limited number of studies (Austin and Johnson, 1997a, 1997b) have shown some positive effects of media literacy programs aimed at affecting perceptions of alcohol advertising and alcohol norms, but there is insufficient evidence to make conclusions about the application of this approach in the context of underage alcohol use.

Approaches that have been demonstrated to reduce youth alcohol use have many program elements in common. However, similar to other approaches recommended in this report, the committee believes that education-oriented interventions should be implemented in the context of a comprehensive approach.

A considerable amount of research (Gottfredson and Wilson, 2003; Hansen and Dusenbury, 2004; Tobler and Strattan, 1997), primarily in primary and secondary schools, has identified several critical elements of successful school-based educational interventions. In addition, research on communitywide alcohol prevention programming (see Chapter 11), such as Project Northland and family-based approaches like the Michigan State University Multiple Risk Outreach Program, offer additional critical elements that can make education interventions more effective (Williams et al., 1999; Nye et al., 1995). Research on such interventions offers a number of lessons about what educational strategies are important for preventing alcohol use and alcohol problems among minors. These lessons, or critical elements, offer a starting place for innovative education interventions and for developing priorities about what kinds of education interventions should be funded. The interventions need to be multicomponent and integrated; sufficient in “dose” and follow-up; establish norms that support nonuse; stress parental monitoring and supervision; be interactive; be implemented with fidelity; include limitations in access; be institutionalized; avoid an exclusive focus on information and avoid congregating high-risk youth; and promote social and emotional skill development among elementary school students.

Multicomponent and Integrated Schools provide a captive population for the delivery of prevention programs and effects can last for 3 to 4 years. Similarly, family and community-based interventions have also produced reductions in the prevalence and intensity of alcohol use. However, prevention effects are maximized when all of these venues are used in concert in a coordinated and mutually supporting manner. For example, meta-analyses (Gottfredson and Wilson, 2003; Tobler et al., 2000) revealed that systemwide change interventions were most effective. Project Northland (see Chapter 11), which included school-based education programs, com-