Social Interaction Setting Events: Experimental Analysis of Contextual Variables
Sasso, Gary M., Peck, Janine, Garrison-Harrell, Linda, Behavioral Disorders
ABSTRACT Severe behavioral problems and social interaction deficits are the two areas of functioning most responsible for the failure of students with emotional and behavioral disorders (E/BD) to adjust adequately and achieve success in inclusive settings. Over the past 10 years, the development and refinement of a technology of functional analysis has allowed educators and researchers to address problem behavior more effectively by matching the function of the response to an intervention. Recently, investigators have begun to turn their attention to the assessment and manipulation of setting events that can elicit appropriate responses. This article describes an experimental hypothesis-testing model of structural analysis designed to identify critical contextual variables associated with successful social interaction. Suggestions for use in general education settings are provided, with an emphasis on matching inclusive social environments to individual student antecedent variables. Strengths and weaknesses of various forms of functional analysis are discussed in conjunction with the need to develop reliable, valid, and acceptable assessments that can be used in natural environments.
Severe problem behaviors pose serious challenges for families and teachers of children with E/BD. Behaviors such as noncompliance, aggression, property destruction, and antisocial responses place these children at risk for exclusion from typical school settings (Dunlap & Kern, 1993; Sasso et al., 1992; Vaughn & Horner, 1997).
In the last 20 years, we have witnessed a reemergence in the behavior analytic literature of assessment and treatment processes based on the functions of severe behavioral problems (Carr, 1977; Iwata, Dorsey, Slifer, Bauman, & Richman, 1982/1994). Functional assessment attempts to identify the maintaining variables that control or support the occurrence of behavioral problems. When these variables are identified accurately, hypothesis-- driven interventions can be devised (Durand & Carr, 1987; Repp, Felce, & Barton, 1988). There is also increasing evidence that interventions matched to the functions of behavioral problems yield more effective and durable results (Carr & Durand, 1985; Iwata, Pace, Cowdery, & Miltenberger, 1994; Mace, Lalli, & Pinter-Lalli, 1991).
The operant model provides two areas of emphasis for assessment of variables maintaining both problem and prosocial behavior. Functional analyses are designed to determine maintaining variables (consequences) that directly control responses (Neef & Iwata, 1994; Sasso & Reimers, 1988). This assessment process can result in a powerful experimental analysis capable of producing direct functional relationships between a behavior and its controlling variables. Structural analyses, on the other hand, provide an assessment of antecedent conditions or discriminative stimuli that set the stage for responses. Until recently, few investigations were reported demonstrating the relationship between an assessment of these contextual variables and antecedent interventions for severe behavior problems (Carr & Carlson, 1993; Touchette, MacDonald, & Langer, 1985). However, recent work employing a hypothesis-driven structural assessment model suggests that variables such as task length (Dunlap, Kern-- Dunlap, Clarke, & Robbins, 1991), task difficulty (Umbreit, 1996a), and preference (Clarke et al., 1995; Umbreit & Blair, 1997) can have significant effects on the occurrence of behavior. In all of the investigations just mentioned, the primary focus of the assessment was severe behavioral problems and the goal was a reduction of these problem behaviors through a careful arrangement of identified antecedent variables.
Another area of behavioral support that may benefit from these types of analyses is social interaction. The recent emphasis on inclusion has opened a number of new environments and interaction opportunities to children with E/BD (Fuchs & Fuchs, 1994; Lewis, Chard, & Scott, 1994). …