Study 1
The evaluation of the Model Smoking Prevention Program (MSPP) by Perry and colleagues (1992) used a quasi-experimental design to assess the net effect of combining a behavioral health education program in school with the communitywide Minnesota Heart Health Program (MHHP) on youths living in the intervention community. Seven annual waves of cohort and cross-sectional behavioral measurements were collected.
The cohort design, which followed students for 6 years, surveyed students annually. The study took place in Sioux Falls, S.D., and was matched to the Fargo–Moorhead community for comparison. All sixth graders enrolled in public schools in both communities were invited to participate in a baseline survey in April 1983, and that grade cohort was surveyed annually each April until its graduation from high school in 1989. Students received the intervention over a 5-year period. The study evaluated annual self-reported smoking history and, in 1986, included the amount of thiocyanate in saliva samples. The self-report consisted of standardized questions regarding daily smoking, weekly smoking, and smoking history. The saliva samples were obtained from a random sample of approximately half of all classrooms in 1986 (analysis was done on 667 students in the intervention community and 409 students in the comparison communities).
The cross-sectional design consisted of a survey for participants to evaluate smoking intensity and history. Cross-sectional students represented those in the community through the study as well as those not present at baseline or new to the community. The survey took approximately 40 minutes and was completed by students in their health, social science, or English classes. Specialized survey staff were trained to administer the surveys. The prevalence of weekly smoking was used to assess the impact of the intervention, and intensity scores were given to quantify the self-reported measures. Smoking intensity was defined as the number of cigarettes smoked per week by each student, which was then calculated as an average of three self-report questions that related to daily smoking, weekly smoking, and smoking history.
The study examined data using nested cohort and nested cross-sectional designs. In this study, students were observed in classrooms, which were nested within schools, which were nested within the towns assigned to the two treatment conditions. For the cross-sectional analyses, all participants were included. For the cohort analyses, only those who also participated in the baseline survey were included. The study did not provide specific sample sizes or demographics for the intervention and comparison community groups. Subgroup analyses were not conducted.