Further Examination of Isolated Versus Combined Contingencies in Functional Analyses

Abstract

Although functional analysis (FA) methodology is the gold standard for determining the function of problem behavior, a major challenge in practice is the safety and efficiency of FAs (Betz & Fisher, 2011; Hanley, 2012; Iwata & Dozier, 2008). To address this, researchers have proposed various procedural and methodological refinements to FAs. A recent methodological refinement to address safety and efficiency involves synthesized (i.e., combined) contingency analyses (SCAs) based on the outcomes of other functional behavioral assessment methods (Hanley et al., 2014). We replicated and extended Holehan et al. (2020) by comparing the outcome of isolated versus synthesized contingencies in functional analyses of precursor behavior and target problem behavior while using a reversal design to replicate the effects, as well as to analyze potential iatrogenic effects (Retzlaff et al., 2020) for four young children. In addition, we collected data on other topographies of problem behavior mentioned in the indirect assessment to see if these behaviors occurred in the same conditions as precursor behavior or target problem behavior or in other conditions to infer a maintaining variable. Furthermore, we examined within-session analyses of FA data to assess under what context precursor behavior or target problem behavior occurred (i.e., establishing operation on, establishing operation off) for isolated and synthesized contingencies. Next, we extended Tsami and Lerman (2019) by evaluating the extent to which FCT+EXT under synthesized contingencies generalized to the different isolated contingencies that were shown to maintain precursor behavior or target problem behavior for participants from Study 1. Additional extensions of Tsami and Lerman included addressing variables not assessed in their study (i.e., combined variables other than escape and tangible), addressing limitations by removing the establishing operation for tangibles during isolated escape test sessions, conducting longer isolated test phases, and examining within-session analyses for synthesized FCT+EXT and isolated test conditions. Results showed that synthesized contingencies were not necessary to show functional relations between precursor behavior or target problem behavior and environmental events for three of four participants. Additionally, intervention results showed that synthesized FCRs did not generalize to all isolated variables.