Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Introduction of Section: Persuasion for the Purpose of Cancer Risk Reduction: Understanding Responses to Risk Communications

Introduction of Section: Persuasion for the Purpose of Cancer Risk Reduction: Understanding... Abstract Risk behaviors and responses to risk communications are complex and multifaceted. Two target articles (1,2) conclude that little longitudinal evidence shows that risk perceptions predict precautionary behaviors. This paper focuses on several questions raised by these perplexing findings that have implications for future research on risk communications. A pressing need exists to understand how people process risk information over time and how such processing may differ as a function of risk status, individual differences, social context, and other factors. I will review evidence and methods from the study of persuasion and attitude change that suggest several ways to study message processing to understand what kinds of thoughts are brought to mind following a persuasive communication, as well as how such thoughts may be related to subsequent beliefs and behaviors as people encounter new information and make risk-relevant choices. Exposure to a risk communication or the perception of some internal cue to action, such as a physical symptom, is supposed to heighten perceived risk of illness, which in turn prompts precautionary behavior. Increasingly, researchers have recognized that human behavior rarely conforms to this ideal sequence. First, information that one is at risk for serious illness may often be processed in ways that downplay one's personal risk. Second, the acknowledgement of high personal risk is no guarantee that one will enact precautionary behavior. In particular, the two target articles yield the same conclusion: the evidence linking high-risk perceptions to increased precautionary behavior is limited, both for behavioral risk-reduction measures (1) and for some kinds of cancer screening (2). Instead, risk perceptions may be unrelated to subsequent behavior change and may even predict decreased precautionary behavior. In particular, for behaviors that are complex and for illnesses that are extremely severe or threatening, the proportion of studies showing a relation of risk perceptions to precautionary behavior is quite low (3,4). How and why do responses to risk communications deviate from the ideal sequence outlined above, and when are such deviations problematic? I will examine some reasons why the literature produces such mixed results concerning (a) the relation of risk communications to risk perceptions and (b) the relation of risk perceptions to precautionary behavior. I start with a brief review of contemporary approaches to understanding persuasion and attitude change that may be useful in developing and testing persuasive messages for cancer risk reduction and screening. I will next discuss the importance of understanding how people high and low in different risk factors for cancer may respond differently to risk communications, with implications for subsequent precautionary behavior. I will conclude with recommendations to focus on understanding a wider set of responses to risk communications and to examine personal, social, and experiential influences on such responses over time. Contemporary Approaches to Persuasion and Attitude Change Persuasion and attitude change have been central topics in social psychology since World War II (5-7). Early studies focused on what came to be known as the “who says what to whom” paradigm of the Yale School; i.e., researchers examined the source of a persuasive message (e.g., expertise, trustworthiness), the message itself (e.g., one-sided versus two-sided arguments), and characteristics of the audience (e.g., education, prior knowledge) as determinants of successful persuasion. Contemporary approaches to attitude change, while they consider such factors, focus mostly on additional determinants of the audience's responses to persuasive communications—not only who says what to whom but also with what effect? Three factors have been found to be especially important in understanding how a target audience or individual will process a message and, accordingly, whether the audience will adopt the attitude advocated in the message: (a) motivation to process the persuasive message, (b) ability to process the persuasive message, and (c) cognitive responses to the message. Motivation and Ability to Process the Persuasive Message The most-studied factor in determining motivation to process messages carefully and extensively is self-relevance or personal involvement (6). In experiments that vary the personal relevance of the message (e.g., by having college students read persuasive communications about requiring comprehensive examinations for graduation that will apply to their university or to another university), it is clear that conditions of high personal relevance create greater attention to message content. Highly involved audiences, among other things, are more sensitive to argument quality, responding favorably only to high-quality arguments and rejecting weak, low-quality arguments, even when they are provided by experts. In contrast, less involved audiences devote less effort to processing the message itself and may base their attitudes on factors other than argument quality, such as the expertise or attractiveness of the source. This distinction between more and less effortful processing has also been termed systematic versus heuristic processing (5). Attitudes that are based on careful, systematic processing of a message are more resistant to persuasion than those that are based on less careful examination of the message. Ironically, when the information in question is highly self-relevant and negative (as is typical of cancer risk communications), it is often the case that those factors that increase motivation to process the message (e.g., factors that highlight the personal relevance of the risk information) may simultaneously decrease people's ability to attend closely to it. For example, Jepson and Chaiken (8) found that people with a chronic fear of cancer were less likely to detect logical errors in cancer-related messages than people without such fears. Messages designed to highlight personal vulnerability may create fear, anxiety, and distraction, which, in turn, have been shown to compromise information processing. These factors may lead people to process messages incompletely or to engage in biased processing to reduce their perceived risk. Researchers have known for some time that fear-arousing communications must be paired with information about offsetting or reducing the threat to be effective (9). Other factors, such as time pressure and message complexity, have also been found to reduce people's ability to process persuasive messages. Cognitive Responses to Persuasion The third factor, which is highly influenced by one's motivation and ability to process a message, as well as one's prior knowledge, is cognitive responses to persuasion. Such responses are typically assessed by asking people to list all of the thoughts that come to mind following exposure to a persuasive communication and then coding these thoughts for agreement or disagreement with various aspects of the message. According to the elaboration likelihood model of persuasion of Petty and Cacioppo (6), the success of a persuasive message depends on the thoughts the audience brings to mind following exposure to the message. If the audience generates thoughts agreeing with the message (termed cognitive elaboration), persuasion is more likely; however, if the audience generates thoughts disagreeing with the message (termed counterarguing), the audience is unlikely to be persuaded by the message and is also less likely to be persuaded in the future. Interventions based on the counterarguing concept have been highly successful. In one study (10), junior high school students taught to develop and rehearse counterarguments to advertisements and to peer pressure to smoke were much less likely to initiate smoking than students at a control school. Such results highlight the importance of understanding the thoughts that are brought to mind as people are exposed to risk-related communications and as they encounter risk-related choices in their daily lives as determinants of future attitudes and behaviors. Unfortunately, it is a common finding that people readily generate counterarguments to communications suggesting that they are at high risk for cancer and other illnesses. One particularly common method of counterarguing involves using a higher standard for evidence that is preference inconsistent (e.g., bad news about the consequences of one's own behavior or information that suggests one's health risk is high) than for evidence that is preference consistent (good news about the consequences of one's own behavior or information that suggests one's health risk is low) (11,12). For example, Liberman and Chaiken (13) exposed women who were frequent coffee drinkers to medical articles examining the link between caffeine and fibrocystic breast disease (FBD). Women who were frequent coffee drinkers were much more critical of the studies supporting the link between caffeine and FBD than women who were not frequent coffee drinkers, but they did not apply this same level of scrutiny to articles refuting the dangers of caffeine consumption. Evidence from recent National Cancer Institute (NCI) focus groups (14) is also consistent with these findings, suggesting that people are aware that they counterargue risk communications and that they require a higher standard of evidence to accept bad news about their risk. From the elaboration likelihood theory of Petty and Cacioppo (6), we can predict that such counterarguing—the generation of disagreeing thoughts in response to risk communications—will make people more resistant to persuasion by future risk communications because disagreeing thoughts may be more easily brought to mind with each successive exposure. The challenge to researchers and practitioners, then, is to understand the conditions that are most likely to trigger elaboration versus counterarguing of risk communications and to change the content or delivery of risk communications to reduce counterarguing and to increase message elaboration. In the following sections, I will discuss the specific research and policy recommendations made in the target articles and suggest some additional