Abstract
The present study constitutes an investigation of tobacco consumption, related attitudes and individual differences in smoking or non-smoking behaviors in a sample of adolescents of different ages in the French-speaking part of Switzerland. We investigated three school-age groups (7th-grade, 9th-grade, and the second-year of high school) for differences in attitude and social and cognitive dimensions. We present both descriptive and inferential statistics. On an inferential level, we present a binary logistic regression-based model predicting risk of smoking. The resulting model most importantly suggests a strong relationship between smoking and alcohol consumption (both regular and sporadic). We interpret this result in terms of both the impact of the actual campaigns and the cognitive processes associated with adolescence.
Acknowledgements
This study was supported by the Tobacco Prevention Office (part of the OFSP) and is part of a two-year project on the creation of relevant warning messages. We are indebted to Marjorie Grivel and Philippe Handschin for helping us to collect the data. Great thanks to Cathy Moeschler, Pierre-Alain Diacon, Jacques Diacon, Karim Errard, Aldo Della Piazza and all the other teachers and students that participated in this phase of the project.
Notes
1. Dichotomized smoking factor/measures (smoker vs. non-smoker) have been often used in previous studies concerning smoking among adolescents (e.g., Zbikowski et al. 2002, De Vries et al. 1988).
2. For clarity reasons, we only limited our analyses on bier, cocktails and alcopops, as all three were most often cited and drank by our participants.
3. Mean scores were calculated on three different items per questions (with a 10-point scale, 1 = Absolutely not important to 10 = Very important): importance to do something now for one's tobacco consumption, readiness to change one's smoking behavior and self-confidence in one's abilities to quit.
4. Because of missing data, logistic regression models only include 262 adolescents (including 38 smokers).
5. Because of missing data on several predictors, the number of adolescents involved in the logistic regression analysis is N=174 (including 35 smokers).