Association for Behavior Analysis International

The continuation of responding in the face of disruption (i.e., response persistence) is a desired outcome in many situations. For example, one goal of early intervention is to ensure that skills acquired in the clinic or home setting generalize and maintain when the student transitions to school. Conversely, persistent responding may be problematic if destructive behavior maintains when treatment is introduced. The presentations in this symposium focus on the factors that contribute to or mitigate against behavioral persistence in applied contexts. Dr. Joel Ringdahl will present on one way to increase the persistence of social skills exhibited by adults with developmental disabilities. Stephen Ryan will discuss reinstatement of appropriate communication responses following periods of extinction. Valdeep Saini will present on the effects of discriminability on persistence of responding during and following the use of time-based reinforcement schedules. Katherine Lichtblau will discuss differential levels of resurgence of destructive behavior following two forms of differential reinforcement.

Several studies in the basic, translational, and applied literatures have demonstrated that the frequency with which reinforcers are delivered (i.e., reinforcement rate) can impact responses persistence. Additional reinforcer parameters, including magnitude of reinforcement, have also been demonstrated to impact response persistence. However, studies relevant to this topic have been conducted in basic laboratories using nonhuman (e.g., Nevin, 1974) and human (e.g., McComas, Hartman, and Jimenez, 2008) participants. Thus, the applied impact is unknown. In the current study, the effect of magnitude of reinforcement on the persistence of socially significant behavior exhibited by adults with mild disabilities was evaluated. Initially, a BST package that included a reinforcement component was used to teach conversation and job interview skills to young adults with developmental disabilities. Performance was scored in terms of percent of steps correct, according to task analyses. The BST packages were paired with either high magnitude of reinforcement or low magnitude of reinforcement, depending on baseline performance (i.e., high magnitude of reinforcement for the skill performed with relatively worse skill during baseline). Following acquisition of the skills, the BST package was discontinued, resulting in termination of the reinforcement contingency. Results indicated that the skill taught using the high-magnitude reinforcer persisted at relatively higher levels during the maintenance period (4 weeks of follow-up) for each participant, though performance remained above criterion regardless of reinforcer magnitude.

Using a human-operant preparation and guided by Shahan and Sweeneys (2011) model of resurgence based on behavioral momentum theory (BMT), we evaluated a refinement to noncontingent reinforcement (NCR) designed to reduce persistence of problem behavior during NCR and mitigate response resurgence during extinction following NCR. The refinement involved increasing the saliency and discriminability of the change from contingent reinforcement (during baseline) to NCR, which BMT predicts should lead to faster reductions in target responding and decrease the likelihood of resurgence. We present data on four subjects, all of whom responded in ways predicted by BMT to varying degrees. Our results are discussed within a translational research framework and broader context of strategies used to mitigate treatment relapse for severe destructive behavior, as NCR is one of the most commonly prescribed interventions for destructive behavior displayed by individuals with intellectual and developmental disabilities.

Nevin and Shahans (2011) Equation 7 of behavioral momentum theory (BMT) predicts that when all other variables are held constant, the rate of alternative reinforcement (Ra) predicts the degree to which responding recurs when disrupted. In the current investigation, we used extinction as a disruptor to evaluate levels of resurgence of problem behavior following two forms of differential reinforcement of alternative behavior (DRA) in which overall rates of reinforcement were equated. In one condition, reinforcement remained available for problem behavior, as well as for a communication response (i.e., DRA without extinction), whereas in the other condition, DRA was implemented with extinction, but we added noncontingent reinforcement to yoke the obtained rates of reinforcement to the DRA without extinction condition. Both DRA conditions suppressed rates of problem behavior with one individual, but we observed higher levels of resurgence following DRA without extinction. We discuss these results in light of BMT, as well as other treatments for problem behavior that do and do not involve terminating of the response-reinforcer relation.