Association for Behavior Analysis International

The Association for Behavior Analysis International® (ABAI) is a nonprofit membership organization with the mission to contribute to the well-being of society by developing, enhancing, and supporting the growth and vitality of the science of behavior analysis through research, education, and practice.


32nd Annual Convention; Atlanta, GA; 2006

Event Details

Previous Page


Symposium #323
CE Offered: BACB
Token Reinforcement Systems: Investigations on the Value of Tokens and the Selection of Back-Up Reinforcers
Monday, May 29, 2006
10:30 AM–11:50 AM
Area: EAB; Domain: Applied Research
Chair: Iser Guillermo DeLeon (Johns Hopkins University)
CE Instructor: Iser Guillermo DeLeon, Ph.D.

In both basic and applied behavior analysis, the consequent stimuli that maintain behavior often derive their reinforcing function from their association with stimuli that have previously been established as reinforcers. Termed conditioned reinforcers, these stimuli play a crucial and ubiquitous role not only in our understanding of learning processes, but in the application of behavioral principles to the solution of socially-relevant behavioral problems, perhaps most conspicuously in the establishment of token reinforcement systems. The present symposium will attempt to expand further our knowledge of the factors that modulate the efficacy of tokens and other behaviorally neutral stimuli as conditioned reinforcers. Specifically, the presentations will discuss (1) token value as a function of response effort, (2) changes in efficacy during the conditioning process to establish tokens as reinforcers, (3) the correspondence between preference assessment outcomes and the selection of back-up reinforcers and (4) direct comparisons of response maintenance between primary and token reinforcers.

Sensitivity to Token Loss as a Function of Earning Requirements.
LISA M. TOOLE (Kennedy Krieger Institute), Iser Guillermo DeLeon (Johns Hopkins University), David M. Richman (University of Maryland, Baltimore County), Gregory A. Lieving (Kennedy Krieger Institute), Melissa J. Allman (Kennedy Krieger Institute)
Abstract: The purpose of the current investigation was to examine the influence of earning requirements on the differential value of token reinforcers. Subjects participated in two phases. In Phase 1, experimental subjects were required to earn 20 tokens by identifying target stimuli in a field of stimuli. Control subjects were simply given 20 tokens (no earning requirement), one at a time, while the times of gains and losses were yoked from an experimental subject. All subjects then had the opportunity in Phase 2 to gamble with their tokens during discrete trials that each could result in a win or loss. The probability of a loss during these trials increased systematically across blocks of 10 trials. Of primary interest was the number of trials in Phase 2 at which the participant opted to escape from the task (i.e., cash out tokens). Generally, those who earned their 20 tokens in Phase 1 spent less time in Phase 2 and opted out sooner than those who received free tokens. These results are discussed in terms of the possible direct, positive relation between the effort required to earn tokens and the subsequent value of those tokens, particularly as value relates to the efficacy of response cost.
An Assessment of the Reinforcing Efficacy of Tokens.
STACIE L. FITCH (New England Center for Children), Jason C. Bourret (New England Center for Children)
Abstract: Participants were 5 children diagnosed with autism. All sessions were conducted in the children’s classroom. Participants were exposed to three conditions (contingent tokens, contingent edibles, and extinction) in a multielement design. Sessions ended following 5 min or following the delivery of the number of reinforcers that matched the number of tokens that the subjects typically earned prior to token exchange. In the contingent tokens and contingent edibles conditions, consequences were delivered according to a random interval (RI) schedule which was arranged such that the number of arranged reinforcers in the session matched the number of tokens that the participant typically earned before exchange in his daily programming (e.g., if a participant usually exchanged tokens after the delivery of 10 tokens, an RI 30-s schedule was used such that, on average, 10 tokens or edibles would be arranged for delivery contingent upon responding). The edibles used were those that typically served as back-up reinforcers for the tokens earned by the participants outside of experimental sessions. The tokens used, were also those already in use for each participant. The data indicated that the tokens served as reinforcers for two participants but not for the other three. The results suggested that the methodology used in the study may be useful in identifying the reinforcing efficacy of tokens currently being used for behavioral programming.
Token Training and Motivating Operation Effects on the Outcomes of Preference and Reinforcer Assessments.
CARLY MOHER EBY (New England Center for Children, Northeastern University), Daniel Gould (New England Center for Children), Richard B. Graff (New England Center for Children), Jason C. Bourret (New England Center for Children)
Abstract: Preference assessments have been shown to be effective procedures for identifying potential reinforcers. Typically, the stimuli included in preference assessments are limited to edible or activity items, rather than conditioned reinforcers (e.g., tokens). The present studies made use of preference- and reinforcer-assessment procedures to evaluate the reinforcing efficacy of tokens as they were being established as conditioned reinforcers (Experiment 1). Three participants underwent a series of paired-stimulus preference assessments with edibles and tokens. In token training, novel stimuli were paired repeatedly with either the participants’ highest- or lowest-preferred edible items. Reinforcer assessments were carried out using a multi-element design across conditions of baseline, novel tokens, and conditioned tokens. Following this, the effects of motivating operations on the reinforcing effectiveness of tokens was tested (Experiment 2). The reinforcing effectiveness of tokens and edibles was assessed in an ABAB design. Controlled-access and satiation conditions were compared. Results for Experiment 1 indicated that the tokens were effectively established as conditioned reinforcers. Results for Experiment 2 indicated that reinforcer effectiveness decreased for both edibles and tokens during satiation relative to the controlled-access condition.
Correspondence Between Preference Assessments and Actual Item Selection.
KIMBERLY SLOMAN (University of Florida), Timothy R. Vollmer (University of Florida), Jorge Rafael Reyes (University of Florida)
Abstract: We conducted a study to evaluate the correspondence between various preference assessments and actual item selection for highly verbal adult males with developmental disabilities. First, we conducted a verbal preference assessment with items in the unit token store and compared the results with actual items purchased. Next, preference assessments were conducted using picture representations of the items and correspondence was again assessed. Finally, preference assessments using the actual items were conducted. There was little to no correspondence between any of the preference assessment methods and actual item selection. Additionally, there was little correspondence between assessment types (e.g. verbal compared to picture preference assessment). However, there was high to perfect correspondence between token store selections across months. Implications for reinforcer assessments with individuals who have extensive verbal repertoires will be discussed.



Back to Top
Modifed by Eddie Soh