fective prevention programming, other content also appears important. Reviews have indicated that approaches aimed at changing normative beliefs about drug use are also effective (Hansen, 1992; Gottfredson, 1997). These approaches often use survey results to correct misperceptions about the prevalence of use, engage youths in discussions to elicit their opinions about the appropriateness, and include testimonials from admired peers emphasizing that use is not acceptable. Instructional programs that incorporate these norm-setting activities have been shown to reduce use (Gottfredson et al., forthcoming; Hansen, 1992), but noninstructional programs that employ these methods outside the context of a broader substance use prevention curriculum are also effective for reducing substance use (Hansen and Graham, 1991; Perry et al., 1996).
Note that these approaches assume that incorrect information about the prevalence or appropriateness of use, rather than poor skills for dealing with social influences to use, increases substance use. On the surface, the effectiveness of these approaches does not square with conclusions of some reviews that programs that provide “information only” about the consequences of substance use do not work. They suggest rather the importance of a credible opinion leader conveying correct information. It may be that the programs giving evidence of harmful effects of providing information only failed to include these key components. Recent commentaries have recommended a return to the information approach, on the assumption that when teens are provided with accurate information, especially from trusted people, they will make good decisions (e.g., Beck, 1998).
The committee recommends additional research to assess the effectiveness of social competency skill development and normative education approaches, which emphasize conveying correct information about the prevalence of drug use and its harmful effects. This research should also assess the interaction between the content of the prevention activity and the risk level of the population targeted, because it is likely that provision of correct information may be especially effective for the subset of the population that is most at risk for higher levels of use. It is likely that research on programs conducted under more tightly controlled or experimental conditions may overestimate the effects that would be observed under normal conditions, and this factor may confound comparisons of the effects of different programs. Therefore the committee recommends additional research on prevention practices implemented under conditions of normal practice so that variability in effects from study to study may be better understood. Finally, the committee recommends further research on alternative methods and targeting mechanisms for teaching social competency skills.