JOURNAL OF APPLIED BEHAVIOR ANALYSIS 1994, 27, 385-392 NUMBER 2 (SUMMER 1994)
THE SIGNIFICANCE AND FUTURE OF FUNCTIONAL ANALYSIS METHODOLOGIES
F. CHARI.ES MAcE THE UNNERSI1Y OF PENNSYLVANIA
Iwata, Dorsey, Slifer, Bauman, and Richman (1982) presented the first comprehensive and stan dardized methodology for identifying operant funaions of aberrant behavior. This essay discusses the significance functional analysis has had for applied behavior analysis. The methodology has lessened the field’s reliance on default technologies and promoted analysis of environment-behavior interactions maintaining target responses as the basis for selecting treatments. It has also contributed to the integration of basic and applied research. Future directions for this research are suggested.
DESCRIPTORS: functional analysis, behavior modification, behavior analysis
The roots of functional analysis methodologies can be traced to the earliest years of applied be havior analysis (e.g., Ayllon & Michael, 1959; Bi jou, Peterson, & Ault, 1968). However, the article by Iwata, Dorsey, Slifer, Bauman, and Richman ( 1982) (reprinted in this issue of JABA) built upon previous theoretical papers (e.g., Carr, 1977) and research methods (e.g., Bijou et al., 1968; Thomas, Becker, & Armstrong, 1968) to formulate the first comprehensive and standardized functional analysis methodology. This methodology, initially applied to the analysis of self-injurious behavior, was soon adapted to analyze environment-behavior inter actions that maintained a wide variety of behavior disorders, such as aggression (Mace, Page, Ivancic, & O’Brien, 1986; Wacker et al., 1990), destructive behaviors (Slifer, Ivancic, Parrish, Page, & Burgio, 1986), disordered speech (Mace & Lalli, 1991; Mace, Webb, Sharkey, Mattson, & Rosen, 1988; Mace & West, 1986), stereotypy (Durand & Carr, 1987; Mace, Browder, & Lin, 1987; Wacker et al., 1990), pica (Mace & Knight, 1986), and tan trums (Carr & Newsom, 1985).
Since being introduced to functional analysis as an intern at the Kennedy Institute in 1982, I have always considered the methodology to be one of the most significant developments in applied be havior analysis. My objectives in this essay are to offer some perspectives on the importance of func-
Requests for reprints may be addressed to F. Charles Mace, Department of Pediatrics, University of Pennsylvania, 3405 Civic Center Blvd., Philadelphia, Pennsylvania 19104.
tional analysis to applied behavioral science and technology development and to suggest some di rections for the future evolution and refinement of functional analysis methodologies.
The Significance of Functional Analysis
Beyond behavior modification: A return to behavior analysis. An assumption common to most applications of learning theory aimed at modifi cation of socially relevant human behavior is that both adaptive and aberrant behaviors are learned through a history of interactions between an indi vidual and the environment (Baer, Wolf, & Risley, 1968; Bijou & Baer, 1961; Krasner & Ullmann, 1965; Skinner, 1953; Tharp & Wetzel, 1969). The vast majority of these interactions are believed to follow operant paradigms of positive and neg ative reinforcement. This history of reinforcement, in tum, influences how an individual responds to current environmental contingencies.
Before applied behavior analysts had a meth odology to identify the conditions maintaining ab errant behavior, the reinforcement histories that gave rise to current behavior-environment inter actions were largely ignored. Instead, existing rep ertoires were altered and new ones established by superimposing reinforcement contingencies, pun ishment contingencies, or both, onto the current environmental contingencies or unknown processes that maintained aberrant behavior. The approach was known generically as behavior modification. However, without the capacity to explicitly inter rupt the events maintaining aberrant behavior, be-
385
http:CHARI.ES
386 F. CHARLES MACE
havioral interventions relied on potent reinforcers and/or punishers that could override the conditions that supported problem behavior. Although effec tive in many cases, this strategy of overriding the maintaining conditions led to concerns about the field· s overreliance on the default technologies of contingent aversive stimulation and artificial posi tive reinforcement (Iwata, 1988; Johnston, 199 la, 1991b; Sherman, 1991).
Behavior modification has also come under fire for its approach to technology development and the treatment philosophies it has spawned (e.g., Dietz, 1978; Hayes, Rincover, & Solnick, 1980; Johnston, 1991a; Mace, 1994; Pierce & Epling, 1980). Technology is directly affected by the scope of research questions posed. Behavior modification research has generally limited itself to the following questions: “What procedures produce behavior change?” “What is the generality of these effects across subjects, behaviors, and settings?” “What are the long-term benefits of this procedure?’· ”What is the relative efficacy of various procedures in treat ing the same problem behavior?’· Although these questions certainly have merit, their focus is limited to technical appl
