The Development of False‐positive Outcomes During Functional Analyses of Problem Behavior

Abstract

False-positive outcomes of functional analyses exist when exposure to the consequences in a condition strengthens a relation that did not previously exist. Two functional analyses (aggression and spitting) were conducted with a child with an intellectual disability. High rates of problem behavior occurred immediately in the attention condition of the functional analysis across response topographies, suggesting a clear functional response class. However, rates of aggression and spitting increased in the demand condition, following the initial block of sessions, indicating the possible creation of a novel contingent relation for spitting. METHOD Participant and settingJordan was an 11-year-old boy diagnosed with an intellectual disability who was admitted to an inpatient hospital for the assessment and treatment of severe problem behavior. The target behaviors during this study were aggression and spitting. All sessions were conducted in a 3 × 3 m padded room with a one-way mirror for observational purposes. Response definitions and interobserver agreementProblem behavior included aggression (hitting with an open hand or closed fist, kicking, and pinching others), and spitting (expelling liquid past the plane of his lips). All problem behavior was recorded using frequency measures and then converted to 51 The development of false-positive outcomes

Reducing Ambiguity in the Functional Assessment of Problem Behavior

“…However, no research has successfully demonstrated a method for determining what behaviors are in the same or different classes. Assessing multiple behaviors is sometimes more efficient (Derby et al ., ) but also may obscure outcomes (Asmus et al ., 2004) or, worse, teach new functions for problem behavior (Jessel et al ., ). Therefore, additional research is needed in identifying behaviors that are in the same class prior to conducting an FA, and this might be accomplished through IA or DA.…”

“…found that this abbreviated assessment identified behavior maintained by automatic reinforcement in 21 of 22 cases; therefore, an alone screen could be conducted to enhance efficiency when problem behavior is hypothesized to be maintained by automatic reinforcement. Conducting an alone screen may reduce the amount of time required to conduct the FA as well as limit the amount of exposure to FA conditions and hence the potential for a false positive (Jessel et al ., ; Rooker et al . ).…”

“…An intervention for escape from demands may not have been necessary, but one could argue that training a teacher not to remove demands following problem behavior (i.e., healthy classroom contingencies) would not likely be harmful. In addition, an FA may yield false‐positive results (Jessel, Hausman, Schmidt, Darnell, & Kahng, ; Rooker, Iwata, Harper, Fahmie, & Camp, ; Shirley, Iwata, & Kahng, ), and having information from multiple assessments could provide further support for conclusions.…”

“…Several procedural variables also merit consideration. When selecting target behavior, conducting the FA on one form of behavior (e.g., only SIB or only aggression) at a time will limit false‐positive results (finding a functional relation when one does not exist; Jessel et al ., ) and false‐negative results (not finding a functional relation when one exists; Asmus, Franzese, Conroy, & Dozier, ; Beavers & Iwata, ). In addition, FA sessions are typically rapidly alternated, using a multi‐element design, to increase the efficiency of the procedure.…”

A translational evaluation of potential iatrogenic effects of single and combined contingencies during functional analysis

“…Other researchers have suggested iatrogenic effects may be most likely during FA if a tangible condition is included (Rooker, Iwata, Harper, Fahmie, & Camp, ). However, Jessel, Hausman, Schidt, Darnell, and Kahng () reported a case of a potential iatrogenic effect during the escape condition of a traditional FA.…”