Tobacco smoking constitutes the largest cause of impaired health and premature death in Norway. Notwithstanding this fact, 26 percent of the Norwegian population smoke daily, 11 percent smoke occasionally and still large numbers of adolescents are recruited to the smoking population each year. Thus, there is evidently a need to develop more effective strategies to increase the quit rate among smokers and to ensure that those who do not smoke refrain from smoking. Several authors have emphasised that to further improve our understanding of smoking uptake and cessation, studies need to be theory-driven. This thesis demonstrates how theoretical perspectives from various social psychological frameworks may contribute to explaining smoking cessation and initiation.
Paper I concerned parents' intentions not to smoke indoors in the presence of their children. Paper II addressed students' intentions to quit smoking (Time 1) and the subsequent making of the quitting attempt six months later (Time 2). Paper III concerned adolescents' intentions to reduce smoking (Time 1) and the subsequent behaviour one year later (Time 2), and Paper IV addressed adolescents' intentions to refrain from smoking (Time 1) and self-reported smoking behaviour one year later (Time 2). The impact of the school-based smoking prevention programme BE smokeFREE was also examined.
Common to the papers was that the Theory of Planned Behaviour (TPB) was used as a theoretical framework. According to the TPB, the more positive a person’s attitude, the stronger the subjective norms and the greater the perceived control over the behaviour (PBC), the stronger the intention to perform the behaviour and correspondingly, the more likely the person is to perform the behaviour. The TPB has been successfully applied in predicting a wide range of behaviours, including smoking. Nevertheless, several authors have argued that the model has some important limitations, e.g., (i) the TPB does not capture affective processes, (ii) it does not account for all sources of social influence, and (iii) the specified predictors are not always sufficient in terms of behavioural prediction. Accordingly, research has shown that moral norms, self-identity, group identity, descriptive norms/group norms, anticipated affective reactions, past behaviour, and action planning are useful additions to the TPB. However, these variables have seldom been applied in the domain of smoking. Thus, one aim of this thesis was to obtain a better understanding of factors that are important with respect to smoking initiation and cessation. A second aim was to empirically illuminate the TPB by including additional variables in the model. Hence, in addition to the TPB components, self-identity was assessed in all papers, along with moral norms. Moreover, the impact of past behaviour (Paper I, II & III), descriptive norms/group norms (Paper III & IV), positive and negative anticipated affective reactions (Paper I & II), group identity (Paper II), action planning (Paper III), smoking by valued others (Paper IV), and alcohol use (Paper IV) was examined. Several interaction effects were also tested.
The thesis is based on empirical data from three different studies. Data were collected by questionnaires in three target groups in Norway: Parents of children born in 1998 (mean age = 33 years, Paper I), university students aged 26 years at Time 1 (Paper II), adolescents aged 14 years at Time 1 (Paper III & Paper IV). The samples contained data from 145-722 participants. Two of the surveys were longitudinal, with 6 and 12 months between the measurement points.
The results from Paper I revealed that subjective norms and PBC accounted for 56% (adjusted R2) of the variance in parents’ intentions not to smoke indoors in the presence of their children, while attitude failed to predict intention. Self-identity, moral norms, and the interactions between parent identity and self-identity as a smoker, and between negative anticipated affect and moral norms accounted for 19% (adj. R2) of the variance in intentions, after the effect of the TPB components had been accounted for. Paper II showed that the combination of attitudes, subjective norms and PBC accounted for 36% (adj. R2) of the variance in students’ intentions to quit smoking. Moral norms, positive anticipated affect, group identity and past behaviour explained 9% (adj. R2) of the variance in quitting intentions, beyond the effect of the TPB components. In Paper III attitude, subjective norms and PBC accounted for 28% (adj. R2) of the variance in adolescents’ intentions to reduce smoking. Moral norms, self-identity as a smoker and the group identification by group norm interaction accounted for 10% (adj. R2), after the TPB component had been controlled for. Finally, in Paper IV subjective norms and PBC accounted for 13% (adj. R2) of the variance in adolescents’ intentions to refrain from smoking, while attitude failed to predict intention. Moral norms, self-identity as a non-smoker, descriptive norms, group identification, and perceived social pressure to smoke explained 13% (adj. R2) of the variance in intentions, beyond the effect accounted for by the TPB components.
Paper II showed that intentions to quit smoking predicted the subsequent making of the quitting attempt six months later (adj. R2 = .12), while PBC failed to predict behaviour. Past behaviour, self-identity as a smoker, moral norms, and the interactions between past behaviour and intention, and between negative anticipated affect and moral norms accounted for 9% (adj. R2) of the variance in behaviour, over and above the impact of the TPB components. In Paper III the TPB components did not have a direct impact on smoking reduction one year later, while the PBC by intention interaction was significant (adj. R2 = .01). Past smoking behaviour, self-identity as a smoker, and action planning explained 23% (adj. R2) of the variance in behaviour, beyond the effect accounted for by the TPB components. Paper IV showed that PBC was a significant predictor of smoking initiation one year later, while the impact of intention was marginally significant (R2 = .05). Past smoking behaviour, smoking by family members, perceived social pressure to smoke and alcohol use predicted smoking initiation one year later, accounting for 19% above the impact of the TPB components. Finally, we were not able to conclude that BE smokeFREE had an independent impact on smoking initiation in this group of adolescents.
In sum, the TPB provided good predictions of intentions (adj. R2 = .13-.56). Nevertheless, the inclusion of the extension variables improved the prediction of intentions substantially (R2 increase = .09-.19). The TPB was less successful in predicting behaviour (adj. R2 = .01-.12). However, several of the extension variables contributed in bridging the intention-behaviour gap (R2 increase = .09-.23). The results from these studies have several theoretical implications. In addition to the cognitive processes captured by attitudes in the TPB, smoking decisions were also influenced by positive and negative anticipated emotions. Moreover, subjective norms as operationalized in the TPB were not able to capture all sources of social influence in relation to smoking cessation and smoking initiation. Overt and covert, distal and proximal normative factors were important motivational sources as well as decisive for subsequent behaviour. Smoking can also provide a means of self-portrayal, i.e., social action is transformed into self-identity. When a smoker perceived smoking as being an important part of his/her self, s/he was more resistant to change his/her behaviour. Thus, motivation and perceived control over the behaviour is not always sufficient to explain why people smoke or not. Consistent with this reasoning, action planning appeared as a promising supplement to intentions and PBC in relation to smoking reduction. Nonetheless, the great intention-behaviour gap in this context clearly illustrates the need for further research concerning self-regulative processes. Finally, the results point to factors that are important in order to prevent smoking, and thus some practical implications for future smoking interventions are suggested.