Smoking attitudes, intentions, and behavior among college student smokers: Positivity outweighs negativity

Abstract

Smokers’ attitudes toward smoking tend to be characterized by ambivalent feelings of positivity and negativity, thereby raising the question of whether the effect of smoking attitudes on smoking behavior is mediated by the absence of negativity and/or by the presence of positivity. In two studies, undergraduate smokers rated how positive and negative they felt about smoking and reported how much they had smoked the previous month. Study 2's participants also reported how much they intended to smoke over the next month and, at 1-month follow-up, how much they actually had smoked during the previous month. Positivity, but not negativity, toward smoking predicted past smoking behavior, concurrent intentions to smoke, and future smoking behavior. Results suggest that weakening positive reactions toward smoking may deter smoking more than strengthening negative reactions does. Potential mechanisms are discussed, as well as the prospect of increasing negativity's deterrent effect.

• Smoking
• attitudes
• ambivalence
• positivity
• negativity

Notes

Notes

1. In order to reduce coercion, Introductory Psychology students can opt to write several short papers rather than participate in research studies. Students who chose to complete the mass survey used in Studies 1 and 2 provided informed written consent. All procedures were approved by the university's Institutional Review Board and all data were kept confidential.

2. In both Studies 1 and 2, Cohen and Cohen's (1983; pp. 53–54) test of the difference between independent rs revealed no gender effects on any of the reported correlations.

3. More students completed the mass survey in Fall 2003 than in Spring 2003 because, as is typical, many more students take Introductory Psychology in the fall than in the spring.

4. As in Study 1, all reported results were also obtained when we included raw positivity, intention, and behavior scores in the analysis rather than the transformed scores.

5. Thompson et al. (1995) found that measures of attitudinal positivity and negativity are more highly negatively correlated when they are assessed closer together in time, ostensibly because of participants’ desire to appear consistent. Thus, positivity and negativity may have been negatively correlated in Study 1 because negativity was assessed immediately after positivity. In contrast, in both Study 2 and Lipkus et al.'s (2001) study, measures of polar opposite constructs (e.g., good vs. bad, like vs. dislike) were separated by two other items. Even if they are moderately negatively correlated, the current findings indicate that positivity and negativity toward smoking are sufficiently independent to have distinct effects on behavior.