Assessing attitudinal ambivalence towards smoking and its association with desire to quit among teen smokers

Abstract

Most smokers have some conflicting thoughts and feelings about their smoking; that is, they feel ambivalent. Whether felt ambivalence can be adequately measured and then used to predict the desire to quit among teen smokers has yet to be explored. Hence, among 402 teen smokers involved in a smoking cessation intervention, we first examined the psychometric properties of an eight-item ambivalence scale via exploratory and confirmatory analyses. After excluding one item, the scale was internally consistent and formed a single factor. We also report on whether felt ambivalence towards smoking assessed at baseline and at four and eight months post-baseline was related to the desire to quit in cross-sectional and prospective analyses. In all cross-sectional and prospective analyses, smokers who felt increasingly ambivalent reported a stronger desire to quit. These results suggest that ambivalence might be a useful construct to explore further and experimentally induce to assess its effects on desire to quit.

Keywords:

• Attitudes
• ambivalence
• smoking
• adolescents

Acknowledgement

Funding for this study was supported through NCI grants CA80262 and CA90716.

Notes

We used a decisional balance scale that was provided to us that, to the best of our knowledge, was being developed for teen smokers; this explains why there was not an equal complement of pro and con items. An item that was noticeably missing from the cons was the importance attached to the negative health effects of smoking. To address this issue, during the eight-month follow-up, we asked teens their absolute and comparative (self vs other) risk of getting a smoking-related disease if they continued to smoke (anchors for absolute risk: 1 = certain not to happen to 7 = certain to happen; anchors for comparative risk: 1 = much less likely to 5 = much more likely), how often they worried about getting a smoking-related disease (1 = never to 5 = always), and how much they worried about getting a smoking-related disease (1 = not at all to 5 = extremely). We expected ambivalence smokers to endorse greater risk and worry. These predictions were not supported in any prospective analyses. However, in the eight-month cross-sectional analyses (N = 174), increasingly ambivalent smokers expressed greater worry (r = 0.29, p < 0.001) and spent more time worrying about getting a smoking-related disease (r = 0.28, p < 0.001). Thus, these data show that ambivalence does relate to worries about smoking-related diseases to some degree.

Due to a clerical error, one of the items of the nonsmoker image scale (It is easy to imagine myself as a nonsmoker.) was not included in approximately 70% of respondents. Hence our analyses are based on the sum of the remaining three items.

For exploratory purposes, we conducted proportional logistic regression models predicting stage of change at four- and eight-months post-baseline from baseline ambivalence, and stage of change at eight-month post-baseline from ambivalence at four months. Individuals who felt more ambivalent at baseline were more likely to be in the preparation stage at four (OR = 1.9, 95% CI = 1.5–2.6; p < 0.001) and at eight-months post-baseline (OR = 1.5, 95% CI = 1.2–2.0; p < 0.001) only. These results provide added support that ambivalence is related to a greater desire to quit using related constructs and measures.