Self-Reinforcement and Persons with Developmental Disabilities

Article excerpt

A considerable body of literature indicates that self-monitoring coupled with self-reinforcement exerts a potent effect on establishing, maintaining, or increasing the frequency of a target response (Agran, Fodor-Davis, Moore, & Deer, 1989; Albion & Salzberg, 1982; Bellamy, 1975; Bolstad & Johnson, 1972; Burgio, Whitman, & Johnson, 1980; Drabman, Spitalnick, & O'Leary, 1973; Glynn, 1970; Hughes & Peterson, 1989; McNally, Kompik, & Sherman, 1984; Ninnes, Fuerst, Rutherford, & Glenn, 1991; Sowers, Verdi, Bourbeau, & Sheehan, 1985). Nevertheless, there is debate on the mechanism responsible for these observed effects. Whereas social learning theorists and others (Bandura, 1977a, 1977b, 1978, 1981; Karoly, 1982; Mahoney & Bandura, 1972; Thoresen & Wilbur, 1976) suggest that self-monitoring and contingent self-delivery of reinforcers are the agents of observed changes in behavior, Catania (1975) and Goldiamond (1976) suggest that these effects are artifacts of the process of social standard setting with its accompanying opportunities for social reinforcement. Catania (1975) argued that circumstances generating the behavior of self-reinforcement (i.e. the reinforcement of one's own response) make it impossible to distinguish behavioral control by self-reinforcement from control attributable to other aspects of the process, such as identification and setting of goals; establishment of the target behavior as a valuable event (conditioned social reinforcer); and learning to discriminate among aspects of one's own behavior.

The implications of Catania's criticism are far reaching. In "The Myth of Self-Reinforcement" (1975) he argued that the factors contributing to behavioral control by self-reinforcement are difficult, if not logically impossible, to distinguish. Review of the empirical literature seems to support this conclusion. Problems include use of experimenter-identified (rather than subject-identified) target responses (Glynn, 1970; Lovitt & Curtiss, 1969); presence of an observer who may exert antecedent control over the subject's behavior (Ninnes et al. 1989); and failure to establish self-delivered reinforcers as functional reinforcers, along with the confounding of self-reinforcement with external (social) reinforcement (Nelson, Hayes, Spong, Jarrett, & McKnight, 1983). There is debate on the basic question of whether personal evaluative standards can be independent of the behavior being evaluated (Goldiamond, 1976; Thoresen & Wilbur, 1976).

Hayes, Rosenfarb, Wulfert, Munt, Korn, and Zettle (1985) attempted to avoid these methodological difficulties in a study designed to separate the effects of self-delivery of reinforcers from those of public standard setting. The first variable they manipulated was whether treatment took place in a public or private context. The second variable was whether there was goal setting with or without the self-delivery of reinforcement. No performance differences between groups were attributable to contingent self-reinforcement; however, there was an effect of public as opposed to private goal setting. Hayes et al. (1985) concluded that "self-reinforcement procedures work by setting a socially available standard against which performance can be evaluated, which then functions as a discriminative stimulus for stringent or lenient social contingencies" (p. 214). This conclusion must be tempered, owing to methodological and statistical considerations. No proof is offered that the public/private ruse used by the authors actually worked, and no attempt was made to ensure the reliability of the independent variable (i.e., that reinforcers were consumed contiguously and in proportion with target behavior). The conclusion of Hayes et al. (1985) that there was no effect of self-reinforcement may also be called into question because there may not have been enough power in the design (i.e., a small number of subjects) to discern a subtle difference where one might really exist. …