantrums, many parents have noted that the tantrums actually got worse. This often occurs because when the child screamed longer than usual or turned bluer than usual, the parents gave in and thereby reinforced the child for that longer response. This will, of course, make extinction of the tantrum behavior more difficult. It also makes it more unlikely that you will be able to ignore it in the future.

A person who has been getting attention for complaining about co-workers may complain more frequently if his manager suddenly ignores his complaints. The manager who doesn't understand that extinction bursts are a predictable aspect of extinction may think that ignoring isn't working, when in fact it is. It should be pointed out again that extinction is not always the best response to a problem like a complaint. It could well be that telling the person "I do not like it when you talk about your co-worker in that manner" may be enough to stop the behavior in your presence. When combined with reinforcement for positive statements about co-workers, it might well be that this would be faster and appear fairer to the performer than a process of extinction that may take awhile.

Emotional behaviors are more likely to accompany extinction on fixed schedules. The length and intensity of the emotional reaction is a function of the schedule and number of reinforcers received on it. Everybody has witnessed at least a mild form of this phenomenon. When a person puts money in a vending machine several times, only to have it returned each time, he may try to shake the machine, or even hit or kick it before giving up. That person is exhibiting emotional behavior characteristic of extinction under fixed fixed ratio (FR) schedules. Parents who say things to their children like "You can cry all you want but you cannot go" are dealing with the same problem.

When people are accustomed to receiving reinforcement on some predictable pattern and suddenly it stops, you can expect an emotional response from them. These responses may range from complaints of "That's not fair" to tantrum-like behavior. These reactions occur at work just like they occur at home. They are easier to recognize in children, but if you look, you will see adult versions at work. With variable schedules there typically is considerably less emotional behavior during extinction.

In regard to Item 3 in our list of effects of extinction, the rate of extinction of behavior previously reinforced under FR 1 is faster than the other schedules. Variable schedules take much longer. Factors affecting the rate of extinction of a behavior under a particular schedule include the size of the ratio or length of the interval, and the number of reinforcers received on the schedule. In any event you might not see an immediate change in the behavior under variable schedules and should be prepared for a gradual, rather than sudden, decline.

A predictable characteristic of extinction that frustrates most people is the phenomenon of resurgence. Resurgence is the re-emergence of previously extinguished behavior. After a period of time when the behavior has not occurred, it often reappears for no apparent reason. In other words, if the new behavior does not get sufficient reinforcement, the old behavior will return. What happens to the old behavior when it returns is critical, since any reinforcement will put the behavior on a much thinner variable schedule and make it more difficult to extinguish in the future.

This occurrence is what has led many people to conclude that you cannot really change people. Their evidence for this is that people may change in the short run but they seem to always go back to their old ways of doing things. They are witnessing a common occurrence in extinction - resurgence. To prevent resurgence, the new behavior must get more reinforcers than are available for the behavior to be extinguished.

Knowledge of the four factors will help you recognize when extinction is occurring. Extinction bursts, emotional behavior, resurgence, and a reduction in performance are all fairly easy to spot. On investigation, if it appears that extinction is the problem, your knowledge of reinforcement will allow you to remedy the situation easily and quickly in most cases.

In addition if you want to extinguish some undesirable behavior, knowing the signs will allow you to anticipate them. If they are anticipated, you can extinguish the behavior in the most efficient manner.

*SOURCE: PERFORMANCE MANAGEMENT: CHANGING BEHAVIOR THAT DRIVES ORGANIZATIONAL EFFECTIVENESS, 4TH ED., 2004, PGS. 233-235*

end