Association for Behavior Analysis International

Direct observation and visual analysis are cornerstones of treating challenging behavior and this symposium evaluates novel methods in this domain. Two studies targeted assessment, with the first comparing assessments with different goals and measurement systems and the second evaluating visual analysis decision making. Specifically, one study evaluated correspondence between a concurrent operant analysis (measuring choice allocation) and a multielement functional analysis (measuring challenging behavior) and conducted treatment comparisons for cases of non-correspondence. The next study evaluated the number of sessions required in extended alone or ignore assessments to make valid conclusions about whether challenging behavior is automatically maintained based on visual analysis of the rate of behavior across sessions. The other studies evaluate direct observation and visual analysis in treatment contexts, one focused on observations across contexts and the other comparing treatments with measurement of various dimensions of behavior. For the observations across contexts, the authors demonstrate whether a treatment for stereotypy shown effective in controlled contexts is also efficacious in naturalistic settings. The next study also targeted repetitive movements, but compared the success of two treatments (blocking and manual guidance) using three behavioral measurements, demonstrating that some measurement systems may be more sensitive in detecting treatment differences.

Professionals working with children who exhibit challenging behavior.

We evaluated the extent to which the results of a response-guided concurrent operant analysis (COA) framework assessing relative preference for attention, tangible, and escape corresponded with results of functional analyses (FA) of problem behavior across six children with or at risk for disabilities. For each participant, we conducted a COA using a simultaneous treatments design, followed by a multi-element FA. In cases where the two assessments did not fully correspond, we used an alternating treatments design to compare the effects of interventions based on each assessment outcome on rates of problem and appropriate behavior. Finally, for cases in which at least one effective intervention was identified, we used a concurrent chain procedure to evaluate participant preference among effective interventions. COA results partially corresponded with FA results (i.e., COA identified one of multiple functions of problem behavior) for all six participants. Treatment results varied by participant. For one participant, both treatments were equally effective; for two participants, neither treatment was effective; and for three participants either the FA- or COA-based treatment was more effective. Participants for whom a superior treatment was identified showed a consistent preference for the more effective intervention.

Several studies have evaluated functional analysis (FA) methods for identifying whether challenging behavior is maintained by automatic reinforcement. Queirim et al. (2013) found that extended alone sessions conducted prior to a multielement FA correctly identified whether challenging behavior was maintained by automatic reinforcement. However, it is unclear how many sessions are needed in this type of an assessment to make confident conclusions. In this study, we conducted a retrospective consecutive case series analysis from clients with intellectual or developmental disabilities that completed extended alone or ignore sessions as part of a day treatment admission for challenging behavior. Using visual analysis, we compared decisions regarding whether challenging behavior was automatically maintained based on data from the first three and six sessions of the assessment to decisions made when all available clinical data were presented. Results from the first 20 cases suggest that reliability between decisions made following three data points compared to all available clinical data was poor, but increased dramatically with the first six data points. This suggests that at least six consecutive sessions with alone/ignore contingencies may be required to draw conclusions about automatically-maintained challenging behavior. Results are discussed in the context of maximizing assessment efficiency.

The participant was a 19-year-old with autism who engaged in high rates of repetitive body movements (RBM) that interfered with and was responsible for the loss of paid employment. We chose two vocational tasks (i.e., cleaning a table and rolling napkins) that the participant performed independently at the job site and that typically took approximately 5- 8 minutes to complete. We used a multi-element design across these tasks to compare the effects of blocking and manual guidance on the number of RBM in, the duration of task completion, and the rate of RBM. Duration of task completion for both tasks decreased, but more so for cleaning the table when manual guidance was used. The frequency of RBM reduced dramatically for both conditions and remained low. Taken together, these two measures imply that both blocking and manual guidance were equally effective. Alternatively the rate of RBM, indicated that blocking was the more effective procedure. Taken collectively, we decided to continue with manual guidance as that was the more effective procedure for decreasing the duration of task completion. Results support the importance of using more than one measurement to determine intervention effects.