A critique of the effectiveness of applied behavior analysis research

Abstract

Since its inception, applied behavior analysis has required that solutions to socially significant problems be effective, though criteria for this dimension have remained largely implicit. This paper reviews three explicit technique for determining the effectiveness of behavioral research: graphical, social validational, and cost analyses. The concept of effect size is introduced as an additional means of comparing the effectiveness of various treatment alternatives. Survey data are utilized to support a bothersome implication of this review, namely that the contingency to produce large effects placed on behavioral researchers may actually decrease the likelihood that a useful technology of application will be produced. Finally, strategies are offered for preserving the effectiveness of behavioral procedures when existing technologies are disseminated to settings of relevance.