Token Economy

Description of the Strategy

The token economy was initially developed by Nathan Azrin and Ted Ayllon in the 1960s to motivate mentally ill patients (e.g., schizophrenics) to perform prosocial behaviors. In the token economy, individuals are provided tokens consequent to performing behaviors that are targeted for improvement. Earned tokens may then be exchanged for various reinforcers, which are provided by individuals who maintain the economy (e.g., spouse, staff). Thus, token economies eliminate unnecessary time delays between performance of target behaviors and delivery of reinforcement.

There are several tasks that must be completed before a token economy can be implemented. First, problem behaviors must be identified by observing the patient in “real-world” settings, as well as from interviews with significant others, staff members, and the patient. Without a clear definition of behavior, there is no assurance of proper dissemination of tokens, potentially resulting in a lack of behavioral improvement. Therefore, it is then important to clearly define each target behavior to facilitate agreement of its occurrence among staff and patient. When a group is the focus of behavioral change, similarities and differences in problem behaviors of group members, as well as personal characteristics, must be considered. Generally, it is easiest to employ token economies with groups who share similar problem behaviors.

After target behaviors have been clearly defined, the number of tokens to be delivered to the patient for successful completion of a response must be determined. In general, performance of desired target behaviors that are observed to occur infrequently prior to the token economy should result in more tokens than less desired target behaviors. Other considerations in the specification of token contingencies include the patient's ability. If the patient lacks the ability to perform the target behavior, the individual should first receive tokens for participation in practice exercises to enhance the individual's behavioral repertoire. As skills develop, standards by which the patient may earn tokens should gradually increase. Quite aside from the decision of how many tokens to provide is the issue of whether to consequence negatively occurring behaviors (e.g., hitting a staff member) by taking away tokens (response cost). If the response cost method is employed, the number of tokens deducted must be significantly lower than the number received.

The time interval between token receipt and exchange of reinforcement is also an issue that will need to be considered. Indeed, an effort should be made to deliver the tokens as soon as the target behavior occurs. Similarly, the patient should be provided an opportunity to exchange earned tokens for the various reinforcers as soon as possible. Public posting of tokens earned and rewards exchanged will be helpful in the monitoring process.

Tokens and other materials (e.g., data sheets to record each patient's behaviors, tokens earned, various reinforcers) must be purchased prior to implementing the system. Tokens must be easily transportable and tangible, such as poker chips, stamps, or checks on a chart. They should also not be too large, easily damaged, or easily reproduced by patients. Tokens can be color coded to indicate specified behaviors (e.g., green tokens reflect performance of social skill behaviors, red tokens reflect performance in leisurely activities), allowing staff member to assess what behaviors are being improved and what behaviors still require intervention.

...