ways to study people's responses to risk communications in light of these challenges. Recommendations for Research on Risk Communication Test Responses to Risk Communications as a Function of Risk Status My first recommendation is to test cognitive responses to a particular risk communication before using it in an intervention or media campaign. If the message triggers counterarguing in any segment of the audience, that segment may become more resistant to persuasion. Knowing the specific disagreeing thoughts that people may generate would allow researchers and practitioners to build these thoughts into the message; i.e., if we knew what kinds of disagreements were common, we could forestall these disagreements by addressing them explicitly in the persuasive communication. Messages that trigger large numbers of disagreeing thoughts should be reconsidered or abandoned, since they may reinforce the very cognitions the message is designed to change. In many respects, this recommendation resembles an initial strategy of elicitation research in which researchers use open-ended measures, focus groups, and other measures to understand the spontaneously accessible information, motivation, and behavioral skills related to precautionary behaviors in specific risk groups (15). More important, however, the present recommendation suggests that an additional step may be necessary. In addition to trying to understand what people know about their cancer risk to provide them with the information that they need, it is important to understand how they will react when such information is provided. These tests of responses to risk communications should be conducted separately in different risk groups. A pressing need exists to understand how people who are high and low in actual risk, by virtue of their family history of cancer, their own health behavior, or other factors, respond to risk communications. As Vernon (2) points out, although studies have addressed fears surrounding and barriers to obtaining a mammogram, very little is known about the predictors of mammography among women at high risk for breast cancer. Both motivation and ability to process risk communications and the specific kinds of counterarguments generated in response to the message are likely to be markedly different for people with different risk factors. In the high-risk groups, the message is more likely to arouse fear and concern; therefore, adding materials to increase perceived vulnerability to cancer is likely to be counterproductive in that people already tend to know that they are at high risk. What people may want to know is how to reduce their risk or how to profit from early detection if their risk cannot be reduced. Likewise, emerging data concerning the prevalence of a pessimistic bias among people in high-risk groups (i.e., those who overestimate their risk) suggest that efforts to increase risk perceptions are not always necessary but that communications to address people's fears and to provide information about risk-reducing behaviors are. Studies of responses to risk communications about mammography and breast self-examination among high-risk women could tell us a great deal about how people manage the emotional consequences of their high-risk status and how such efforts are related to beliefs about screening and self-examination. In contrast, in low-risk groups, the goal of the communication may be to increase personal vulnerability but only to the level required for people to see the message as personally relevant. Such studies may also provide valuable information about why the predicted relation of risk perceptions to precautionary behavior is so elusive—this relation may be different among high- and low-risk people. Specifically, high perceptions of risk may interfere with precautionary behaviors, and low risk perceptions may not necessarily preclude precautionary behaviors. For these reasons, the goal of risk communication should not necessarily be to make people appreciate their risk of serious illness but instead to increase risk-relevant knowledge, adherence to risk-reduction behaviors, and compliance with screening recommendations in ways that will improve their health and reduce their chances of illness. As noted earlier, the relation of risk perceptions to precautionary behavior is especially weak for highly severe illnesses. One approach that has been successful in increasing precautionary behaviors for other life-threatening illnesses, such as human immunodeficiency virus (HIV) infection, has been to present precautionary behaviors in the context of less serious and less stigmatized risks. For example, in an intervention to reduce young women's risk of contracting HIV, Bryan et al. (16) geared the intervention toward reduction of sexually transmitted diseases (STDs) in general instead of HIV infection specifically. Of course, precautions that reduce one's vulnerability to STDs in general also reduce one's risk of HIV infection, but the recruiting materials and risk communications do not have to work against a strong negative emotional response to HIV or acquired immunodeficiency syndrome to be effective. People engaged in managing a less severe risk may be more able to process risk-relevant information and less motivated to engage in counterarguing. In this way, people may learn and practice risk-reduction behaviors without having to acknowledge their risk of a life-threatening illness. With respect to cancer risks, communications targeted to less threatening health concerns, but that still address the same behavioral risks (diet, smoking, etc.), may be more successful than those that emphasize cancer vulnerability, especially among people at high risk. Similarly, cancer screening might be presented as part of an age-appropriate overall health assessment rather than as a particular set of tests exclusively for cancer. Examine Individual Differences That May Influence Responses to Risk Communication In addition to being negative and threatening, cancer risk communications often have a second set of hurdles to overcome. As Gerrard et al. (1) point out, communications about behavioral risk reduction may also present information that reflects poorly on the self, namely, that one's own behavior is risky or foolish. Gerrard et al. suggest that such information is likely to trigger defensive or reactive responses, especially among people with high self-esteem. On the basis of studies of smoking cessation and sexual risk behavior, they suggest that people with high self-esteem are less likely to acknowledge that their own behavior increases their risk. These findings suggest that some people's motivation to counterargue risk communications is especially high. Therefore, specific risk communications must be more sensitive to the possibility that a message may promote counterarguing, not only to reduce perceptions of personal vulnerability but also to save face, either to oneself or to others. Increasing evidence (17) suggests that many health-risk behaviors, such as excessive dieting, failure to use sunscreen, and failure to wear such protective devices as helmets or life jackets, are related to self-presentational concerns. In addition to being right, people wish to appear “cool,” attractive, strong, and not overly concerned with risks, and such goals may compete against goals to prevent illness. For these reasons, continued efforts to present precautionary behaviors as more socially acceptable and positively, instead of negatively, self-defining are especially important. The findings by Gerrard et al. (1) concerning self-esteem suggest that another important direction for future research will be to examine how people respond to negative information about themselves and to the conditions that may help people to process and elaborate, rather than counterargue, such information. In an experimental test of this question (18), we found that giving people a boost in one area of the self-concept unrelated to health behavior before exposure to information about health risks promoted more open, less biased processing of risk-relevant information. Specifically, women who were frequent caffeine consumers were randomly assigned to complete a 10-item survey about their prior acts of kindness and generosity or to a neutral survey before their exposure to passages confirming or disconfirming the link between caffeine and FBD. The kindness survey was constructed so that all participants would receive high scores. Compared with women in the neutral condition, women who completed the kindness affirmation oriented more quickly to the information suggesting that caffeine was linked to FBD, reported greater beliefs in the link between caffeine and FBD, and reported greater perceived control over reducing their caffeine consumption. It is interesting that they also reported lower intentions to reduce their caffeine consumption at the end of the first session. However, at a 1-week follow-up, women who completed the affirmation were less likely to recall risk-disconfirming information than women in the neutral condition. Put differently, women who did not complete the self-affirmation showed the typical pattern of enhanced recall for information suggesting that their behavior was not risky, but this tendency was eliminated among women in the self-affirmation condition. This particular self-affirmation method has not yet been tested extensively in the health domain, but several studies in other domains reviewed by Aspinwall (19) similarly suggest that boosting people's sense of important self-beliefs, values, or competence in one area may help them manage self-relevant negative information in a second area. These findings suggest that it may be possible to reduce the tendency of people high in self-esteem to counterargue risk information in one domain by reminding them of their triumphs or strengths in other domains. Such affirmations, when carefully constructed so that all participants receive high scores, may also increase the ability of people with low self-esteem to manage negative information about themselves. Understand Responses to Communications About Different Risk Factors for Cancer Understanding how people process information that suggests that their own behavior places them at risk for illness is important, but it may be just as important to study how people respond to information that suggests that there is something else about them—their genetic makeup, their family history, or previous environmental exposure—that increases