Simple slopes of the belief that smoking enhances a man’s Cad features for men and women on Number of cigarettes smoked daily 

Context in source publication

Context 1

... capable to hit a woman, and impulsive (all above .82”). The third factor was loaded by attractive, kind, physically healthy, intelligent, strong, and good possible long-term partner or husband (all above .47”). The resulting factors were examined for similarity in content and were offered descriptors believed to re fl ect their overall theme: 1) “Cad” (borrowed from Kruger, Fisher, & Jobling, 2003 to describe a competitive, dominant, brave, and promiscuous “dark hero”), 2) Poor impulse-control, 3) “Dad” (“compassionate, industrious, monogamous, proper hero”, Kruger, Fisher, & Jobling, 2003) aspects of a man. The items loading each factor were summed to form composite measures representing those aspects. The alpha coef fi cients for the composite measures were .87, .83, and .79 respectively. We then addressed the question whether the extent to which participant embraced each of the three belief dimensions (the composite measures) could be predicted from the Number of cigarettes smoked daily and the socio-demographic variables of Gender and Age. We assumed it was possible for in fl uence of the independent variables to be moderated, and therefore also entered all interactive terms produced with these variables into our analyses: Gender × Age, Gender × Number of cigarettes, Age × Number of cigarettes, and Gender × Age × Number of cigarettes. We then ran three multiple regression analyses in which all independent variables were entered simultaneously. In Table 2, standardized regression coef fi cients and standard errors for the coef fi cients of each predictor are reported. We found that all three composite dimensions could be directly predicted from the Number of cigarettes smoked daily. Namely, the extent to which people cherished the beliefs that a man’s smoking makes him appear to women as possessing more “Cad” features, as well as the extent to which they believed that smoking enhanced his “Dad” features. Finally, the Number of cigarettes smoked daily was a signi fi cant negative predictor of the belief that smoking makes a man appear as having Poor impulse- control. The more cigarettes people smoked per day the stronger the beliefs about the positive self-presentational impact of cigarette smoking they held. We also found that for the belief that smoking enhanced the “Cad” dimension of a man in female perception was predicted by Age and that in fl uence of Number of smoked cigarettes was moderated by Gender. Using the MODPROBE macro (Hayes & Matthes, 2009), we calculated simple slopes (see Figure 1) for men and women and found that the Number of cigarettes smoked daily predicted the belief that smoking enhanced a man’s “Cad” features only in women ( ß = .255, SE = .082, t = 3.106, p = .002), while in men it had no signi fi cant in fl uence ( ß = .046, SE = .061, t = .748, p = .455). These results indicate that those who most strongly cultivate the belief that smoking enhances Cad features are younger female smokers. The Number of cigarettes smoked daily variable used in the regression analyses was skewed ( M = 3.63, SD = 7.41, min = 0, max = 40, g = 2.36) and any transformations of it (including 1 Cox-Box transformation) did not yield a normally distributed variable. Accordingly, we also tested the current associations using two alternative strategies. In the fi rst, we replaced the Number of cigarettes variable with a dichotomized variable – Smoking status of the participant (participants who smoked zero cigarettes had a score of zero on this variable and participants who smoked at least one cigarette per day had a score of one). The results were perfectly congruent with those reported above obtained using regression with the original Number of cigarettes smoked daily variable - we detected the same effects. The second alternative strategy consisted in a robust method based on an MM-type estimator (Yohai, 1987). Results were exactly the same as the results of ordinary linear regression analysis in the case of the Cad variable, while in the case of Dad features, apart from the reported effects, we also found signi fi cant interaction effects of participant’s Gender with Number of cigarettes smoked daily ( ß = -.125, SE = .0415, t =- 3.013, p = .003) and Age with Number of cigarettes smoked daily ( ß = -.060, SE = .022, t =- 2.776, p = .003). Further inspection of the interaction effects and simple slopes indicated that the Number of cigarettes smoked daily had a stronger effect on female beliefs that smoking enhances Dad aspects of a man ( ß = .432, SE = .069, t =- 6.297, p < .001) than on respective male beliefs ( ß =.182, SE = .045, t = 4.015, p < .001). Independently, an increase in Age served to decrease the effect created by the Number of cigarettes smoked 1 . Similarly, analysis of Poor impulse control variable also yielded one additional interaction effect apart from the one found in our ordinary regression analysis: Age with Number of cigarettes smoked daily, ( ß = .057, SE = .024, t = 2.353, p = .019). The older the participant, the weaker the (negative) effect of Number of cigarettes smoked daily on their belief that smoking impacts impulse control. × 2 (Participant’s smoking status) design, with Age controlled as a covariate. The analysis yielded signi fi cant main effects of Age, F (1,391) = 10.365, p = .001, η p2 = .026, Smoking status, F (1,391) = 16.257, p < .001, η p2 = .040, signi fi cant effect of repeated measures, F (2,782) = 13.333, p < .001, η p2 = .033, and signi fi cant interactive effects of Gender with Smoking status, F (1,391) = 4.562, p = .033, η p2 = .012, Age with repeated measures, F (2,782) = 9.028, p < .001, η p2 = .023, and Smoking status with repeated measures F (2,782) = 20.920, p < .001, η p2 = .051. Smokers held signi fi cantly more positive beliefs about the impact of smoking on all aspects of a man than non-smokers. Closer examination of a marginally insigni fi cant ( p = .08) interactive effect of Gender × Smoking status × repeated measures showed that women’s views were signi fi cantly more polarized than men due to their smoking status, especially on the Cad and Dad dimensions. For means and signi fi cant contrasts see Figure ...