Association for Behavior Analysis International

This symposium presents a series of studies each featuring a novel methodological approach to the prevention, assessment, and treatment of challenging behavior. Tara Fahmie will present the effects of differential reinforcement and delay training on behavior maintained by access to tangibles in the prevention of severe problem behavior. Prevention in this study is demonstrated by a post-intervention decrease in moderate behaviors in the intervention setting and by a continued increase in the severity of behaviors in untreated settings. Flavia Julio will present a high-resolution post-effect analysis to evaluate changes in problem behavior maintained by automatic reinforcement following the discontinuation of noncontingent matched stimuli. This strategy may help to identify short-term post-effects consistent with either establishing or abolishing effects of matched stimuli over the putative sensory reinforcer. Jade Wightman will present an evaluation of the usability and reliability of electromyography-aided approach to the identification of bruxism episodes in individuals with profound intellectual disability. The study features a minimally invasive system for the direct recording on muscle activity that could provide more accurate measures during the functional analysis and treatment of bruxism. Finally, Javier Virues-Ortega will present an intervention protocol for tracking and reducing excessive head movement in children with autism undergoing neuroimaging examinations. Our discussant, Dr. Brian A. Iwata, will provide his informed perspective on these contributions to the prevention, assessment, and treatment of challenging behavior.

Noncontingent matched stimulation is intended to reduce behavior maintained by automatic reinforcement by emulating the sensory products of problem behavior, thereby inducing satiation (reinforcer replacement hypothesis). By contrast, matched stimuli may simply interfere with the sensory products of the behavior, thereby establishing the value of the sensory reinforcer over the duration of the matched stimuli (interference hypothesis). Post-effect analyses could help to reveal which behavioral process may be operating during noncontingent matched stimulation. Specifically, increments of problem behavior following the discontinuation of matched stimuli would be consistent with the interference hypothesis. By contrast, gradual recovery of problem behavior following the termination of matched stimuli would be consistent with the sensory reinforcer replacement hypothesis. Published post-effect analyses (e.g., Rapp, 2007) do not provide definitive evidence in favor or against either hypothesis, chiefly because changes in motivating operations induced by noncontingent matched stimuli may be short lived and could be masked by the overall variability of the behavior. In order to evaluate subtle post-effects induced by matched stimuli we conducted a high-resolution post-effect analysis in a series of cases of problem behavior maintained by automatic reinforcement. Our results indicated that both the interference and the reinforcer replacement hypotheses were consistent with the post-effects of matched stimuli in different subjects. Clinical recommendations based on our findings will be discussed.

There are few studies on bruxism in the behavioral literature. Functional analyses of bruxism in individuals with developmental disabilities have relied on audible responding to assess the occurrence of bruxism. However, bruxism can also be inaudible. An electromyography (EMG) measures the frequency, duration, and amplitude of bruxism and allows for the assessment of both audible and inaudible bruxism events. We used a minimally invasive EMG wireless device to record bruxism episodes. Bruxism is typically defined as episodes of masseter muscle activity over 30% of muscle voluntary contraction (MVC). We present a proxy of the 30% MVC index for individuals with profound intellectual disability that are not able to follow verbal instructions skill necessary to estimate MVC. Over a series of free operant sessions we recorded a minimum of 100 EMG activity events and obtained the 30th percentile of the distribution of events as a proxy of the 30% MVC value. Subsequently, we used this threshold to train observers to identify bruxism events based on real-time EMG recordings. Our study indicates that it is possible (and time-efficient) to estimate a proxy to the 30% MVC value for individuals with profound intellectual disability. Our results also show that it is possible to train observers to use the proxy threshold to reliably identify bruxism events based on EMG recordings. A functional analysis and treatment study using this methodology is currently under way.

Magnetic resonance imaging (MRI) is a non-invasive medical procedure that is an essential diagnostic tool for a myriad of neurological conditions. During the scanning session the individual is required to remain still throughout the procedure, which can last from a few seconds to several minutes. This task is particularly difficult for children with autism as they often present with diminished capacity to comprehend instructions and may also present with sensory sensitivities. Traditionally, individuals with disabilities have been sedated in order to undergo an MRI scan, which has been correlated with numerous medical side effects including death. Moreover, sedation has a success rate below 70%. There is some evidence suggesting that it is possible to promote steadiness in children using exposure-based methods. However, none of these studies have included participants with autism. Also, the studies available frequently fail to include a generalization test in a real scanner. We used an exposure-based intervention to promote tolerance to a mock scanner, followed by a 5-min DRO adjusting schedule to establish steadiness. Training was followed by a real scanner session. The results of this study are promising in that we were able to establish steadiness in all participants. Moreover, the real scanner quality indices resulting from the real scanner sessions were all within the acceptable range.