their risk but cannot be directly modified. The results of recent NCI focus groups (14) suggest an acute need to understand how people think about genetic risk (e.g., is there an inevitable one-to-one re-lation of genetic risk to the development of cancer, or do behaviors and environment still make a difference?). Research (20,21) on adjustment to cancer, as well as to negative life events more generally, suggests that people have a strong tendency to attribute their misfortune to factors within their control. Such beliefs may provide hope that one can reduce the chances of recurrence through one's own behavior, and they are generally associated with good psychologic adjustment, as long as one blames controllable aspects of the situation, such as one's behavior, instead of less mutable aspects, such as one's character or personality. A pressing need exists to communicate advances in genetic testing and other aspects of genetic risk in ways that do not lead people to feel hopeless about their increased risk. Assess Constellations of Risk Perceptions and Related Health Beliefs The preceding sections discussed responses to cancer risk communications in terms of risk status, individual differences, and beliefs regarding the mutability of one's risk. These factors highlight a need to understand the effect of risk communications on a wide set of beliefs related to risk and precautionary behaviors. It will be especially important to be sure that a risk communication that heightens perceived vulnerability does not simultaneously create unfavorable changes in other health beliefs related to precautionary behavior. This possibility may be especially important in the case of several widely studied individual differences that seem to be associated with different effects for different health beliefs. For example, although it appears from the studies by Gerrard et al. (1) that people high in self-esteem are less likely than those low in self-esteem to acknowledge the riskiness of their behavior, high self-esteem may have benefits in terms of people's beliefs about whether they can successfully enact the recommended risk-reduction behavior. Similarly, although people with low self-esteem may not counterargue their risk as vigorously or as successfully as those with high self-esteem, they may not respond as well to risk communications in terms of other beliefs, such as self-efficacy, that have been shown to be important to enacting and maintaining behavior change. Thus, risk communications to increase perceived vulnerability that do not address other health beliefs may easily backfire, making the high self-esteem respondents more defensive and the low self-esteem respondents less confident. A similar set of questions—and potential pitfalls—arises in research on optimistic beliefs. A great deal of research has focused on the possibility that optimists will underestimate their risks of negative events and, correspondingly, on the development of interventions to increase their perceived risk. However, optimism appears to be associated with increased, rather than decreased, attention to health-risk information (22) and to more constructive responses to illness and other stressful events (23), suggesting that it is beneficial on balance. The corresponding question of whether pessimism may undermine risk-reduction behavior has yet to be fully examined. In several studies reviewed by Scheier and his colleagues (23), people who are pessimistic are more likely to engage in avoidant coping, denial, and disengagement when faced with stressful circumstances. People who are pessimistic may acknowledge their risk to a greater extent (and perhaps even overestimate it, although this has yet to be determined), but they may fail to act to reduce their risk because they do not believe such responses will be successful in reducing their risk. As these examples illustrate, it may be important to consider the relation of any given individual difference to a wide range of relevant beliefs that may affect precautionary behaviors, not just perceived risk, and to develop risk communications and other interventions that address these related health beliefs. Assess Multiple Influences on Responses to Risk Communications Over Time Both target articles (1,2) listed several methodologic reasons why cross-sectional studies fail to provide convincing or useful information about the relation between risk perceptions and precautionary behavior. To these vital and frequently overlooked points, I would like to add another point that has been virtually unaddressed in the literature. Often we simply do not know what happens in the interim between exposure to a risk communication and some subsequent behavioral outcome that would explain a positive, negative, or null relation between risk perceptions and precautionary behavior. There are reasons to believe that several factors may influence such relations, including how the risk communication was processed at the time of exposure, the success or failure of subsequent efforts to enact risk-reducing behavior, the processing of subsequent risk-related information, and social influences on both responses to risk information and precautionary behavior. A comprehensive review of such factors is beyond the scope of this paper, but I will provide some examples to illustrate how these factors may influence risk perceptions, precautionary behavior, and subsequent processing of risk communications. An important initial question may be, “Where does the risk information go and why?” That is, what determines whether a risk communication is counterargued or weakened by conflicting information, as opposed to being elaborated and bolstered by additional consistent information? Interesting and potentially important differences in recall for risk-related information can be found over very short periods of time, yet little is known about how such differences are created and maintained. For example, in a 1-week follow-up to the study of women who were frequent caffeine drinkers (18), those who did not complete the affirmation reported greater numbers of risk-disconfirming facts on a surprise free recall test than those who completed the affirmation. What intervening process(es) maintained such differential recall? It is unlikely that such differences could be explained simply by a single instance of defensive processing or counterarguing at the time of exposure; instead, it is possible that these participants reminded themselves of the reasons caffeine was not harmful to their health with each cup of coffee, thus repeatedly counterarguing the risk communication in the week following exposure. Conversely, how did those who completed the affirmation differ in the thoughts they brought to mind each time a choice about consuming caffeine arose? Understanding how different aspects of the persuasive message may be brought to mind as people make risk-relevant choices will be essential to understanding how risk communications may influence beliefs and behaviors over time. A second, closely related factor is understanding how risk perceptions are related to successful or unsuccessful attempts to enact precautionary behaviors. Risk perceptions may indeed prompt initial efforts to undertake precautionary behaviors, but people may experience frustration, embarrassment, distress, or failure in their efforts. These difficulties may influence subsequent risk perceptions. Recall that the smokers studied by Gerrard et al. (1) who failed in their attempts to quit smoking were the ones who subsequently downplayed the risks of cigarette smoking. Risk perceptions and behaviors are likely to be reciprocally related as people encounter risk-relevant choices on a frequent—and even daily—basis. Experience sampling or daily diary studies might be very useful in understanding how people think about risk information following exposure, especially when that information involves daily behavior and choices, such as diet, exercise, and smoking. Identifying the stumbling blocks on the path to precautionary behavior and how such difficulties affect subsequent risk perceptions and processing of risk information is essential. Third, as these examples suggest, it is important to understand how initial responses to a risk communication may influence the processing of subsequent risk information. Does initial disagreement with a risk communication make it more difficult for subsequent messages to be processed veridically? Are such effects compounded with repeated exposure to risk communications? Such phenomena may be especially amenable to analysis in terms of theories of responses to persuasion, in that counterarguing makes disagreeing thoughts more accessible and therefore more likely to be evoked by subsequent risk communications. Related topics for future study might include understanding other aspects of how people manage multiple risk messages. People are exposed to repeated messages about diet, exercise, screening, and other health-related topics, not just those from a planned intervention. It will be important to understand how prior message processing affects responses to this new information, as well as people's strategies for dealing with an ever-increasing volume of information about health risks. Do people become discouraged as they realize that they cannot simultaneously address dozens of potential threats to health? Do people choose to focus on one or two risks and ignore the rest? Studies of people's strategies for dealing with such information on a daily basis would likely give insights into such processes. In this vein, the recent NCI focus group report (14) on how the public perceives, processes, and interprets risk information is a valuable start. Understanding how these processes may differ as a function of risk status may also reveal some important information about the relation of risk perceptions to multiple message processing over time. A final factor that is relatively understudied in understanding responses to risk communications is how the social interactions of people exposed to risk communications bolster or degrade the effects of the message. For cancer and many other illnesses, the beliefs and behaviors of other people have been shown to be highly influential in how people understand their own risk and whether they enact precautionary behavior (24,25). It will be critical to examine how specific risk communications may be supported or discounted by similar others' views and experiences. Such responses may be especially important