Functional analysis patterns of automatic reinforcement: A review and component analysis of treatment effects.

Abstract

Functional analysis (FA) conditions include different antecedent or consequent events that may disrupt responding. Thus, varying patterns of FA differentiation may predict treatment outcomes of problem behavior maintained by automatic reinforcement. These patterns could be used to inform the development of individualized interventions. An approach to classifying these patterns is to categorize FA outcomes as attention condition lowest, demand condition lowest, and play condition lowest, according to the condition in which problem behavior is most disrupted. In Study 1, we applied this criterion to 120 datasets finding that 60% could be classified using this method, whereas 89% of datasets showed a disruption of 50% or higher. In Study 2, we conducted a treatment component analyses for 3 individuals whose FAs each exhibited one of the 3 distinct patterns. The results indicated that specific elements of the FA conditions could reduce problem behavior. The predictive utility of these disruption patterns is discussed.