Purpose: The goal of the present study was to determine the variables from the theory of planned behavior that are responsible for differentiating the stages of smoking acquisition.
Methods: A self-administered questionnaire was distributed to 360 secondary I level students from a possible population of 500 students. This questionnaire was designed by the researchers according to Ajzen and Fishbein's and Ajzen's recommendations regarding the predictive model variables. It also contained sociodemographic and personal variables. Data from this cross-sectional study were analyzed using multivariate analysis.
Results: The multivariate analysis of variance indicated an overall significant difference in terms of the three psychosocial variables (Pillais approximation F12,867 = 25.22; p < or = 0.000). The analyses of variance showed that the variables attitude (F4,289 = 112.25; p = 0.0001), subjective norm (F4,289 = 23.48; p = 0.0001), and perceived behavioral control (F4,289 = 87.92; p = 0.0001) contributed to differentiating the stages of smoking acquisition. The discriminant function analysis indicated that attitude (0.826) and perceived behavioral control (0.725) dominate the construct; the contribution of the subjective norm (0.375) is less important. Finally, these variables from the theory of planned behavior permitted the correct classification of 68.4% of adolescents into the stages of smoking acquisition.
Conclusion: The application of the theory of planned behavior seems to increase understanding of the stages of smoking acquisition. The distribution of students in the five stages shows that in the advanced stages, refraining from smoking is more difficult, and that smoking advantages surpass disadvantages. These findings represent only a step toward the development of interventions aimed at warning adolescents against smoking.