Discussion

         The current study predicted that personal involvement in smoking would lead individuals to misperceive threatening information about the behavior.  Specifically, smokers who are exposed to negative health information about cigarette smoking would be motivated by cognitive dissonance to misperceive this information.  Consequently, it was hypothesized that this group would rate their risk of smoking related diseases to be less than smokers not exposed to negative information and nonsmokers.  These findings demonstrate that smokers exposed to anti-smoking information did not differ in their perception of risk in comparison to smokers in the control condition.  Therefore, smokers exposed to anti-smoking information did not rate their risk of smoking related diseases to be less than those smokers who read facts about Multiple Sclerosis as predicted.  Accordingly, this does not support the prediction that anti-smoking information may motivate smokers to underestimate their perception of risk.

        Possible explanations for these results may be that participants were not highly involved in smoking and did not consider cigarette use a vital aspect of their identity.  Prior research (Freeman et al., 2001) demonstrates that increased personal involvement in a behavior amplifies the amount of discomfort that a dissonant message produces.  A result of this increased dissonance is perceiving the information as threatening to one’s self-concept.  In contrast, less personal involvement in a behavior is associated with more accurate interpretation of information (Freeman et al., 2001).  Accordingly, the findings of the present study suggests that individuals were not involved in smoking enough to perceive the negative smoking information as threatening.  They were not motivated to misperceive the information and underestimate their personal risk for disease.  Instead, smokers’ level of involvement in smoking prompted them to accurately interpret the information.

        Related to these results smokers rated their chances of developing these diseases to be greater than did nonsmokers.  On average smokers rated their risk to be “about the same” as a normal smoker.  This adds support for the notion that smokers who were exposed to the anti-smoking message were able to accurately interpret the information.  They read the information and made reasonable predictions about there own risk for disease.

        Personal involvement in smoking was further assessed by measuring participants’ adherence to smoker stereotypes and attitudes towards smoking. It was hypothesized that smokers’ self-concept would be more consistent with the smoker stereotype than nonsmokers.  In opposition to past research (Freeman et al., 2001), findings suggest that smokers in the current study did not identify with the labels frequently associated with the stereotypical smoker.  These traits included: brave, unhealthy, wise, sociable, smart, beautiful, sexy, popular, outgoing, exciting, and likes groups.  Results illustrated no significant differences between smokers’ and nonsmokers’ adherence to these traits as a whole.  One explanation is that smoking was not an important aspect of the self-concept of smokers in the current study.  Therefore, they did not identify with the label and traits attributed to stereotypical smoker.  This further supports the idea that the sample of smokers was not highly involved in smoking behaviors enough to adhere to related stereotypes.

        When analyzed individually, smokers rated themselves higher on many of traits associated with the stereotypical smoker.  Specifically, smokers scored themselves significantly higher on the traits of brave, unhealthy, and like to be in groups than did nonsmokers.  Although not significant, smokers also indicated higher ratings on the traits of wise, sociable, and smart than did non smokers.  These results suggest that smoking was only slightly relevant to smokers’ self-concept.  Furthermore, only some characteristics attributed to the stereotypical smoker were important to their self-concept. Again this defends the idea that smoking was not a highly relevant behavior to the sample of smokers' identity.

        The current study also predicted that smokers would hold more positive attitudes towards smoking and this would cause them to underestimate their risk for disease.  In concurrence with recent literature (Cui et al., 2002), current findings illustrate the relationship between attitudes, behavior, and risk perception.  As hypothesized, smokers indicated more positive attitudes towards smoking on the overall Smoking Attitudes Scale and on each of its four subscales than nonsmokers.  Specifically, smokers cited more positive views towards interpersonal relationships with smokers, the sale and marketing of cigarettes and health related issues; and more negative views towards smoking restrictions.  This supports the idea that smoking related attitudes are associated with one’s involvement in the behavior.

