Assessment and Treatment of Elopement Utilizing a Trial-by-Trial Format

Background: Elopement, or leaving caregiver supervision, is a behavior with potentially dangerous consequences for individuals with autism (Matson & Rivet, 2008). As with other forms of problem behavior, understanding the function of elopement is important for successful treatment. Individuals have been shown to elope for various reasons, including to access attention, to access preferred items/activities, or to escape from nonpreferred demands. Piazza et al. (1997) presented a method of identifying the function of an individual’s elopement by systematically presenting specific antecedent events and delivering an associated consequence when elopement occurred. A retrieval procedure was used to reinstate the antecedent condition following elopement. Elevated rates of elopement in one or more test conditions relative to a control condition suggested that the consequence maintained within that condition was responsible for maintaining elopement. However, because some preferred items or activities are conditionally available (e.g, elopement towards an ice-cream truck can only occur when the ice-cream truck approaches), it can be difficult to assess certain forms of elopement utilizing this type of assessment.

Objectives: The purpose of this study was to develop a novel procedure for identifying the function of elopement when it is necessary to do so in a more naturalistic setting.

Methods: The participant had a history of elopement toward elevators and other items with buttons. Thus, all sessions were conducted a lobby and adjoining elevators. Attention, demand, and toy play conditions were conducted in the elevator. In these conditions, RM was able to press buttons and ride the elevator continuously. In all sessions, one instance of the elevator doors opening constituted the beginning of a trial. A trial ended when RM either eloped or the elevator doors closed. Each session consisted of five trials. RM received continuous attention during the toy play condition. Throughout each session of the demand condition RM was required to complete non-preferred demands. A 20 s break from demands was delivered contingent upon elopement. During the attention condition, no attention was delivered unless elopement occurred; in which case the therapist delivered 20 s of attention. During the tangible condition, a trial started as soon as the therapist and RM began walking through the lobby area. The trial ended following either the occurrence of elopement toward the elevator, in which case RM was provided access to the buttons and one elevator ride, or when RM and the therapist finished walking through the lobby area without the occurrence of elopement.

Results: Low percentages of elopement were observed in the attention (M=10%), demand (M=15%), and toy play (M=0%) conditions. In contrast, a high percentage of trials containing elopement occurred in the tangible condition (M=100%). These results suggested that elopement was maintained by access to the elevator.

Conclusions: This study demonstrates a novel methodology for assessing the function of elopement when typical methods are unfeasible. Future research should focus on adapting this trial-by-trial format to assessing elopement hypothesized to be maintained by other idiosyncratic sources of reinforcement.