Best, J. Allan, Flay, Brian R., Towson, Shelagh M. J., Ryan, Katherine B., Perry, Cheryl L., Brown, K. Stephen, Kersell, Mary W., and d'Avernas, Josie R.
Sixth-grade children in 22 schools received either a social-influences smoking- prevention program or routine health education. The social-influences program was designed to teach youth about peer, parent, and media influences affecting smoking onset and to provide them with skills m resisting these influences. Comparison schools were given no program, but were permitted to continue their usual provision of health education. Program impact was evaluated as a function of pretreatment risk of future smoking. Risk was defined with respect to both (a) the prevalence of social models who smoked and (b) previous smoking experience. Two-and-a-half-year results show program impact to vary with both kinds of risk. Smoking-experience risk interacted such that, at first, there was greater impact on children with experience, but on later follow-up the pattern reversed, with the greater treatment effects seen for those initially with limited experience. Social-models risk showed a direct relationship, with greater risk being associated with greater program impact. Implications both for evaluation research and prevention programs are discussed. [ABSTRACT FROM AUTHOR]