The retest reliability of nicotine dependence measures

Abstract

Aim: As reliable instruments are needed for longitudinal research on the development of nicotine addiction, we evaluated the test–retest reliability of eight survey measures (three scales and five individual item measures).Methods: A paper survey was administered to convenience samples of adolescents in two schools and adults in five workplaces in Argentina and re-administered 1–2 weeks later. Test–retest reliability was assessed with Pearson correlations for continuous measures and kappa statistics for dichotomous measures.Results: Both surveys were completed by 95 adolescent and 88 adult smokers. The modified Fagerström Tolerance Questionnaire (r = 0.88–0.91), the Hooked on Nicotine Checklist (r = 0.85–0.91), the Autonomy Over Smoking Scale (0.91–0.96) and items assessing latency to needing a cigarette (r = 0.85–0.94), the pleasure obtained from smoking (r = 0.84–0.94), and relaxation obtained from the first cigarette (κ = 0.63–0.67), all showed very good to excellent reliability in both populations. Items assessing experiencing the need to smoke more often than previously and the percent of time smoking because of need appear to be more reliable with adolescents (κ = 0.52 and r = 0.76, respectively) than with adults (κ = 0.32 and r = 0.51, respectively).Conclusions: This is the first study to simultaneously compare the reliability of multiple smoking measures within individual adolescent and adult smokers. All the eight measures showed adequate reliability for use in longitudinal studies with adolescents and all but one were adequate for use with adults.