Association for Behavior Analysis International

Token reinforcement systems have endeared themselves in educational settings and other core domains of applied behavior analysis, including developmental disabilities, since the beginnings of the field (Ayllon & Azrin, 1968; Kazdin, 1977; Kazdin, 1978). In addition to providing opportunities to expand the scope and relevance of our field, such systems also provide an important set tools to behavior analysts studying the conceptual underpinnings of phenomena such as reinforcement, conditioned reinforcement, and punishment. The four papers presented in this symposium continue this tradition by demonstrating the usefulness of token systems in diverse applications. Donaldson, DeLeon, Fisher, and Kahng demonstrate the use of a token procedure to evaluate differences and preferences for gaining versus losing rewards. Russell, Ingvarsson and Haggar investigate the use of a token system in the context of behavioral economic procedures designed to contrast the value of different categories of reinforcers. Boyle, Samaha, and Bloom evaluate maintenance of responding under schedule thinning using fixed and variable schedules of token exchange. Finally, Parry and Anderson examine the effectiveness of a token reinforcement system augmented with self-managed check-in/check-out.

Token economies involve individuals working to earn tokens that are exchangeable for back-up reinforcers. Token economies produce meaningful changes in behavior, however an important issue related to token schedules is how best to fade the schedule of back-up reinforcement to practical and cost-effective levels. Although research shows that token economies continue to be effective during and after schedule thinning, little research exists regarding the best way to do so. The concept of unit price from behavioral economics predicts that responding will be the same given the same ratio of cost (e.g., ratio requirements) to benefits (back-up reinforcers earned), and treats fixed- and variable-ratio requirements the same in terms of cost. Despite this prediction, early research on variable-ratio schedules suggests that they maintain responding at higher ratio requirements than fixed-ratio schedules. Thus, the purpose of this study was to compare fixed- and variable-ratio schedules of reinforcement while increasing response requirements to earn back-up reinforcers (i.e., unit price) in the context of a token reinforcement system for compliance with academic tasks. Results showed no consistent differences between the schedules at any unit price.

Check-in/check-out is an evidence-based Tier II intervention used widely in schools. CICO is a Tier II intervention designed for use within a comprehensive, school-wide framework. Components of CICO include frequent feedback on social behavior and a token economy. We investigated a modification of this intervention, incorporating the use of self-management into check-in/check-out. The intervention included the following features: morning and afternoon feedback sessions with a mentor, self-monitoring of performance throughout the day, randomly scheduled accuracy checks, and rewards contingent upon earned points. We used an ABAB reversal design to assess experimental control. Across participants, the self-management intervention resulted in decreases in disruptive behavior and increases in academic engagement. Social validity data were mixed with students and parents rating the intervention as valued and the coordinator rating the intervention as not useful. Possible reasons for the mixed social validity will be discussed with an emphasis on systems-variables necessary for successfully embedding an intervention in the school culture.