Association for Behavior Analysis International

This symposium covers extensions of using a multiple schedule to increase the practicality of function-based treatments. Topics include (a) methods for leaning the duration of the extinction component and whether systematic and gradual progressions are necessary, (b) the portability of schedule-correlated stimuli and the extent to which treatment gains generalize across individuals and settings, and (c) the effects of incorporating the stimulus correlated with the extinction component on the level of communication responses when used as the continuous signal during delays to reinforcement. Together, the presentations demonstrate how multiple schedules serve as an efficient and efficacious method to sustain acquired functional communication responses and low levels of problem behavior across different contexts and how the schedule-correlated stimuli can be used to enhance the efficacy of another schedule thinning method. In addition, we are fortunate to have Dr. Wayne Fisher serve as the discussant.

Previous studies indicate that teaching a functional communicative response in a multiple schedule arrangement provides an effective method for thinning reinforcement when attempting to produce a shift in responding in individuals from problem behavior to socially appropriate responses for the same reinforcer(s) (Fisher, Kuhn & Thompson, 1998; Hanley, Iwata, & Thompson, 2001; Tiger & Hanley, 2004, 2005). Although previous research has demonstrated beneficial effects of multiple-schedules for reaching terminal schedule criteria (Hanley et al., 2001), it is unclear whether the thinning procedures established to reach the schedule values are necessary. The purpose of the current study was to evaluate the: (a) effects of contingency-correlated stimuli during FCT, (b) necessity of systematic and gradual fading steps during schedule thinning under multiple schedule components and, (c) extent to which contingency-correlated stimuli facilitated generalization across therapists and environments. Participants were 4 individuals referred for severe problem behavior and whose basic skills set included ability to communicate vocally and follow multiple-step instructions. Each participant was exposed to mixed vs. multiple schedule arrangements with and without fading of lean and rich schedules of reinforcement. Results suggest schedule correlated stimuli are necessary and sufficient for FCT schedule thinning and they promote rapid generalization.

Studies have indicated that multiple schedules can produce discriminated manding while maintaining low rates of problem behavior shown to be maintained by the same reinforcer as those mands (Fisher, Kuhn, & Thompson, 1998; Hanley, Iwata, & Thompson, 2001; Tiger & Hanley, 2004, 2005). However, few studies have examined systematic methods of determining the duration of the components of the multiple schedule (i.e., periods SD and SDELTA) and subsequent schedule thinning. The purpose of the current investigation was to examine whether a multiple schedule in which the duration of the SD and SDELTA components were based upon the rate of mands during prior functional communication training (FCT) would produce discriminated manding while maintaining low rates of problem behavior. Participants were 4 individuals with developmental disabilities who engaged in severe behavior. The initial SDELTA interval was double that of the average interresponse interval for mands during FCT. This study also investigated whether discriminated manding would maintain when the schedule of reinforcement was the thinned to a terminal goal of a 10 min SDELTA. Overall, discriminated manding was observed with all 4 participants and manding maintained when the schedule of reinforcement was systematically thinned.

We used multiple schedule training to reduce high rates of manding and problem behavior emitted by individuals with developmental disabilities and autism during reinforcement schedule thinning procedures. Once mands were under discriminative control, a test for transfer of stimulus control to a signaled delay to reinforcement was implemented. During the multiple schedule training, rates of problem behavior were reduced to zero and manding was reduced to near zero during extinction intervals while maintaining steady rates of mands during reinforcement intervals. Following multiple schedule training, results showed transfer of stimulus control to signaled delays when mands during the delay interval decreased to near zero levels. Transfer of control from multiple schedules to signaled delay to reinforcement procedures may be a critical component of the treatment of excessive manding in natural environments when a variety of preferred items are available on varying schedules.