The current research aimed to examine the implicit biases of smokers and non-smokers to others who did or did not smoke. Study 1 presented adult smokers and non-smokers with an Implicit Relational Assessment Procedure (IRAP) that assessed bias toward or against smokers and non-smokers. Study 2 replicated this with adolescent smokers and non-smokers. Both studies also presented self-report measures. Both adult and adolescent smokers produced IRAP effects that indicated pro-smoker biases; non-smokers' biases were relatively neutral. Trends in the data from Studies 1 and 2 led to a post hoc analysis of the non-smoker data to investigate the potential impact of parental smoking status on nonsmokers' biases. Both the IRAP and self-report measures data suggested that parental smoking status increased positivity in attitudes toward smokers among non-smokers. Hierarchical logistic regression analyses indicated that the IRAP data in Study 1, but not Study 2, predicted smoking status above and beyond the self-report measures. The post-hoc analyses showed a similar trend. The consistency of the findings with the only existing IRAP study of attitudes toward smokers, as well as with the broader literature, supports the view that response biases toward smokers may not change fundamentally from adolescence to adulthood, and that parental smoking status may having a moderating influence on these biases.
KEYWORDS IRAP, Smokers Attitudes, Parental InfluenceAttitudes toward smokers have changed considerably in recent decades (Chapman & Freeman, 2008). The traditional positive characteristics of smokers as glamorous and independent have been replaced with malodorous and selfish stereotypes (Farrimond & Joffe, 2006). Goldstein (1991), for example, found that non-smokers favored non-smokers over smokers as measured by self-report.Indeed, the literature on attitudes toward smokers has relied extensively on questionnaires, which have been criticized for their potential sensitivity to extraneous sources of influence (e.g., self-presentation, see Barnes-Holmes et al., 2006;de Jong, 2002). In contrast, there are now numerous latency-based behavioral measures, assumed by some to reveal implicit attitudes. The most common of these measures is known as the Implicit Association Test (IAT; Greenwald, McGhee, & Schwartz, 1998).The basic idea is that participants should be faster when pairing two closely associated than weakly associated categories. However, while much research in this regard has investigated attitudes toward smoking, little has employed these measures when investigating attitudes toward smokers.Of the limited research on attitudes toward smokers, Swanson, Rudman, and Greenwald (2001) found that while smokers demonstrated anti-smoking biases, they identified themselves as being a "smoker" on an IAT. In other words, smokers engaged in behavior they did not appear implicitly positive toward, but this did not impact negatively on their self-esteem levels (measured by another IAT). This basic effect was replicated in a study...