in understanding responses to risk communications among people at high risk for cancer, as the desire for information from and about other people has been shown to increase sharply as a function of distress or uncertainty. A daily diary study could answer some of these questions by tracking how people share risk-related information with others following exposure to a risk communication and are influenced by their reactions to the information and its recommendations. Such a study would provide valuable information about how social factors, for better or for worse, influence subsequent emotions, thoughts, and behaviors related to the risk. One promising approach to creating social support for risk perceptions and precautionary behaviors is to conduct risk-communication interventions in small peer groups (16,26). In this way, people see that others like them are similarly concerned about their risk. Such approaches may increase support for beliefs and behaviors related to risk reduction, while at the same time decreasing self-presentational concerns that may interfere with precautionary behaviors. Conclusion The points raised in the target articles and in this brief discussion of responses to persuasive messages suggest that there are many reasons that people deviate from the ideal sequence in which exposure to risk communications leads to perceived risk and, ultimately, to precautionary behavior. First, there may be a complex series of trade-offs that influence people's responses to risk communications. Factors that increase people's motivation to process a message may interfere with their ability to do so; similarly, factors that increase one's sense of personal vulnerability may also create feelings of pessimism and hopelessness that interfere with risk-reduction behavior. As Rothman and Kiviniemi (27) point out, people who acknowledge increased personal vulnerability may also simultaneously change other perceptions of the health threat, such as its seriousness or prevalence. It is unknown how these different factors work together to influence precautionary behavior and subsequent processing of risk-related information. The recognition of this complex—and often conflicting—set of responses to risk communications suggests several different approaches for future research. One approach would be to subject major risk communication campaigns to careful pilot studies of responses to risk communications by use of multiple measures of information processing, cognitive elaboration, and related health beliefs to understand what people attend to, what thoughts are generated by the message (agreeing? disagreeing?), how the information is evaluated (convincing? trustworthy?), what people recall following exposure, and how and whether people believe they can modify their risk. It will be essential to conduct these tests in high- and low-risk groups and among people with different risk factors for cancer. Second, persuasive communications to increase perceived risk of cancer must be sensitive to several factors that, unless handled carefully, could result in extensive counterarguing and, therefore, increased resistance to change. Special care should be taken to present information about risk behaviors that may be self-defining or that have important self-presentational or other interpersonal functions. It is difficult to alter such behaviors without making the recommended risk-reduction behaviors more socially acceptable or normative and without addressing reasons other than health-related ones that people practice a particular health-risk behavior. A third pressing need is to understand how people manage risk-related information over time and what happens as they try to incorporate this information and the behavioral changes it suggests into their daily lives. Where does the risk information “go,” and why do risk perceptions bear an inconsistent—and sometimes counterproductive—relation to precautionary behaviors? These factors may be different for high- and low-risk groups in ways that may inform the design and delivery of risk communications and of other interventions to maintain support for precautionary behaviors. Additionally, individual differences, such as self-esteem and optimism, may play an important role in such responses, but they may have complex relations to subsequent risk perceptions and behavior. Only prospective studies that examine a range of outcome measures can tell us how different responses to risk communication may be related to risk beliefs and precautionary behaviors over time. Such studies are essential in that the target behaviors, such as changes in diet and exercise and adherence to screening recommendations, must be practiced consistently over time to be effective. Daily diary studies may provide valuable information about how people manage risk information in their social interactions and behavioral choices and how such processes affect risk perceptions and precautionary behaviors over time. Finally, it may be well worthwhile to consider one of the major conclusions of Vernon's (2) review. Her suggestion to focus directly on social and structural factors that may influence cancer screening, such as cultural beliefs, socioeconomic status, insurance coverage, and physician recommendations, is an important reminder that precautionary behaviors are not always purely personal or individual but may also depend on a broader context of social and structural support for such behaviors. I thank Meg Gerrard for her helpful comments on an earlier version of this manuscript. References (1) Gerrard M, Gibbons FX, Reis-Bergan M. The effect of risk communication on risk perceptions: The significance of individual differences. Monogr Natl Cancer Inst  1999; 25: 94-100. Google Scholar (2) Vernon SW. Risk perception and risk communication for cancer screening behaviors: a review. Monogr Natl Cancer Inst  1999; 25: 101-19. Google Scholar (3) Gerrard M, Gibbons FX, Bushman BJ. Relation between perceived vulnerability to HIV and precautionary sexual behavior. Psychol Bull  1996; 119: 390-409. Google Scholar (4) Weinstein ND. The precaution adoption process. Health Psychol  1988; 7: 355-86. Google Scholar (5) Eagly AH, Chaiken S. The psychology of attitudes. Fort Worth (TX): Harcourt, Brace Jovanovich; 1993. Google Scholar (6) Petty RE, Cacioppo JT. Communication and persuasion: central and peripheral routes to attitude change. New York (NY): Springer-Verlag; 1986. Google Scholar (7) Petty RE, Wegener DT. Attitude change: multiple roles for persuasion variables. In: Gilbert DT, Fiske ST, Lindzey G, editors. Handbook of social psychology. 4th ed. Boston (MA): McGraw-Hill; 1998. p. 323-90. Google Scholar (8) Jepson C, Chaiken S. Chronic issue-specific fear inhibits systematic processing of persuasive communications. J Soc Beh Pers  1990; 5: 61-84. Google Scholar (9) Leventhal H. Findings and theory in the study of fear communications. Adv Exp Soc Psychol  1970; 5: 119-86. Google Scholar (10) McAlister A, Perry C, Killen J, Slinkard LA, Maccoby N. Pilot study of smoking, alcohol and drug abuse prevention. Am J Public Health  1980; 70: 719-21. Google Scholar (11) Ditto PH, Lopez DF. Motivated skepticism: use of differential decision criteria for preferred and nonpreferred conclusions. J Pers Soc Psychol  1992; 63: 568-84. Google Scholar (12) Ditto PH, Scepansky JA, Munro GD, Apanovitch AM, Lockhart LK. Motivated sensitivity to preference-inconsistent information. J Pers Soc Psychol  1998; 75: 53-69. Google Scholar (13) Liberman A, Chaiken S. Defensive processing of personally relevant health messages. Pers Soc Psychol Bull  1992; 18: 669-79. Google Scholar (14) National Cancer Institute, Office of Cancer Communications. How the public perceives, processes, and interprets risk information: Findings from focus group research with the general public. Bethesda (MD): Natl Cancer Inst; 1998 Jun Report No.: POS-T086. Google Scholar (15) Fisher JD, Fisher WA. Changing AIDS-risk behavior. Psychol Bull  1992; 111: 455-74. Google Scholar (16) Bryan AD, Aiken LS, West SG. Increasing condom use: evaluation of a theory-based intervention to prevent sexually transmitted diseases in young women. Health Psychol  1996; 15: 371-82. Google Scholar (17) Leary MR, Tchividjian LR, Kraxberger BE. Self-presentation can be hazardous to your health: impression management and health risk. Health Psychol  1994; 13: 461-70. Google Scholar (18) Reed MB, Aspinwall LG. Self-affirmation reduces biased processing of health-risk information. Motivation Emotion  1998; 22: 99-132. Google Scholar (19) Aspinwall LG. Rethinking the role of positive affect in self-regulation. Motivation Emotion  1998; 22: 1-32. Google Scholar (20) Taylor SE. Adjustment to threatening events: a theory of cognitive adaptation. Am Psychol  1983; 38: 1163-73. Google Scholar (21) Janoff-Bulman R. Assumptive worlds and the stress of traumatic events: applications of the schema construct. Soc Cogn  1989; 7: 113-36. Google Scholar (22) Aspinwall LG, Brunhart SM. Distinguishing optimism from denial: optimistic beliefs predict attention to health threats. Pers Soc Psychol Bull  1996; 22: 993-1003. Google Scholar (23) Scheier MF, Carver CS, Bridges MW. Optimism, pessimism, and psychological well-being. In: Chang EC, editor. Optimism and pessimism: theory, research, and practice. Washington (DC): American Psychological Association. In press 1999. Google Scholar (24) Buunk BP, Gibbons FX, editors. Health, coping, and well-being: perspectives from social comparison theory. Mahwah (NJ): Erlbaum; 1997. Google Scholar (25) Croyle RT. Appraisal of health threats: cognition, motivation, and social comparison. Cogn Ther Res  1992; 16: 165-82. Google Scholar (26) Fisher JD, Fisher WA, Misovich SJ, Kimble DL, Malloy TE. Changing AIDS risk behavior: effects of an intervention emphasizing AIDS risk reduction information, motivation, and behavioral skills in a college population. Health Psychol  1996; 15: 114-23. Google Scholar (27) Rothman AJ, Kiviniemi MT. “Treating people with information”: An analysis and review of approaches to communicating health risk information. Monogr Natl Cancer Inst  1999; 25: 44-51. Google Scholar Oxford University Press http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JNCI Monographs Oxford University Press

Introduction of Section: Persuasion for the Purpose of Cancer Risk Reduction: Understanding Responses to Risk Communications

Aspinwall, Lisa G.
JNCI Monographs , Volume 1999 (25) – Jan 1, 1999
Download PDF
6 pages

Loading next page...
