Association for Behavior Analysis International

Restricted and repetitive behavior is among the diagnostic characteristics of autism spectrum disorders (ASD). To the extent that the behavior of individuals with ASD can be conceptualized as problems of invariability, our understanding of environmental variables that influence restricted and repetitive behavior and methods of increasing variability may be informed by basic and applied literature on response variability. Slocum et al. compared levels of rigid behavior in groups of individuals with and without an ASD. Following the group comparison, a percentile schedule of reinforcement was used to treat rigid and inflexible behaviors within the ASD group. In their first study, Peterson, Rodriguez, and Pawich compared the effects of modeling rote versus variable responses during the teaching of intraverbal categorization. The effects of programming lag contingencies on response variability were later evaluated within a second study. Caccavale, Lechago, and Sweatt demonstrated how lag schedules could be used to increase variability in greetings. Finally, Gayman et al. targeted the appropriate use of mands frames while increasing the variability in the number of mand frames used for three participants with ASD. Dr. Allen Neuringer, the leading researcher on response variability, will serve as the discussant.

Autism spectrum disorders (ASD) are defined, in part, by behavior that can be characterized as restricted and inflexible. Such behavior is exemplified by the so-called "higher-order" restricted repetitive behaviors characterized by their insistence on sameness or resistance to change. These behaviors can significantly interfere with opportunities to develop functional behaviors and more complex repertoires. The current study was conducted in two parts. The first study compared the level of rigid behavior of a group of 20 individuals who are typically developing with the behavior of 20 individuals who are diagnosed with ASD. Following that group comparison, the second study involved the treatment of those rigid and inflexible behaviors within the ASD group using a percentile schedule of reinforcement. We treated both within-activity and between-activity rigidity. To date, we have been able to demonstrate the effectiveness of this treatment for 4 out of our 5 subjects.

Children with autism spectrum disorders often require direct, systematic instruction to learn new skills (e.g., Discrete-trial instruction [DTI]). DTI has been criticized for producing rote responding (e.g., Cihon, 2007). Over the course of a DTI program, a single appropriate response (e.g., "hello") may be selectively strengthened to the exclusion of other appropriate responses ("hi", "howdy","good day"; Lee, McComas, & Jawor, 2002). In the first of two studies, we assessed the effects of having the therapist model variable versus rote responses (using a progressive prompt delay) on response acquisition and variability of intraverbal-categorization responding during DTI. For two of the four participants, acquisition was slower in the variable relative to the rote prompting condition. For all participants, any initial variability observed decreased during treatment in both conditions. In the second study, we evaluated the effects of adding a Lag-1 contingency to the variable-model condition on increasing variability. Variability increased for all four participants with the Lag-1 schedule but only after the therapist modeled variable responding using a progressive-prompt delay. Results are discussed in terms of improving the lack variability that can occur with DTI.

The results of the current study extend the literature on lag schedules of reinforcement and behavioral variability by demonstrating that lag schedules of reinforcement were effective in increasing variability in greeting responses. Our participant was an 8-year old boy diagnosed with autism. There was little variability in responding during baseline. We taught him six new greeting responses during a second baseline condition to demonstrate that teaching new responses alone was not sufficient in promoting variability in responding. Three lag schedules were introduced (Lag 1, Lag 2, and Lag 3) to promote emission of four or more greeting responses. There was a corresponding increase in the number of different responses with the introduction of each lag schedule of reinforcement, providing evidence for the efficacy of lag schedules of reinforcement in producing variability in greeting responses. Variability in responding maintained during a reversal to the baseline and generalization conditions, during which a continuous reinforcement schedule was used. Other sources of social reinforcement have likely maintained variability in responding. We hypothesize that responding will be similar with future participants.

According to the current version of the Diagnostic and Statistical Manual of Mental Disorders (5th ed., text rev.; DSM-V; American Psychiatric Association [APA, 2013], one of the core features of autism is persistent deficits in social communication. These social communication deficits often become apparent when children diagnosed with autism fail to demonstrate a functional form of communication, specifically manding for preferred items and activities. The development of a manding repertoire increases the likelihood of contacting reinforcement from a listener. Often single word mands are developed first. The development of multiple word mands or mand frames (e.g., "I want," "May I have") may further increase the likelihood of contacting reinforcement by clarifying the function of the speaker's vocalizations and, therefore, effective interventions to produce functional mand frames is needed. The current investigation targeted the appropriate use of mands frames while increasing the variability in the number of mand frames used for 3 participants with autism. Data show that using textual prompts with text fading effectively extinguished one participant's use of an incorrect mand frame while simultaneously increasing the variability of correct mand frame usage. The addition of a lag schedule increased variability of mand frame usage for two of the participants.