Token economy programs are common place in many psychiatric facilities operating in the U.S. today. There is an increasing significance on “holding patients accountable for their behaviors, with level systems and earned privileges being integral components of treatment plans for many patients” (Glynn, 1990, p. 403). Many behavioral programs used are based on operant principles that compensate good behavior and penalize bad behavior are standard in many facilities. The use of the token economy has transformed mental health services, even though its use has been limited (Glynn, 1990).
Historical Overview of the Token Economy
In the 1960s and 1970s operant learning principles were widely used among behaviorists to help modify the behaviors of psychiatric patients. Lindsey and Skinner (1954) were the first behaviorists to show that operant learning principles could be used to treat the behaviors of these patients. Through their positive results many studies were conducted using operant learning principles. Even though these early studies did not center on treatment, they did demonstrate that “the laws of learning could apply in the clinical setting” (Glynn, 1990, p. 388). In a clinical study led by Ayllon and Azin (1965, 1968) the work performance of a group of female psychiatric patients multiplied due to the use of tokens as an incentive. They used reversal and multiple-baseline designs in this study. These were not the only behavioral techniques used by these researchers.
Using extinction and differential reinforcement proved that nursing staff could use these techniques to improve the rate of appropriate behaviors and reduce inappropriate behaviors. Their work was used as a basis for establishing a detailed token economy that was meant “to cover all aspects of the patient’s life” instead of certain behaviors (Atthowe & Krasner, 1968, p. 38). This study was conducted over a 20-month period, in which a baseline was established during the first 6 months, then 3 months of modeling, and the final 11 months of “active treatment on the token economy” (Glynn, 1990, p. 389). The researchers found an increase in attendance during group activities and a reduction in rule violations as a part of “token reinforcement” (Glynn, 1990, p. 389).
During the next 10 years, researchers conducted studies using similar research designs to show the advantages of token economy systems with other populations in psychiatric settings (Glynn, 1990). Research conducted by Gripp and Margo (1971) showed that a token economy ward improved the behaviors of schizophrenic patients in a state hospital compared to those patients on control wards. The staff evaluated the program as having a more positive working environment in comparison to the other ward staff.
Research conducted by Hersen, Eisler, Alford, and Agras (1973) showed that participating in a token economy improved the depressive behavior of unstable patients suffering with depression. A multiple-baseline across-behaviors design was used by Nelson and Cone (1979) to demonstrate that personal management, personal hygiene, ward work, and social skills are behaviors that can be improved by using a token economy program with acute psychiatric patients. Another study was conducted comparing patients on a traditional ward versus a token system ward. Results showed that patients on the token system ward had increased participation in activities, spent less time in bed, better hygiene, and a reduction in disturbing comments (Gershone, Errickson, Mitchell & Paulson, 1977).
Further research was conducted by Stoffelmayr, Faulkner, and Mitchell (1979) in comparing the outcomes of a token economy program and a “milieu-like social therapy” for “hard core” patients and discovered a more positive affect of the token economy system (Glynn, 1990, p. 389). Hofmeister, Scheckenbach, and Clayton (1979) showed that patients who were recently admitted earned discharge from the token economy within 2 months. During this study long-term patients who transitioned over from other wards obtained the same outcomes, but it took longer.
Fullerton, Cyner, McLaughlin-Reidel (1978) decided to look at the success rates of those patients who discharged back to the community after leaving a token economy system. The results showed that 72 percent of 125 patients ready for assessment were still successful 3 years after being discharged from the program. Token economies began to come under critical observation during the mid-1970s. The results of these economies were encouraging but not enough research using a “rigorous scientific method” had been done surrounding the token economy (Glynn, 1990, p. 390). These studies using a token economy used research designs that did not eliminate bias from using single-case and small-sample designs comparing pre-post behaviors of patients. These studies did not “randomly assign patients to treatments”, which skewed the results (Glynn, 1990, p. 390).
It was hard to recreate these studies due to a lack of detail. Reports were often missing diagnostic information, specific data on psychotic drugs being administered, and information on methods used in the token economy and alternative therapies. Researchers began to emphasize the need for a more rigorous scientific examination of the token economy for psychiatric patients. This would entail random assignment of participants to a token economy, collection of follow-up data to ascertain the continuation and generalization of all acquired outcomes (Glynn, 1990). Detailed attention to the specification of the methodology and vigilant observation to the intricate assessment of treatment outcomes would be essential.
Paul and Lentz (1977) directed an extensive study that met this criterion. The researchers randomly assigned 84 patients with schizophrenia to state hospital wards using “social learning, milieu therapy, or traditional custodial care approaches” (Glynn, 1990, p. 391). As patients were discharged new participants were added to the study, ending with a sample size of 102. The token economy in the social learning program had highly structured educational activities planned each day (Glynn, 1990). The milieu program had 9 to 10 person living groups modeled after a therapeutic community.
These groups would be assigned to recognizing any issues and encouraging change within the community by applying “social and group pressure” (Glynn, 1990, p. 391). The participants in the milieu group had to attend academic or life skill classes during the day. Participants in both programs were in traditional treatment 85 percent of the time. At the end of the study the patients that were in the social learning group had spent less time in the hospital, attained better discharge rates, needed less psychotropic medications, and were supported in the community longer than either one of the comparison groups (Glynn, 1990).
These contrasts were remarkable. Once all patients had been released, the rates of successful placement within the community were similar (Glynn, 1990). Even with such good results there were still limitations to the token economy that was reported as a small percentage of patients between 3 to 20 percent that did not react to the token economy. Kazdin (1983) gives his view of treatment failures and interprets the causation as being from flaws in the application of the token economy instead of its overall principles. Kazdin (1983) believes that these failures can be avoided by pinpointing effective reinforcers for each patient, establishing integrity of treatment, influencing the economy to decrease inflation, and including participants in the management and sharing of repercussions.
The level system is a behavioral method that modifies the status of participants dependent on their behavior; every status level varies in the extent of reinforcement (Hagopian, Rush, Richman, Kurtz, Contrucci & Crosland, 2002). There have been few studies conducted on the implementation of Level Systems; however, over the years level systems have been used by teachers to inspire and strengthen students with emotional and behavioral disorders to succeed in school (Scheuermann, Webber, Partin, & Knies, 1994). Participants in a level system program will advance from one level to the next more displaying appropriate behaviors and fewer inappropriate behaviors. The participants will receive access to more privileges