 
/lp/oxford-university-press/introduction-of-section-persuasion-for-the-purpose-of-cancer-risk-CvIMFYECBs

References (29)

Publisher
Oxford University Press
Copyright
Oxford University Press
ISSN
1052-6773
eISSN
1745-6614
DOI
10.1093/oxfordjournals.jncimonographs.a024216
Publisher site
See Article on Publisher Site

Abstract

Abstract Risk behaviors and responses to risk communications are complex and multifaceted. Two target articles (1,2) conclude that little longitudinal evidence shows that risk perceptions predict precautionary behaviors. This paper focuses on several questions raised by these perplexing findings that have implications for future research on risk communications. A pressing need exists to understand how people process risk information over time and how such processing may differ as a function of risk status, individual differences, social context, and other factors. I will review evidence and methods from the study of persuasion and attitude change that suggest several ways to study message processing to understand what kinds of thoughts are brought to mind following a persuasive communication, as well as how such thoughts may be related to subsequent beliefs and behaviors as people encounter new information and make risk-relevant choices. Exposure to a risk communication or the perception of some internal cue to action, such as a physical symptom, is supposed to heighten perceived risk of illness, which in turn prompts precautionary behavior. Increasingly, researchers have recognized that human behavior rarely conforms to this ideal sequence. First, information that one is at risk for serious illness may often be processed in ways that downplay one's personal risk. Second, the acknowledgement of high personal risk is no guarantee that one will enact precautionary behavior. In particular, the two target articles yield the same conclusion: the evidence linking high-risk perceptions to increased precautionary behavior is limited, both for behavioral risk-reduction measures (1) and for some kinds of cancer screening (2). Instead, risk perceptions may be unrelated to subsequent behavior change and may even predict decreased precautionary behavior. In particular, for behaviors that are complex and for illnesses that are extremely severe or threatening, the proportion of studies showing a relation of risk perceptions to precautionary behavior is quite low (3,4). How and why do responses to risk communications deviate from the ideal sequence outlined above, and when are such deviations problematic? I will examine some reasons why the literature produces such mixed results concerning (a) the relation of risk communications to risk perceptions and (b) the relation of risk perceptions to precautionary behavior. I start with a brief review of contemporary approaches to understanding persuasion and attitude change that may be useful in developing and testing persuasive messages for cancer risk reduction and screening. I will next discuss the importance of understanding how people high and low in different risk factors for cancer may respond differently to risk communications, with implications for subsequent precautionary behavior. I will conclude with recommendations to focus on understanding a wider set of responses to risk communications and to examine personal, social, and experiential influences on such responses over time. Contemporary Approaches to Persuasion and Attitude Change Persuasion and attitude change have been central topics in social psychology since World War II (5-7). Early studies focused on what came to be known as the “who says what to whom” paradigm of the Yale School; i.e., researchers examined the source of a persuasive message (e.g., expertise, trustworthiness), the message itself (e.g., one-sided versus two-sided arguments), and characteristics of the audience (e.g., education, prior knowledge) as determinants of successful persuasion. Contemporary approaches to attitude change, while they consider such factors, focus mostly on additional determinants of the audience's responses to persuasive communications—not only who says what to whom but also with what effect? Three factors have been found to be especially important in understanding how a target audience or individual will process a message and, accordingly, whether the audience will adopt the attitude advocated in the message: (a) motivation to process the persuasive message, (b) ability to process the persuasive message, and (c) cognitive responses to the message. Motivation and Ability to Process the Persuasive Message The most-studied factor in determining motivation to process messages carefully and extensively is self-relevance or personal involvement (6). In experiments that vary the personal relevance of the message (e.g., by having college students read persuasive communications about requiring comprehensive examinations for graduation that will apply to their university or to another university), it is clear that conditions of high personal relevance create greater attention to message content. Highly involved audiences, among other things, are more sensitive to argument quality, responding favorably only to high-quality arguments and rejecting weak, low-quality arguments, even when they are provided by experts. In contrast, less involved audiences devote less effort to processing the message itself and may base their attitudes on factors other than argument quality, such as the expertise or attractiveness of the source. This distinction between more and less effortful processing has also been termed systematic versus heuristic processing (5). Attitudes that are based on careful, systematic processing of a message are more resistant to persuasion than those that are based on less careful examination of the message. Ironically, when the information in question is highly self-relevant and negative (as is typical of cancer risk communications), it is often the case that those factors that increase motivation to process the message (e.g., factors that highlight the personal relevance of the risk information) may simultaneously decrease people's ability to attend closely to it. For example, Jepson and Chaiken (8) found that people with a chronic fear of cancer were less likely to detect logical errors in cancer-related messages than people without such fears. Messages designed to highlight personal vulnerability may create fear, anxiety, and distraction, which, in turn, have been shown to compromise information processing. These factors may lead people to process messages incompletely or to engage in biased processing to reduce their perceived risk. Researchers have known for some time that fear-arousing communications must be paired with information about offsetting or reducing the threat to be effective (9). Other factors, such as time pressure and message complexity, have also been found to reduce people's ability to process persuasive messages. Cognitive Responses to Persuasion The third factor, which is highly influenced by one's motivation and ability to process a message, as well as one's prior knowledge, is cognitive responses to persuasion. Such responses are typically assessed by asking people to list all of the thoughts that come to mind following exposure to a persuasive communication and then coding these thoughts for agreement or disagreement with various aspects of the message. According to the elaboration likelihood model of persuasion of Petty and Cacioppo (6), the success of a persuasive message depends on the thoughts the audience brings to mind following exposure to the message. If the audience generates thoughts agreeing with the message (termed cognitive elaboration), persuasion is more likely; however, if the audience generates thoughts disagreeing with the message (termed counterarguing), the audience is unlikely to be persuaded by the message and is also less likely to be persuaded in the future. Interventions based on the counterarguing concept have been highly successful. In one study (10), junior high school students taught to develop and rehearse counterarguments to advertisements and to peer pressure to smoke were much less likely to initiate smoking than students at a control school. Such results highlight the importance of understanding the thoughts that are brought to mind as people are exposed to risk-related communications and as they encounter risk-related choices in their daily lives as determinants of future attitudes and behaviors. Unfortunately, it is a common finding that people readily generate counterarguments to communications suggesting that they are at high risk for cancer and other illnesses. One particularly common method of counterarguing involves using a higher standard for evidence that is preference inconsistent (e.g., bad news about the consequences of one's own behavior or information that suggests one's health risk is high) than for evidence that is preference consistent (good news about the consequences of one's own behavior or information that suggests one's health risk is low) (11,12). For example, Liberman and Chaiken (13) exposed women who were frequent coffee drinkers to medical articles examining the link between caffeine and fibrocystic breast disease (FBD). Women who were frequent coffee drinkers were much more critical of the studies supporting the link between caffeine and FBD than women who were not frequent coffee drinkers, but they did not apply this same level of scrutiny to articles refuting the dangers of caffeine consumption. Evidence from recent National Cancer Institute (NCI) focus groups (14) is also consistent with these findings, suggesting that people are aware that they counterargue risk communications and that they require a higher standard of evidence to accept bad news about their risk. From the elaboration likelihood theory of Petty and Cacioppo (6), we can predict that such counterarguing—the generation of disagreeing thoughts in response to risk communications—will make people more resistant to persuasion by future risk communications because disagreeing thoughts may be more easily brought to mind with each successive exposure. The challenge to researchers and practitioners, then, is to understand the conditions that are most likely to trigger elaboration versus counterarguing of risk communications and to change the content or delivery of risk communications to reduce counterarguing and to increase message elaboration. In the following sections, I will discuss the specific