Author's note: Kim Witte (Ph.D., University of California) is professor in the Department of Communication at Michigan State University. Discussions with Thom Jayne and comments from Scott Ratzan improved this manuscript substantially

Abstract

Researchers and academics have accumulated a great deal of knowledge on how to manipulate individuals' behaviors. While health communication researchers and practitioners prefer to call their work public health campaigns or health education interventions, the truth is that our ultimate goal is to manipulate people into practicing healthy behaviors. As health communication research and practice moves into the 21st century, we must face this issue squarely and directly and develop strategies for the ethical use of manipulation techniques to promote health and prevent disease.

The Manipulative Nature of Health Communication Research:

Ethical Issues and Guidelines

The one fact I would cry from every housetop is this: the Good Life is waiting for us--here and now! At this very moment we have the necessary techniques, both material and psychological, to create a full and satisfying life for everyone.

- BF Skinner, 1948

Researchers and academics have accumulated a great deal of knowledge on how to manipulate individuals' behaviors (Bandura, 1977; Graeff, Elder, & Booth, 1993; Janz & Becker, 1984; Prochaska, 1979). Manipulation is not a popular word, yet it is really a major part of what health promotion and disease prevention is all about. Broadly defined, to manipulate or persuade means to influence people into doing what we want them to do through direct (i.e., obviously persuasive, as through advertisements or PSAs) or indirect (i.e., not obviously persuasive, as through "information" or "awareness" campaigns) strategies (see Cialdini, 1988; Miller & Steinberg, 1975; Pratkanis & Aronson, 1991). As Brock, Shavitt, and Brannon (1994) note, "No one escapes...the constant reconfiguring of our beliefs, attitudes, intentions, and behavior by unrelenting and ubiquitous forces" (p. 1). They continue, "Persuasion is constantly remaking us into persons who are measurably changed. Sometimes imperceptibly--ofttimes dramatically" (Brock, Shavitt, & Brannon, 1994, p. 1).

While health communication researchers and practitioners prefer to call their work "health education interventions" or "public health campaigns," the truth is that their ultimate goal is to manipulate people into practicing healthy behaviors. For example, doctor-patient researchers study how to make patients feel satisfied with medical encounters (i.e., manipulate patients into feeling a certain way about a medical interaction) or how to make patients comply or adhere to medical advice (i.e., manipulate patients into doing what physicians want them to do) (Barnlund, 1993; Thompson, 1990). Public health communication researchers study how to develop campaigns or interventions that produce the greatest amount of behavior change in the desired direction (Rice & Atkin, 1989). Even in cases where communicators simply try to "inform" they are still influencing and manipulating perceptions and behaviors -- albeit inadvertently. Salmon (1989) has noted, any health campaign -- informational, factual, or otherwise -- is inherently persuasive in that it frames messages in a certain way which result in audience perceptions being molded (or manipulated) in a certain manner. Let us acknowledge the manipulative nature of our profession and use available theories to guide the development of messages in order to best promote health and prevent disease. As health communication research and practice moves into the 21st century, we must face the fact that much of health communication research and practice is manipulative in nature and develop strategies for the ethical use of manipulation techniques to promote health and prevent disease.

Shaping Perceptions, Beliefs, and Behaviors

There are three key ways in which our perceptions, beliefs, and behaviors are shaped and manipulated by verbal and nonverbal messages. First, the amount of information given about a topic influences our perceptions. It is impossible to fully inform individuals about a health risk because new research is constantly being released throughout the world. Most often, time and space constraints limit the amount of information given to patients or audiences in any medical encounter or public health message. Because of this, some facts or pieces of information are presented to the exclusion of others. Generally, physicians see patients for only brief stretches of time (Ray, 1993). Similarly, the public health practitioner typically has only 30-60 seconds to convey a health risk and recommended response in a public service announcement (Rice & Atkin, 1989).

Because the amount of information presented will influence or manipulate patients or audiences's perceptions and behaviors, health communicators need to give some thought as to what information should be presented and what should not. Again, it is important to note that even if health communicators do not intend to influence or manipulate perceptions, they do so by virtue of presenting some facts and excluding others. The key ethical question here is, which pieces of "information" should be chosen to positively influence health?

Second, the framing of messages becomes a crucial issue to consider given the fact that only a limited amount of information is presented to individuals. In other words, words and messages must be consciously and strategically chosen by campaigners and physicians in order to produce intended outcomes (i.e., healthy behaviors, compliance, etc.). Health communicators who do not consciously or strategically consider their word choices may inadvertently produce harmful outcomes. By using one word over another we (perhaps unintentionally) emphasise one aspect of a situation or issue over another. For example, we can tell patients they have a good chance of dying from cancer, or a good chance of surviving cancer. Similarly, we can describe condoms as working most of the time to prevent HIV transmission, or failing some of the time. The research evidence strongly supports the position that framing of messages has a profound influence on perceptions and decisions. For example, McNeil, Pauker, Sox, and Tversky (1982) found that patients' treatment preferences shifted given differences in

1. whether the risks were presented as chances of survival or chances of dying,
2. whether the specific treatments were identified or not (e.g., surgery or radiation therapy, or Treatment A or Treatment B), or
3. whether the data were presented in terms of life-expectancy figures or cumulative-probability figures.

