Treatment Integrity in Single-Subject Research
Frank M. Gresham University of California--Riverside
The empirical demonstration that changes in behavior are functionally related to manipulated changes in the environment is a fundamental principle of behavior analytic research ( Baer, Wolf, & Risley, 1968). In other words, researchers must demonstrate that changes in a dependent variable are functionally related to systematic, controlled, and measurable changes in an independent variable. This demonstration requires the representation of the independent variable by known physical parameters of an environmental event. Without clear and unambiguous specification of the independent variable, there can be no definitive conclusions regarding a functional analysis of behavior ( Johnston & Pennypacker , 1980; Sidman, 1960).
Recently, researchers and practitioners have demonstrated a heightened interest in what is known as treatment integrity ( Gresham, Gansle, & Noell, 1993; Gresham, Gansle, Noell, Cohen, & Rosenblum, 1993; Moncher & Prinz, 1991; Peterson, Homer, & Wonderlich, 1982). Treatment integrity (sometimes referred to as treatment fidelity) refers to the degree to which treatments are implemented as intended ( Peterson et al., 1982; Yeaton & Sechrest, 1981). A subset of treatment integrity is treatment differentiation, which refers to the degree to which treatments or treatment conditions differ from one another ( Kazdin, 1986).
Treatment integrity is concerned with the accuracy and consistency with which independent variables are implemented. Treatment integrity is necessary, but not sufficient, for the demonstration of functional relationships between experimenter-manipulated independent variables and dependent variables. That is, some independent variables may be implemented with perfect integrity yet show no functional relationship