Adolescent Nonsmokers, Triers, and Regular Smokers' Estimates of Cigarette Smoking Prevalence: When Do Overestimations Occur and by Whom?1

Abstract

The present study tested three assumptions often made by smoking prevention researchers and practitioners regarding the nature and causes of smoking prevalence estimates. Specifically, it is often assumed that (a) adolescent prevalence estimates of all levels of smoking are inflated, (b) youths at different stages of smoking onset all overestimate smoking prevalence, and (c) the predictors of these estimates are the same for youths at different stages of smoking onset. Measures of smoking behavior based on self‐reports using a bogus pipeline procedure, estimates of cigarette smoking prevalence (trying behavior and weekly use), and reports of smoking in significant others were obtained by the same youths when they were in eighth and ninth grades. Results indicated that, taken over all youths, estimates of trying cigarettes were accurate, whereas estimates of weekly use were inflated. When responders were defined by their stage of smoking onset, either as a nonsmoker, trier, or regular smoker, analyses revealed that triers were accurate regarding prevalence estimates of trying, whereas non‐smokers made gross underestimations and regular smokers made overestimations of trying. All groups of youths overestimated prevalence of weekly smoking. Analyses of variance revealed that both youths' smoking status and interpersonal smoking environment additively predicted prevalence estimates, and estimates of smoking prevalence were relatively higher in youths who began to smoke, or increased their levels of smoking, within the next year. Thus, previous assumptions were partially confirmed. Future research and prevention programming should focus more on regular smoking as the focus of normative smoking distortion effects and correction efforts.