A comparison of predictors of the adolescent intention to smoke with adolescent current smoking using discriminant function analysis.

Abstract

This paper explores one possible operationalization of smoking intention to assist development of adolescent smoking intervention programmes. Such programmes usually focus on predictors of current smoking, ignoring predictors of intentions to smoke, or how intentions relate to actual future smoking. Respondents took part in two stages of a repeated measures national survey on Australian adolescent smoking behaviour (12-month interval between administrations). A total of N=1419 adolescents provided matched data, measuring personal smoking habits, intentions of smoking behaviour, stress (seven subscales), self-esteem and response to the General Health Questionnaire (GHQ-12). A comparison of discriminant function predictors showed similarity between the current smoking and intention to smoke, with only one predictor differing between the functions. Non-smokers were more behaviourally consistent with their stated non-smoking intention over 12 months (i.e. remain non-smokers) than smokers (i.e. intention seemed unrelated to actual behaviour). While the predictors give no general indication of whether the processes behind the intention to smoke differ from current smoking, the intention to smoke may be a better focus for intervention with smokers given the apparent inconsistency with stated intention. Intervention for non-smokers may be usefully focused on maintaining non-smoking rather than preventing smoking.