How we change what others think, feel, believe and do

Reward Extinction

Techniques > Conditioning > Reward Extinction

Description | Example | Discussion | See also

Description

A way to help stop an unwanted action is to take a two-stage approach:

1. Reward the unwanted action. Give them something they like every time they do it.
2. When the reward is confidently expected, suddenly stop. Do not reward at all.

If the action increase after stopping rewarding, ignore this completely. Have patience and determination. Just wait.

Example

A dog barks. The owner feeds the dog every time this happens. Soon, the dog is barking to get food. Then the owner just stops rewarding the barking, completely ignoring it. At first, the dog barks more, but eventually it gives up and stops barking.

Some boys play soccer loudly in the street outside a person's house. The person comes out and gives them some money, saying they it is lovely to see them play so well. The next week, the person claims they are rather short of money and gives them less. The boys complain, but still play. The following week, the person says they have had enough of paying and want to watch the boys play for free. Disgusted, the boys take their ball and go to play somewhere else.

Discussion

While reward extinction may seem unusual and possible problematic, it does add another possible method when the subject is not responding to more common means of encouraging cessation. It works by making the reward extinct and, by association, the unwanted action.

In more detail, this method works by the principle of extinction, where a non-reward leads to non-action. It amplifies this effect by making the unwanted action more pleasurable in anticipation of the reward. When the reward disappears, the anticipated pleasure turns to anticipated disappointment, making the unwanted action unpleasant.

See also

Extinction