research and policy recommendations made in the target articles and suggest some additional ways to study people's responses to risk communications in light of these challenges. Recommendations for Research on Risk Communication Test Responses to Risk Communications as a Function of Risk Status My first recommendation is to test cognitive responses to a particular risk communication before using it in an intervention or media campaign. If the message triggers counterarguing in any segment of the audience, that segment may become more resistant to persuasion. Knowing the specific disagreeing thoughts that people may generate would allow researchers and practitioners to build these thoughts into the message; i.e., if we knew what kinds of disagreements were common, we could forestall these disagreements by addressing them explicitly in the persuasive communication. Messages that trigger large numbers of disagreeing thoughts should be reconsidered or abandoned, since they may reinforce the very cognitions the message is designed to change. In many respects, this recommendation resembles an initial strategy of elicitation research in which researchers use open-ended measures, focus groups, and other measures to understand the spontaneously accessible information, motivation, and behavioral skills related to precautionary behaviors in specific risk groups (15). More important, however, the present recommendation suggests that an additional step may be necessary. In addition to trying to understand what people know about their cancer risk to provide them with the information that they need, it is important to understand how they will react when such information is provided. These tests of responses to risk communications should be conducted separately in different risk groups. A pressing need exists to understand how people who are high and low in actual risk, by virtue of their family history of cancer, their own health behavior, or other factors, respond to risk communications. As Vernon (2) points out, although studies have addressed fears surrounding and barriers to obtaining a mammogram, very little is known about the predictors of mammography among women at high risk for breast cancer. Both motivation and ability to process risk communications and the specific kinds of counterarguments generated in response to the message are likely to be markedly different for people with different risk factors. In the high-risk groups, the message is more likely to arouse fear and concern; therefore, adding materials to increase perceived vulnerability to cancer is likely to be counterproductive in that people already tend to know that they are at high risk. What people may want to know is how to reduce their risk or how to profit from early detection if their risk cannot be reduced. Likewise, emerging data concerning the prevalence of a pessimistic bias among people in high-risk groups (i.e., those who overestimate their risk) suggest that efforts to increase risk perceptions are not always necessary but that communications to address people's fears and to provide information about risk-reducing behaviors are. Studies of responses to risk communications about mammography and breast self-examination among high-risk women could tell us a great deal about how people manage the emotional consequences of their high-risk status and how such efforts are related to beliefs about screening and self-examination. In contrast, in low-risk groups, the goal of the communication may be to increase personal vulnerability but only to the level required for people to see the message as personally relevant. Such studies may also provide valuable information about why the predicted relation of risk perceptions to precautionary behavior is so elusive—this relation may be different among high- and low-risk people. Specifically, high perceptions of risk may interfere with precautionary behaviors, and low risk perceptions may not necessarily preclude precautionary behaviors. For these reasons, the goal of risk communication should not necessarily be to make people appreciate their risk of serious illness but instead to increase risk-relevant knowledge, adherence to risk-reduction behaviors, and compliance with screening recommendations in ways that will improve their health and reduce their chances of illness. As noted earlier, the relation of risk perceptions to precautionary behavior is especially weak for highly severe illnesses. One approach that has been successful in increasing precautionary behaviors for other life-threatening illnesses, such as human immunodeficiency virus (HIV) infection, has been to present precautionary behaviors in the context of less serious and less stigmatized risks. For example, in an intervention to reduce young women's risk of contracting HIV, Bryan et al. (16) geared the intervention toward reduction of sexually transmitted diseases (STDs) in general instead of HIV infection specifically. Of course, precautions that reduce one's vulnerability to STDs in general also reduce one's risk of HIV infection, but the recruiting materials and risk communications do not have to work against a strong negative emotional response to HIV or acquired immunodeficiency syndrome to be effective. People engaged in managing a less severe risk may be more able to process risk-relevant information and less motivated to engage in counterarguing. In this way, people may learn and practice risk-reduction behaviors without having to acknowledge their risk of a life-threatening illness. With respect to cancer risks, communications targeted to less threatening health concerns, but that still address the same behavioral risks (diet, smoking, etc.), may be more successful than those that emphasize cancer vulnerability, especially among people at high risk. Similarly, cancer screening might be presented as part of an age-appropriate overall health assessment rather than as a particular set of tests exclusively for cancer. Examine Individual Differences That May Influence Responses to Risk Communication In addition to being negative and threatening, cancer risk communications often have a second set of hurdles to overcome. As Gerrard et al. (1) point out, communications about behavioral risk reduction may also present information that reflects poorly on the self, namely, that one's own behavior is risky or foolish. Gerrard et al. suggest that such information is likely to trigger defensive or reactive responses, especially among people with high self-esteem. On the basis of studies of smoking cessation and sexual risk behavior, they suggest that people with high self-esteem are less likely to acknowledge that their own behavior increases their risk. These findings suggest that some people's motivation to counterargue risk communications is especially high. Therefore, specific risk communications must be more sensitive to the possibility that a message may promote counterarguing, not only to reduce perceptions of personal vulnerability but also to save face, either to oneself or to others. Increasing evidence (17) suggests that many health-risk behaviors, such as excessive dieting, failure to use sunscreen, and failure to wear such protective devices as helmets or life jackets, are related to self-presentational concerns. In addition to being right, people wish to appear “cool,” attractive, strong, and not overly concerned with risks, and such goals may compete against goals to prevent illness. For these reasons, continued efforts to present precautionary behaviors as more socially acceptable and positively, instead of negatively, self-defining are especially important. The findings by Gerrard et al. (1) concerning self-esteem suggest that another important direction for future research will be to examine how people respond to negative information about themselves and to the conditions that may help people to process and elaborate, rather than counterargue, such information. In an experimental test of this question (18), we found that giving people a boost in one area of the self-concept unrelated to health behavior before exposure to information about health risks promoted more open, less biased processing of risk-relevant information. Specifically, women who were frequent caffeine consumers were randomly assigned to complete a 10-item survey about their prior acts of kindness and generosity or to a neutral survey before their exposure to passages confirming or disconfirming the link between caffeine and FBD. The kindness survey was constructed so that all participants would receive high scores. Compared with women in the neutral condition, women who completed the kindness affirmation oriented more quickly to the information suggesting that caffeine was linked to FBD, reported greater beliefs in the link between caffeine and FBD, and reported greater perceived control over reducing their caffeine consumption. It is interesting that they also reported lower intentions to reduce their caffeine consumption at the end of the first session. However, at a 1-week follow-up, women who completed the affirmation were less likely to recall risk-disconfirming information than women in the neutral condition. Put differently, women who did not complete the self-affirmation showed the typical pattern of enhanced recall for information suggesting that their behavior was not risky, but this tendency was eliminated among women in the self-affirmation condition. This particular self-affirmation method has not yet been tested extensively in the health domain, but several studies in other domains reviewed by Aspinwall (19) similarly suggest that boosting people's sense of important self-beliefs, values, or competence in one area may help them manage self-relevant negative information in a second area. These findings suggest that it may be possible to reduce the tendency of people high in self-esteem to counterargue risk information in one domain by reminding them of their triumphs or strengths in other domains. Such affirmations, when carefully constructed so that all participants receive high scores, may also increase the ability of people with low self-esteem to manage negative information about themselves. Understand Responses to Communications About Different Risk Factors for Cancer Understanding how people process information that suggests that their own behavior places them at risk for illness is important, but it may be just as important to study how people respond to information that suggests