Association for Behavior Analysis International

While the introduction of functional analysis (Iwata et al 1982/1994) forever changed the field of Applied Behavior Analysis, modifications to behavior analytic procedures are a necessity to have true utility in developing effective treatments for all individuals in all settings with a wide variety of problem behaviors. Individuals with disabilities may exhibit problem behaviors that are difficult to analyze using typical analogue assessments due to intensity or rate (Davis et al, 2012), or the problem behavior may be maintained by idiosyncratic variables that are observed in the natural setting but not during a standard functional analysis (Hanley, Iwata & McCord, 2003). Finally, undifferentiated results may require additional analysis to determine a function. Hagopian et al. (2013) reported that the percentage of functional analyses with differentiated results increased from 47% to 87% when modifications to standard analogue conditions were made. This symposium seeks to extend recent research on modifications to functional analysis methodology to result in more accurate identification of behavioral function. First, the utility of a mand assessment to clarify inconclusive functional analysis results is examined. Second, an assessment of idiosyncratic variables evoking problem behavior during transitions is described. The third presentation focuses on elopement, a behavior that is difficult to address via functional analysis. This study is a systematic replication of the Lehardy et al. (2013) single-room functional analysis of elopement. The final presentation presents data comparing functional analyses of single and multiple response topographies to identify behavioral function of multiple problematic behaviors.

Mands often precede problem behavior and the responses may be related in several ways. Specifically, (a) mands may be one response in a response chain with problem behavior, (b) mands may be the first behavior in a response class hierarchy with problem behavior (Lalli, 1995), or (c) problem behavior may be a precurrent response increasing the probability that future mands will receive reinforcement (Fisher et al. 2001). In this study, functional analyses of problem behavior either failed to evoke problem behaviors or yielded undifferentiated results for four participants with autism. An assessment was then developed to empirically identify the relationship between repetitive mands and problem behaviors. Following a functional analysis to identify the specific reinforcer maintaining mand responses, the consequences for mands (e.g. specific reinforcement, extinction, verbal No, and nonspecific reinforcement) were experimentally manipulated (as potential antecedents to problem behavior), while problem behavior produced access to reinforcement for the mand. Across all participants, problem behavior reliably occurred in one or more of the test conditions relative to conditions during which mands produced specific reinforcement. Data suggest that the problem behaviors for these individuals served as a precurrent contingency for mands.

Although problem behavior during transitions is a commonly described occurrence and various strategies to reduce these problem behaviors, a limited number of studies have analyzed the function of problem behaviors during transitions. Previous research has suggested that problem behavior during transitions may be reinforced by escape from an environment with demands or low attention (Kern and Vorndran, 2000), the unpredictability of the transitions (Flannery & Horner 1994), initiating or terminating an activity or changing locations (McCord et al. 2001). In the present study, a functional analysis was conducted for two participants, during which demands to transition to a room containing academic materials were provided. Problem behavior produced escape from the transition. Both participants exhibited problem behavior during the escape from transition condition relative to a control condition, resulting in no contact with the work in the room. After examining variables specific to the transition, data suggested that it was not the presence of academic materials evoking the problem behavior, but specific features of the environment that may have served as conditioned aversive stimuli (e.g., presence of people and, small spaces). Results will be discussed in terms of conditioned motivating operations, as well as implications for treatment.

The functional reinforcers maintaining problematic behavior may vary across different response topographies. Several studies have recommended that clinicians graph functional analysis data for each topography separately when multiple responses are reinforced during the assessment (e.g., Derby et al. 2000) or conduct separate functional analyses for each topography (e.g. Mace et al. 1986). However, no studies have directly compared assessments reinforcing a single topography with assessments reinforcing multiple topographies of behavior. It is possible that conducting separate functional analyses may unintentionally obscure the results of functional analyses (e.g. by placing one response topography on extinction during the assessment). In the current study, we conducted functional analyses with 3 participants, each displaying at least 2 topographies of problem behaviors. In two separate assessments, either all topographies of problem behavior or a single topography of problem behavior produced reinforcement. Assessment order was counterbalanced across participants. Results indicated that the functions of separate topographies were identified in both the single and multiple topography functional analyses. In fact, data for all three participants showed that reinforcing only one topography of behavior in an assessment also provided a clear identification of function for a second behavior, not producing reinforcement.