Background: We conducted a psychometric analysis of an adapted secondhand smoke (SHS) questionnaire by testing the three-component structure of the original scale that measures SHS exposure in home, work and social environments.
Methods: The 15-item questionnaire was administered to 839 daily smokers participating in a multi-site randomized controlled trial. Following parallel analysis, we conducted a confirmatory factor analysis specifying a three-factor structure. Cronbach's alphas and fit indices were calculated to assess internal consistency. Criterion validity was assessed by comparing the Social environments subscale to the Brief Wisconsin Inventory of Smoking Dependence Motives Social/Environmental Goads subscale. Predicative validity of the questionnaire was assessed using linear regressions and tobacco biomarkers of harm; NNAL, expired carbon monoxide and total cotinine.
Results: Five items did not load onto any factor and were dropped, resulting in a 10-item questionnaire. The Cronbach's alphas were (0.86), (0.77) and (0.67) for the Work, Social, and Home subscales, respectively. The WISDM subscale was moderately correlated with scores on the Social subscale (r = 0.57, p < 0.001). The questionnaire demonstrated predictive validity of smoke exposure above individual's own reported use as measured by cigarettes smoked per day.
Conclusions: Three constructs emerged; results indicate that a shortened 10-item scale could be used in future studies.
Keywords: cigarette smoke exposure; environmental tobacco smoke assessments; environmental tobacco smoke exposure; psychometric analysis; risk assessment.