Primary prevention: Toward an experimental paradigm sensitive to contextual variables.

This paper argues that the experimental paradigm which has dominated research and intervention in the prevention of psychosocial disorders needs to become more sensitive to contextual variables. Traditionally, this paradigm has (1) provided a piecemeal method to study and, ultimately, understand reality; (2) focused attention on outcomes at the expense of processes in behavior; and (3) assumed that the results obtained under experimental conditions are externally valid and will generalize across persons, settings, and time. Evidence from studies aimed at the primary prevention of delinquency is used to illustrate these issues. It is concluded that, while it is important to know which factors are reliably associated with the development of psychosocial disorders and interfere with prevention efforts, it is equally important to understand how such factors are organized and, in unison, influence behavior. Implications of this conclusion are discussed, with particular emphasis on emerging methodologies to study behavioral organization (structural equation modeling, meta-analysis), and on the possibility that, rather than searching for universal principles of behavior which would be applicable at all times, we may have to be satisfied with a search for preventive models that are consistent with the cultural context in which they are applied.