Mass Media Intervention to Reduce Youth Smoking, a joint program between Baylor College of Medicine and the University of Vermont, sought to develop methods of design and delivery of comprehensive, theory-based media campaigns to reduce the prevalence of cigarette smoking among ethnically diverse adolescents, and to assess the effects of these methods on the prevention and cessation of tobacco use among these populations. The study has three project components: Project 1, “message development using audience research” to conduct formative research to develop theory-based smoking prevention and cessation messages for media campaigns; Project 2, “reducing youth smoking using mass media” to develop media messages in four-year campaigns to prevent smoking among adolescents and evaluate their effectiveness in four intervention mass media markets versus four comparison sites; and Project 3, “youth smoking cessation using mass media” to develop a three-year campaign to help adolescents stop smoking and evaluate their effectiveness with a cohort of weekly smokers established at a baseline in four intervention mass media markets versus four comparison sites. Baylor College of Medicine is responsible for Project 1’s staff training and coordination for formative research activities.
Locations included the University of Vermont in Burlington, Vt., Baylor College of Medicine in Houston, Texas, University of Minnesota’s Data Collection Support Service Center in Minneapolis, Minn., Macro International in Silver Spring and Calverton, Md., San Antonio, Texas, Miami, Fla., and schools and media outlets in: Amarillo and Lubbock, Texas, Charleston and Columbia, S.C., Ft. Myers and West Palm Beach, Fla., and Grand Rapids and Saginaw, Mich.
Students in San Antonio’s Northside Independent School District and Miami-Dade County’s school district, which both have strong Hispanic, African American, and white student populations, participated in PRYSM ad rating sessions – what the investigators refer to as pre-testing of our media campaigns. This helped develop anti-smoking advertisements that were later tested on a much larger scale in California, Texas, Florida, and Washington, D.C. The PRYSM ad development process relies on information provided by students about their hobbies, likes and dislikes, experiences with smoking, and ad opinions. Each study year, ads were developed relying on information collected during previous diagnostic testing, where students filled out a survey about their media preferences (favorite singer, TV shows, etc.), hobbies, interests, wishes, and attitudes and habits about cigarette smoking. The data collected from the diagnostic testing provided information to producers to develop concepts for TV and radio ads. The information was presented by grade level, or “campaign,” in the form of the PRYSM Writer’s Notebook. The first ad campaign occurred in the four states in January 2002, and subsequent campaigns were developed each year. In the final 2005 cycle, 60 concepts were submitted by producers and reviewed by a panel of experts who determined that 22 of them would go into production. Once produced, the PRYSM pre-testing survey was administered in classrooms in Miami and San Antonio. Ads were geared toward three grade levels: 4-6, 7-8, and 9-12. Student input from each level helped PRYSM researchers determine the ads that went on to final production. Students were asked to view or listen to each ad, and then identify which of PRYSM’s four main objectives the ad met (good things happen when you don’t smoke, bad things happen when you do smoke, most kids don’t smoke, and ways to refuse a cigarette). They were then asked to rate how much they liked each ad. Students also were asked to discuss the ads out loud in order to provide ideas about how to make them better. Of the 22 ads students saw, 14 went into final production. In addition to rating the ads, students were asked general demographic questions and about their smoking experience. The answers to these questions determined whether smokers or at-risk smokers preferred different ads than those that were not and whether ad preferences differed by race and gender. The 14 final ads aired between January and April 2005 in the communities participating in the larger study. An analysis sample was obtained from 1,230 youth ages 9-18 in four public school districts in California, Texas, Florida, and Washington, D.C. The districts provided similar proportions of grades 4-6, 7-8, and 9-12, comprised of 34% African American, 31% Hispanic, and 34% non-Hispanic white or other students. Forty-eight classrooms with students from four classrooms in each targeted age category in each district were recruited with a student participation rate of 92.5%. Of those sampled, 55% were classified at high risk of smoking and 45% at low risk.