|Basic and Applied Research on Token Schedules
|Saturday, May 24, 2008
|1:00 PM–2:20 PM
|Area: EAB; Domain: Basic Research
|Chair: Jason C. Bourret (The New England Center for Children)
|CE Instructor: Jason C. Bourret, Ph.D.
Token schedules have been extensively examined by both "basic" and "applied" researchers. However, these two research traditions have developed largely independently of one another. This symposium is an attempt to foster contact and highlight research on tokens that tells us something new about the schedules and also has important implications for their application.
|Saving and Spending in a Token-Accumulation Procedure with Pigeons.
|RACHELLE L. YANKELEVITZ (University of Florida), Timothy D. Hackenberg (University of Florida)
|Abstract: The present research concerns reinforcer accumulation by pigeons in a token-reinforcement context. In a standard token-accumulation procedure, three keys are arranged: token-production, on which responses illuminate lights (tokens); exchange-production, on which responses produce access to the exchange key; and exchange, on which responses produce one unit of food per earned token. At any point after producing one token, pigeons choose to continue producing tokens or to produce exchange. Previous research has shown that reinforcer accumulation is inversely related to the cost of producing reinforcers and directly related to the cost of producing exchange, but the relationship between accumulation and the exchange schedule itself remains uninvestigated. In the current procedure, the costs of producing tokens and producing exchange were held constant while the cost of exchanging each token was varied from 1 to 150. The extent to which accumulation frequency (percent of cycles with accumulation) and magnitude (mean number of tokens accumulated per cycle) were systematically related to exchange ratio was investigated. Results are interpreted with respect to unit price, a composite measure of the costs (ratio size) and benefits (reinforcer amount) associated with different courses of action.
|Comparison of Responding Under Token and Tandem Schedules in a Clinical Setting.
|JONATHAN SEAVER (The New England Center for Children), Jason C. Bourret (The New England Center for Children)
|Abstract: Relatively little work has been done evaluating the effects of second-order token schedules on response rates and patterns in clinical settings. In the current study, patterns of responding on second-order token and tandem schedules of reinforcement were examined with two participants. All participants were students at a residential school for individuals with developmental disorders. Token schedule values were similar to those used typically in clinical application. Response requirements on the tandem schedule were yoked to the number of responses required for token exchange on the second-order token schedule. Results showed longer mean durations to complete the initial components relative to successive components on both tandem and token schedules. Response rates maintained on both schedules were equivalent. Implications in terms of expected response patterns on token schedules used with humans in a clinical setting and overall clinical benefits of the use of token schedules are discussed.
|Conditioned Reinforcement: Schedule Thinning Pilot Study.
|CHERYL LYNN CAMASSA (The New England Center for Children), Daniel Gould (The New England Center for Children)
|Abstract: Token reinforcement can be thought of as a second-order schedule in which responses produce tokens according to one schedule (production schedule) and then the opportunity to exchange tokens for other reinforcement occurs after a specified number of tokens has been earned (exchange schedule). This pilot study explored thinning of schedules of conditioned reinforcement by examining production and exchange schedules. Two participants with autism were exposed to reinforcer assessments using various combinations of production and exchange schedules. For example, in the FR1?FR4 schedule [FR4(FR1) second-order schedule], the participant received a token after each hand-raising response, and after 4 tokens were earned they were exchanged for a backup edible reinforcer. This schedule was compared in an alternating treatments design to FR4?FR1 [FR1(FR4) second-order schedule], in which the participant received a token after 4 hand-raising responses, and after 1 token was earned it was immediately exchanged for a backup edible reinforcer. Schedule comparisons included FR1(FR2) vs. FR2(FR1); FR1(FR4) vs. FR4(FR1); FR1(FR8) vs. FR8(FR1); and FR1(FR16) vs. FR16(FR1). Results showed little systematic difference at lower schedule values. At higher schedule values [FR1(FR8) vs. FR8(FR1) and FR1(FR16) vs. FR16(FR1)] there was a higher rate of responding when the production schedule was FR1 as compared to FR8 or FR16.
|Punishment in Token-Based Procedures for Treating Addiction.
|JOHN M. ROLL (Washington State University)
|Abstract: Token-based (i.e., contingency management) interventions are quite effective at initiating and maintaining abstinence. These procedures are generally believed to rely solely on positive reinforcement. However, most do include a punishment contingency for failures to abstain. In this presentation data from several studies will be presented demonstrating that this punishment enhances the efficacy of the procedures. Additional data will be presented showing that punishment contingencies may not be appropriate unless they are combined with strong reinforcement procedures as individuals in treatment will “escape the punisher”. An exception to this may occur when treatment is provided in a criminal justice context in which “escape” is not an option. This will be illustrated with clinical data from several recent trials.