        Cognitive dissonance theory provides one possible explanation for these results.  According to this model (Freeman et al., 2001), it is logical to observe a consistency between positive attitudes towards smoking and cigarette use.  Any discrepancy between these two would result in psychological discomfort, which would drive the individual to change the problem behavior or modify their attitudes.  Previous findings (Cui et al., 2001) linking tobacco use with positive attitudes towards the behavior have important implications for the current results.  According to these prior findings, when participants indicated involvement in smoking, they rated their attitudes towards the behavior more favorably than periods when they considered themselves nonsmokers throughout the 7 year study (Cui et al., 2001). The  relationship between smoking attitudes and current smoking behavior demonstrates individuals’ needs for balance between their behavior and cognitions.

        Furthermore, current findings indicated that more positive attitudes towards smoking were associated with decreased perception of personal risk. When asked to consider personal risk for developing lung cancer in comparison to an average smoker, participants with more positive attitudes towards smoking indicated the least amount of risk.  This suggests that favorable attitudes towards an smoking motivated individuals to misperceive contrary information in order to maintain a balance between their cognitions.

        The current findings have important implications and applications for the field of health psychology.  Specifically, these results add further support to the existent literature linking attitudes and behavior, and suggest that health campaigns will be most effective when instituted prior to the attainment of positive attitudes towards a harmful activity.  Once an individual becomes highly involved in the behavior, health campaigns may motivate the misperception if the information, instead of attitudinal change.  Thus, the present study implies that positive attitudes towards a harmful behavior will decrease the likelihood that one will accurately interpret dissonant information.  Specifically, this relationship between attitudes and behavior provides valuable information when evaluating the effectiveness of anti-smoking campaigns.  Since attitudes are so strongly embedded in behavior, these findings suggest that the most efficient way of preventing smoking initiation may be to target individuals before they develop positive attitudes toward smoking.

        The current research had several limitations which may have had an effect on the results.  A small sample size is one factor that may have influenced the result of the study.  As previously noted, when asked to rate the probability of developing lung cancer, smokers exposed to the negative smoking information indicated less chance than did smokers in the control condition.  Although these differences were only noted as a trend, it is likely that a larger sample may have shown significant differences between the groups.  Consequently, it is recommended that future research address this issue and recruit a larger sample of participants to investigate if this trend may be significant.

         Also, the participants’ smoking status and level of involvement in smoking may have limited the findings of the current research.  Only 24.1% of the participants indicated daily smoking habits and reported smoking a mean of 4.86 cigarettes each day.  Therefore, the sample of regular smokers may have been inadequate to test the hypotheses of the current study.  Occasional and light smokers who indicated infrequent cigarette use were not highly involved in the behavior. Therefore, they did not perceive smoking to be an integral part of their identity and did not perceive the anti-smoking information as threatening.  Accordingly, they were able to accurately process the information and arrive at an a reasonable conclusion.

        In response to this limitation, it is suggested that future studies focus primarily on daily smokers.  It is recommended that researchers screen participants before their involvement in the study to ensure that smoking is a relevant aspect of their daily lives.  In addition, it would be interesting to compare individuals’ level of smoking involvement and their perception of risk to determine how involved one must be in smoking to be threatened by the anti-smoking information.

         A third limitation of the study was its inability to note any behavioral changes in reaction to the anti-smoking messages.  Since it is proposed that many of the smokers were able to accurately interpret the information, it may be possible that they decided to change their smoking behaviors in response to the dissonant information.  The tendency of dissonance to produce behavioral changes could not be observed in the current research.  Therefore, future research should conduct longitudinal studies that periodically expose smokers to anti-smoking messages and observe cessation attempts throughout the study.

         In conclusion, the present research provides strong support for the relationship between attitudes, behavior, and perception of personal risk.  Specifically, attitudes are reflective of an individual’s involvement in a given behavior and this involvement often prompts the misperception of dissonant information.  When applied to health campaigns, specifically anti-smoking messages, it is apparent that these efforts must be implemented prior to the attainment of positive attitudes and beliefs to ensure the accurate interpretation of the information.