that there is something else about them—their genetic makeup, their family history, or previous environmental exposure—that increases their risk but cannot be directly modified. The results of recent NCI focus groups (14) suggest an acute need to understand how people think about genetic risk (e.g., is there an inevitable one-to-one re-lation of genetic risk to the development of cancer, or do behaviors and environment still make a difference?). Research (20,21) on adjustment to cancer, as well as to negative life events more generally, suggests that people have a strong tendency to attribute their misfortune to factors within their control. Such beliefs may provide hope that one can reduce the chances of recurrence through one's own behavior, and they are generally associated with good psychologic adjustment, as long as one blames controllable aspects of the situation, such as one's behavior, instead of less mutable aspects, such as one's character or personality. A pressing need exists to communicate advances in genetic testing and other aspects of genetic risk in ways that do not lead people to feel hopeless about their increased risk. Assess Constellations of Risk Perceptions and Related Health Beliefs The preceding sections discussed responses to cancer risk communications in terms of risk status, individual differences, and beliefs regarding the mutability of one's risk. These factors highlight a need to understand the effect of risk communications on a wide set of beliefs related to risk and precautionary behaviors. It will be especially important to be sure that a risk communication that heightens perceived vulnerability does not simultaneously create unfavorable changes in other health beliefs related to precautionary behavior. This possibility may be especially important in the case of several widely studied individual differences that seem to be associated with different effects for different health beliefs. For example, although it appears from the studies by Gerrard et al. (1) that people high in self-esteem are less likely than those low in self-esteem to acknowledge the riskiness of their behavior, high self-esteem may have benefits in terms of people's beliefs about whether they can successfully enact the recommended risk-reduction behavior. Similarly, although people with low self-esteem may not counterargue their risk as vigorously or as successfully as those with high self-esteem, they may not respond as well to risk communications in terms of other beliefs, such as self-efficacy, that have been shown to be important to enacting and maintaining behavior change. Thus, risk communications to increase perceived vulnerability that do not address other health beliefs may easily backfire, making the high self-esteem respondents more defensive and the low self-esteem respondents less confident. A similar set of questions—and potential pitfalls—arises in research on optimistic beliefs. A great deal of research has focused on the possibility that optimists will underestimate their risks of negative events and, correspondingly, on the development of interventions to increase their perceived risk. However, optimism appears to be associated with increased, rather than decreased, attention to health-risk information (22) and to more constructive responses to illness and other stressful events (23), suggesting that it is beneficial on balance. The corresponding question of whether pessimism may undermine risk-reduction behavior has yet to be fully examined. In several studies reviewed by Scheier and his colleagues (23), people who are pessimistic are more likely to engage in avoidant coping, denial, and disengagement when faced with stressful circumstances. People who are pessimistic may acknowledge their risk to a greater extent (and perhaps even overestimate it, although this has yet to be determined), but they may fail to act to reduce their risk because they do not believe such responses will be successful in reducing their risk. As these examples illustrate, it may be important to consider the relation of any given individual difference to a wide range of relevant beliefs that may affect precautionary behaviors, not just perceived risk, and to develop risk communications and other interventions that address these related health beliefs. Assess Multiple Influences on Responses to Risk Communications Over Time Both target articles (1,2) listed several methodologic reasons why cross-sectional studies fail to provide convincing or useful information about the relation between risk perceptions and precautionary behavior. To these vital and frequently overlooked points, I would like to add another point that has been virtually unaddressed in the literature. Often we simply do not know what happens in the interim between exposure to a risk communication and some subsequent behavioral outcome that would explain a positive, negative, or null relation between risk perceptions and precautionary behavior. There are reasons to believe that several factors may influence such relations, including how the risk communication was processed at the time of exposure, the success or failure of subsequent efforts to enact risk-reducing behavior, the processing of subsequent risk-related information, and social influences on both responses to risk information and precautionary behavior. A comprehensive review of such factors is beyond the scope of this paper, but I will provide some examples to illustrate how these factors may influence risk perceptions, precautionary behavior, and subsequent processing of risk communications. An important initial question may be, “Where does the risk information go and why?” That is, what determines whether a risk communication is counterargued or weakened by conflicting information, as opposed to being elaborated and bolstered by additional consistent information? Interesting and potentially important differences in recall for risk-related information can be found over very short periods of time, yet little is known about how such differences are created and maintained. For example, in a 1-week follow-up to the study of women who were frequent caffeine drinkers (18), those who did not complete the affirmation reported greater numbers of risk-disconfirming facts on a surprise free recall test than those who completed the affirmation. What intervening process(es) maintained such differential recall? It is unlikely that such differences could be explained simply by a single instance of defensive processing or counterarguing at the time of exposure; instead, it is possible that these participants reminded themselves of the reasons caffeine was not harmful to their health with each cup of coffee, thus repeatedly counterarguing the risk communication in the week following exposure. Conversely, how did those who completed the affirmation differ in the thoughts they brought to mind each time a choice about consuming caffeine arose? Understanding how different aspects of the persuasive message may be brought to mind as people make risk-relevant choices will be essential to understanding how risk communications may influence beliefs and behaviors over time. A second, closely related factor is understanding how risk perceptions are related to successful or unsuccessful attempts to enact precautionary behaviors. Risk perceptions may indeed prompt initial efforts to undertake precautionary behaviors, but people may experience frustration, embarrassment, distress, or failure in their efforts. These difficulties may influence subsequent risk perceptions. Recall that the smokers studied by Gerrard et al. (1) who failed in their attempts to quit smoking were the ones who subsequently downplayed the risks of cigarette smoking. Risk perceptions and behaviors are likely to be reciprocally related as people encounter risk-relevant choices on a frequent—and even daily—basis. Experience sampling or daily diary studies might be very useful in understanding how people think about risk information following exposure, especially when that information involves daily behavior and choices, such as diet, exercise, and smoking. Identifying the stumbling blocks on the path to precautionary behavior and how such difficulties affect subsequent risk perceptions and processing of risk information is essential. Third, as these examples suggest, it is important to understand how initial responses to a risk communication may influence the processing of subsequent risk information. Does initial disagreement with a risk communication make it more difficult for subsequent messages to be processed veridically? Are such effects compounded with repeated exposure to risk communications? Such phenomena may be especially amenable to analysis in terms of theories of responses to persuasion, in that counterarguing makes disagreeing thoughts more accessible and therefore more likely to be evoked by subsequent risk communications. Related topics for future study might include understanding other aspects of how people manage multiple risk messages. People are exposed to repeated messages about diet, exercise, screening, and other health-related topics, not just those from a planned intervention. It will be important to understand how prior message processing affects responses to this new information, as well as people's strategies for dealing with an ever-increasing volume of information about health risks. Do people become discouraged as they realize that they cannot simultaneously address dozens of potential threats to health? Do people choose to focus on one or two risks and ignore the rest? Studies of people's strategies for dealing with such information on a daily basis would likely give insights into such processes. In this vein, the recent NCI focus group report (14) on how the public perceives, processes, and interprets risk information is a valuable start. Understanding how these processes may differ as a function of risk status may also reveal some important information about the relation of risk perceptions to multiple message processing over time. A final factor that is relatively understudied in understanding responses to risk communications is how the social interactions of people exposed to risk communications bolster or degrade the effects of the message. For cancer and many other illnesses, the beliefs and behaviors of other people have been shown to be highly influential in how people understand their own risk and whether they enact precautionary behavior (24,25). It