Factor structure of the Smoking Cessation Self-Efficacy Questionnaire among smokers with and without a psychiatric diagnosis

Abstract

Cessation self-efficacy has been shown to be a consistent predictor of smoking cessation outcomes. To date, no scale assessing cessation self-efficacy has been validated across smokers with and without a psychiatric diagnosis (current or past). Smokers with a psychiatric diagnosis are typically heavy smokers, have a more difficult time quitting, and are more prone to experience lower self-efficacy. Determining whether smoking cessation self-efficacy scores are invariant across these populations is crucial for future research and intervention strategies. Data from the Flexible and Extended Dosing of Nicotine Replacement Therapy (NRT) and Varenicline in Comparison to Fixed Dose NRT for Smoking Cessation: The FLEX Trial, a randomized control trial for smoking cessation, was used to assess the factor structure of the Smoking Cessation Self-Efficacy Questionnaire (SEQ-12), a 12-item scale assessing an individual's confidence to refrain from smoking. Confirmatory factor analysis (CFA) was used to compare the model's fit between the original factor structure and the present data, and to test for measurement invariance across with a current, past, or no psychiatric diagnosis. Initial support was found for both a 2- and 3-factor structure. Using CFA, only the 3-factor model displayed adequate fit indices (Global Fit Index [GFI] = 0.924). Results from the model comparisons showed no differences between those with a current, past, or no psychiatric diagnosis (cmin (30) = 38.64, p = .134). The 3 factors were highly correlated, indicative of an underlying global factor. The SEQ-12 was found to be measurement invariant across treatment-seeking smokers, with preliminary evidence suggesting it is a valid measurement scale for evaluating overall cessation self-efficacy, regardless of psychiatric status. (PsycINFO Database Record