Association for Behavior Analysis International

Functional analyses have become standard of care for the treatment of problem behaviors exhibited by individuals with developmental disabilities. Functional analyses have allowed behavior analysts to isolate the reinforcers that maintain problem behaviors; thus, resulting in more effective behavioral interventions. Since the early 1980s, there has been tremendous growth in the amount of research on the evaluation, refinement, and extension of functional analyses methodology. The purpose of this symposium is to present four studies that have examined extensions of functional analysis methodology. The first two set of papers will examine methods of expediting the evaluation of behavioral function. One study examined a method of quickly determining if problem behaviors are maintained by automatic reinforcement. The other examined if latency measures could be used to elucidate escape-maintained problem behaviors. The third paper examined a method for refining functional analyses by evaluating test-specific control conditions during functional analyses. The final paper examined methods of identifying idiosyncratic variables that may maintain problem behavior.

A typical functional analysis (FA) includes a series of 10-min conditions that identify different sources of reinforcement for problem behavior (Iwata et al. 1994/1982). Results of several studies suggest that shortening session durations, reducing the number of sessions, or limiting the number of comparisons might be considered as assessment options when time is limited (Derby et al., 1992; Kahng & Iwata, 1999; Northup et al., 1991; Wallace & Iwata, 1999). A general finding in previous research is that behavior maintained by automatic reinforcement persists in the alone condition, whereas behavior maintained by social contingencies extinguishes. Thus, initial exposure to only the alone condition may represent an efficient screening procedure when maintenance by automatic reinforcement is suspected. We conducted a series of 5-min alone sessions with individuals who exhibited varied topographies of problem behavior and subsequently conducted complete FAs to verify initial predictions based on results of the screening. Results to date indicate that the screening procedure is a useful and extremely efficient assessment tool when time is limited.

Although research has been conducted that evaluated an efficient method of identifying a task preference hierarchy (i.e., used latency to the first problem behavior as dependent measure; Call, Pabico, & Lomas, 2009), a direct relationship between response latency and rate as measures of problem behavior has not been evaluated. In the current study, a demand assessment and subsequent demand analysis were conducted with 6 participants who were admitted to an inpatient unit for the assessment and treatment of severe problem behavior. The demand assessment was conducted to identify a task preference hierarchy and to evaluate the relationship between response latency and rate. Less aversive (i.e., tasks associated with the longest latency, [LA]), and high aversive (i.e., tasks associated with the shortest latency, [HA]) tasks were identified and included into a demand analysis. For 5 of the6 participants, an inverse relationship between response latency and rate during the demand assessment was shown. Additionally, the demand analysis yielded differentiated levels of problem behavior per LA and HA task condition for3 of the6 participants. These results suggest that response latency may be predictive of response rate, providing an efficient measure of problem behavior when identifying a task preference hierarchy.

Functional analyses of problem behavior typically include a play condition to control for the effects of both positive and negative reinforcement. However, results of2 studies highlight the potential advantage of using different, test-specific conditions to control for each source of reinforcement. Fischer, Iwata, & Worsdell (1997) compared responding in the attention, play, and alone conditions for 36 individuals with attention-maintained SIB; responding was lowest in the play condition for 31 individuals. Kahng and Iwata (1998) compared responding in the demand, play, and alone conditions for 55 individuals with escape-maintained SIB; responding was lowest in the alone condition in every case. We compared responding across4 potential control conditions (alone, ignore, play, and DRO) for individuals with problem behavior maintained by either positive or negative reinforcement. Results to date suggest that the play and DRO conditions best suppress problem behavior maintained by attention, whereas the alone and ignore conditions best suppress problem behavior maintained by escape. Additional tests will be conducted to determine whether test-specific controls are better than a common control in a pairwise functional analysis when the function of problem behavior is unknown.

By conducting a functional analysis, clinicians can identify behavioral function and develop effective treatment. Although functional analysis (FA) often results in clear outcomes, it sometimes yields ambiguous response patterns that cannot be interpreted. When this occurs, clinicians may need to modify functional analysis conditions to include idiosyncratic antecedent or consequent events. To date, a systematic strategy for identifying variables to include in modified conditions has not been reported. In this study, we evaluated a systematic progression of indirect and descriptive analyses for identifying idiosyncratic variables to test following an initially ambiguous FA outcome for4 individuals diagnosed with an autism spectrum disorder. We conducted2 indirect assessment formats across multiple respondents to identify reliable idiosyncratic antecedent and consequent variables to test. We also conducted descriptive observations to confirm the occurrence of these events in the natural environment. We used the results from these assessments to inform modified FA conditions that were subsequently evaluated. For all participants, a differentiated outcome was observed during the modified functional analyses, illustrating the utility of this method for identifying idiosyncratic events.