Association for Behavior Analysis International

Functional Analyses (FAs) have been widely used in the field of Applied Behavior Analysis to determine the underlying function(s) of problem behavior. Advancements in functional analysis research suggest that functions of behaviors may change for a multitude of reasons (e.g., time, settings). The current symposium includes four papers focusing on findings from functional analyses evaluations. The first talk presents a study on the analysis of individuals who have had 2 FAs conducted. The second talk presents a study on a comparison between first admission and second admission FAs. The third talk presents a study on the investigation of the extent to which interaction effects occur with participants from a university based early childhood center. The final talk presents a study on the comparison of documented FA outcomes to retrospective ongoing visual inspection (OVI) results. All four studies will be discussed by Dr. Amanda Zangrillo who has extensive experience in implementation as well as research regarding FAs.

Submitted in the Basic level.

Few studies have reported the outcomes of functional analyses (FAs) completed at different points in time to evaluate the stability of behavioral function (Lerman et al., 1994). In the current study, we analyzed archival FA data from electronic health records (EHRs) in a large nonprofit behavioral health provider in order to evaluate whether documented functions of behavior remained stable over time. A search of the EHR produced 207 distinct FAs documented between 2018 and 2023. A review of these records identified 21 individuals who had at least 2 FAs completed for the same target behavior at least 6 months apart. Five more individuals with at least 2 FAs were identified through a review of additional clinical records. Of the 26 total individuals identified, FA results had exact agreement for 14 individuals, partial agreement for 8 individuals, and no agreement for 4 individuals. For the individuals whose FA results did not match, 3 out of 4 of the data sets included at least one inconclusive FA result. Variables that may account for the different outcomes (e.g., protocol differences) will be discussed.

Functional analyses are considered the gold standard for identifying the function of problem behavior. Once the functional analysis is implemented and a function is identified, a function-based intervention can be developed and implemented. Typically, if a social function is identified, some form of functional communication (FCT; Tiger et al., 2008) is implemented. Previous research has demonstrated FCT to be an effective treatment that both decreases problem behavior while increasing appropriate alternative responses (i.e., Functional Communication Responses; FCRs). Although FCT is an effective intervention for problem behavior, there may be times where FCT becomes ineffective over the course of its implementation (Ringdahl & St. Peter, 2017). One possible explanation for decreases in the effectiveness of FCT includes a change in the function of problem behavior. Specifically, it is possible that if the initially identified function shifts, the intervention will no longer be relevant. Possible implications of this decrease in effectiveness of treatment may include lack of caregiver treatment adherence, resulting in possible readmissions due to an increase of problem behavior. The purpose of this study was to compare the results of a functional analyses conducted during an initial admission into a severe behavior clinic to the results of a functional analysis conducted during a second admission.

Structured criterion lines have become an objective measurement tool to accurately apply visual inspection to functional analysis (FA) results (Roane et al., 2013). Additionally, ongoing visual inspection (OVI) has been applied as an objective measure to interpret ongoing results of a functional analysis, with the use of “stop” criteria (Saini et al., 2018). In this current study, we applied OVI to over 200 archival functional analyses documented in an electronic health record (EHR) at a large nonprofit behavioral health provider from 2018-2023. Results from the OVI analysis were compared with the documented FA outcomes and data on agreement was recorded. Data were also collected on whether the FA was terminated at the appropriate time based on the OVI criteria. Results suggest that applied practitioners often correctly interpreted the results of experimental functional analysis, despite having very little training in the structured criteria or OVI. It was more difficult to apply the OVI “stop” criteria to the completed FAs due to methodological differences (e.g., protocol modifications, inclusion of an automatic screening)..