will be critical to examine how specific risk communications may be supported or discounted by similar others' views and experiences. Such responses may be especially important in understanding responses to risk communications among people at high risk for cancer, as the desire for information from and about other people has been shown to increase sharply as a function of distress or uncertainty. A daily diary study could answer some of these questions by tracking how people share risk-related information with others following exposure to a risk communication and are influenced by their reactions to the information and its recommendations. Such a study would provide valuable information about how social factors, for better or for worse, influence subsequent emotions, thoughts, and behaviors related to the risk. One promising approach to creating social support for risk perceptions and precautionary behaviors is to conduct risk-communication interventions in small peer groups (16,26). In this way, people see that others like them are similarly concerned about their risk. Such approaches may increase support for beliefs and behaviors related to risk reduction, while at the same time decreasing self-presentational concerns that may interfere with precautionary behaviors. Conclusion The points raised in the target articles and in this brief discussion of responses to persuasive messages suggest that there are many reasons that people deviate from the ideal sequence in which exposure to risk communications leads to perceived risk and, ultimately, to precautionary behavior. First, there may be a complex series of trade-offs that influence people's responses to risk communications. Factors that increase people's motivation to process a message may interfere with their ability to do so; similarly, factors that increase one's sense of personal vulnerability may also create feelings of pessimism and hopelessness that interfere with risk-reduction behavior. As Rothman and Kiviniemi (27) point out, people who acknowledge increased personal vulnerability may also simultaneously change other perceptions of the health threat, such as its seriousness or prevalence. It is unknown how these different factors work together to influence precautionary behavior and subsequent processing of risk-related information. The recognition of this complex—and often conflicting—set of responses to risk communications suggests several different approaches for future research. One approach would be to subject major risk communication campaigns to careful pilot studies of responses to risk communications by use of multiple measures of information processing, cognitive elaboration, and related health beliefs to understand what people attend to, what thoughts are generated by the message (agreeing? disagreeing?), how the information is evaluated (convincing? trustworthy?), what people recall following exposure, and how and whether people believe they can modify their risk. It will be essential to conduct these tests in high- and low-risk groups and among people with different risk factors for cancer. Second, persuasive communications to increase perceived risk of cancer must be sensitive to several factors that, unless handled carefully, could result in extensive counterarguing and, therefore, increased resistance to change. Special care should be taken to present information about risk behaviors that may be self-defining or that have important self-presentational or other interpersonal functions. It is difficult to alter such behaviors without making the recommended risk-reduction behaviors more socially acceptable or normative and without addressing reasons other than health-related ones that people practice a particular health-risk behavior. A third pressing need is to understand how people manage risk-related information over time and what happens as they try to incorporate this information and the behavioral changes it suggests into their daily lives. Where does the risk information “go,” and why do risk perceptions bear an inconsistent—and sometimes counterproductive—relation to precautionary behaviors? These factors may be different for high- and low-risk groups in ways that may inform the design and delivery of risk communications and of other interventions to maintain support for precautionary behaviors. Additionally, individual differences, such as self-esteem and optimism, may play an important role in such responses, but they may have complex relations to subsequent risk perceptions and behavior. Only prospective studies that examine a range of outcome measures can tell us how different responses to risk communication may be related to risk beliefs and precautionary behaviors over time. Such studies are essential in that the target behaviors, such as changes in diet and exercise and adherence to screening recommendations, must be practiced consistently over time to be effective. Daily diary studies may provide valuable information about how people manage risk information in their social interactions and behavioral choices and how such processes affect risk perceptions and precautionary behaviors over time. Finally, it may be well worthwhile to consider one of the major conclusions of Vernon's (2) review. Her suggestion to focus directly on social and structural factors that may influence cancer screening, such as cultural beliefs, socioeconomic status, insurance coverage, and physician recommendations, is an important reminder that precautionary behaviors are not always purely personal or individual but may also depend on a broader context of social and structural support for such behaviors. I thank Meg Gerrard for her helpful comments on an earlier version of this manuscript. References (1) Gerrard M, Gibbons FX, Reis-Bergan M. The effect of risk communication on risk perceptions: The significance of individual differences. Monogr Natl Cancer Inst  1999; 25: 94-100. Google Scholar (2) Vernon SW. Risk perception and risk communication for cancer screening behaviors: a review. Monogr Natl Cancer Inst  1999; 25: 101-19. Google Scholar (3) Gerrard M, Gibbons FX, Bushman BJ. Relation between perceived vulnerability to HIV and precautionary sexual behavior. Psychol Bull  1996; 119: 390-409. Google Scholar (4) Weinstein ND. The precaution adoption process. Health Psychol  1988; 7: 355-86. Google Scholar (5) Eagly AH, Chaiken S. The psychology of attitudes. Fort Worth (TX): Harcourt, Brace Jovanovich; 1993. Google Scholar (6) Petty RE, Cacioppo JT. Communication and persuasion: central and peripheral routes to attitude change. New York (NY): Springer-Verlag; 1986. Google Scholar (7) Petty RE, Wegener DT. Attitude change: multiple roles for persuasion variables. In: Gilbert DT, Fiske ST, Lindzey G, editors. Handbook of social psychology. 4th ed. Boston (MA): McGraw-Hill; 1998. p. 323-90. Google Scholar (8) Jepson C, Chaiken S. Chronic issue-specific fear inhibits systematic processing of persuasive communications. J Soc Beh Pers  1990; 5: 61-84. Google Scholar (9) Leventhal H. Findings and theory in the study of fear communications. Adv Exp Soc Psychol  1970; 5: 119-86. Google Scholar (10) McAlister A, Perry C, Killen J, Slinkard LA, Maccoby N. Pilot study of smoking, alcohol and drug abuse prevention. Am J Public Health  1980; 70: 719-21. Google Scholar (11) Ditto PH, Lopez DF. Motivated skepticism: use of differential decision criteria for preferred and nonpreferred conclusions. J Pers Soc Psychol  1992; 63: 568-84. Google Scholar (12) Ditto PH, Scepansky JA, Munro GD, Apanovitch AM, Lockhart LK. Motivated sensitivity to preference-inconsistent information. J Pers Soc Psychol  1998; 75: 53-69. Google Scholar (13) Liberman A, Chaiken S. Defensive processing of personally relevant health messages. Pers Soc Psychol Bull  1992; 18: 669-79. Google Scholar (14) National Cancer Institute, Office of Cancer Communications. How the public perceives, processes, and interprets risk information: Findings from focus group research with the general public. Bethesda (MD): Natl Cancer Inst; 1998 Jun Report No.: POS-T086. Google Scholar (15) Fisher JD, Fisher WA. Changing AIDS-risk behavior. Psychol Bull  1992; 111: 455-74. Google Scholar (16) Bryan AD, Aiken LS, West SG. Increasing condom use: evaluation of a theory-based intervention to prevent sexually transmitted diseases in young women. Health Psychol  1996; 15: 371-82. Google Scholar (17) Leary MR, Tchividjian LR, Kraxberger BE. Self-presentation can be hazardous to your health: impression management and health risk. Health Psychol  1994; 13: 461-70. Google Scholar (18) Reed MB, Aspinwall LG. Self-affirmation reduces biased processing of health-risk information. Motivation Emotion  1998; 22: 99-132. Google Scholar (19) Aspinwall LG. Rethinking the role of positive affect in self-regulation. Motivation Emotion  1998; 22: 1-32. Google Scholar (20) Taylor SE. Adjustment to threatening events: a theory of cognitive adaptation. Am Psychol  1983; 38: 1163-73. Google Scholar (21) Janoff-Bulman R. Assumptive worlds and the stress of traumatic events: applications of the schema construct. Soc Cogn  1989; 7: 113-36. Google Scholar (22) Aspinwall LG, Brunhart SM. Distinguishing optimism from denial: optimistic beliefs predict attention to health threats. Pers Soc Psychol Bull  1996; 22: 993-1003. Google Scholar (23) Scheier MF, Carver CS, Bridges MW. Optimism, pessimism, and psychological well-being. In: Chang EC, editor. Optimism and pessimism: theory, research, and practice. Washington (DC): American Psychological Association. In press 1999. Google Scholar (24) Buunk BP, Gibbons FX, editors. Health, coping, and well-being: perspectives from social comparison theory. Mahwah (NJ): Erlbaum; 1997. Google Scholar (25) Croyle RT. Appraisal of health threats: cognition, motivation, and social comparison. Cogn Ther Res  1992; 16: 165-82. Google Scholar (26) Fisher JD, Fisher WA, Misovich SJ, Kimble DL, Malloy TE. Changing AIDS risk behavior: effects of an intervention emphasizing AIDS risk reduction information, motivation, and behavioral skills in a college population. Health Psychol  1996; 15: 114-23. Google Scholar (27) Rothman AJ, Kiviniemi MT. “Treating people with information”: An analysis and review of approaches to communicating health risk information. Monogr Natl Cancer Inst  1999; 25: 44-51. Google Scholar Oxford University Press

Journal

JNCI Monographs – Oxford University Press

Published: Jan 1, 1999

There are no references for this article.