McNeil, Weichselbaum, and Pauker (1978) reported that operable lung cancer patients preferred treatments that assured survival now, as opposed to prolonging survival later. Similarly, McNeil, Pauker, and Tversky (1988) found that the framing of risks influenced whether subjects chose radiation therapy or surgery. When risks for radiation therapy were framed as mortalities (e.g., 0 out of 100 people will die during radiation therapy, or 78 will die by 5 years), 47% of the subjects chose radiation therapy over surgery. However, only 18% of the sample chose radiation over surgery when risks were stated in survival terms (e.g., 100 out of 100 people will survive radiation therapy, or 22 will be alive in 5 years) (McNeil, Pauker, & Tversky, 1988). This line of research shows that how one presents a risk or recommended response strongly influences what individuals decide to do about that health threat.

Health messages are especially influential in terms of their impact on individuals' decisions and behaviors when new information is disseminated to people. Specifically, when people do not hold strong prior beliefs about a health risk or a recommended response "they are at the mercy of the way that the information is presented. Subtle changes in the way that risks are expressed can have a major impact on perceptions and decisions" (Slovic, Fischoff, & Lichtenstein, 1984, p. 184). In sum, the evidence is clear that how health communicators frame their health-risk messages manipulates individuals' perceptions and ultimately their decisions and behaviors.

Third, the order in which information is given to individuals also influences their perceptions and behaviors. In some cases, people will remember what is presented at the beginning of a message and forget what is presented at the end. In other cases the reverse is true. Research has shown that when a health topic is important and/or relevant to individuals, a primacy effect typically operates (Bettinghaus & Cody, 1994; Bostrom, 1983). That is, people will remember information given at the beginning of a presentation and forget information given at the end of a message. If the health topic is perceived to be irrelevant or not very important, then a recency effect operates (Bettinghaus, 1980; Bostrom, 1983). With recency effects, information given at the end of a presentation is remembered better than information given at the beginning of a presentation.

The order of information presented in a message also is critical when the information is particularly frightening or emotionally arousing. When bad or scary news about a health threat is given, many people stop processing the message any further. For example, cancer patients often tell the story that when their doctor first told them of their diagnosis, they didn't remember anything past being told "you have cancer." Their emotional arousal was so high due to their perceived "fatal" diagnosis that they ceased to cognitively process the treatment information their doctors gave them. Much research has shown that when people are overwhelmed with frightening information, they simply "shut down" and stop listening to and/or processing messages (Witte, 1992).

In sum, ethical guidelines are needed for both health communicators who focus on behavior change strategies (intended manipulation of behavior) as well as for those who intend to merely "inform" their clients or audiences. The research evidence suggests that all messages (either intentionally or unintentionally) manipulate an audience's perceptions because in every health communication (a) only a limited amount of information is ever presented, (b) the information presented is always framed in a certain manner, and (c) the information given is always ordered in a particular way (Pratkanis & Aronson, 1991).

The Ethics of Choosing

Which Perceptions, Beliefs, and Behaviors to Shape

As B.F. Skinner said in the opening quotation, we have the necessary techniques to create the "good life," but who decides what the good life or healthy life is? This is the key ethical question facing health communication researchers today. How do we go about choosing which perceptions, beliefs, and behaviors to shape?

Some have said that one cannot not communicate (Watzlawick, Beavin, & Jackson, 1967). While the degree of truth in this statement has been argued vigorously (e.g., Motley, 1990), I believe the available scientific evidence demonstrates that one cannot not manipulate when communicating about health and disease. By virtue of framing messages in a certain manner, in a certain order, with a certain amount of information, communicators manipulate receivers into certain mindsets -- sometimes intentionally and sometimes unintentionally. Public service announcements attempt to influence audiences into thinking that skin cancer is a significant and likely event for many people, physicians try to convince their patients that they must take their medication or stop smoking. Because one cannot not manipulate when communicating about health, health communicators have a tremendous ethical responsibility to first determine what appropriate health messages are and then to craft their messages to promote health and prevent disease. Following are some preliminary ideas for carrying out this task.

1. Health communicators must decide in advance what their health-related goal is, and then construct their messages to fit that goal. Of the many messages people receive daily, we do not know which ones will be attended to and which ones will not. Therefore, it is important to craft each message as if it will be attended to, comprehended, and acted upon. If practitioners do not consider in advance what their health-related goals are, they run the risk of composing a message that results in more harm than good. For example, in response to some media reports in 1993 about a Swedish study that found mammograms may cause an increase in cancer (due to the radiation in mammography), a mammogram scare was started in the United States that resulted in many women shunning life-saving mammograms. What the study really found was that the risk of radiation from mammography may be greater for women under age 40 than the discovery of breast cancer. After age 40, mammograms were clearly life-saving.
   
   A personal example also illustrates the need to decide in advance health-related message goals. Currently, I am a member of the "Radon Awareness Advisory Committee" for the state of Michigan. The committee is composed of a group of experts -- physicians, epidemiologists, natural scientists, environmental specialists. I am the only communication expert. Some of the research evidence about radon is downright scary and maximises how harmful it is while other research minimises its significance.
   
   The "hard" scientists (i.e., those who study physiology, pollutants, etc.) argue that we should simply release all available information about radon to the community and then let individuals decide what to do about the risk. The problem with this approach is that it assumes people are rational decision makers and that they will sift through all available information much like computers before making decisions. However, research shows that people typically behave in non-rational ways and may be overly influenced by some types of evidence (i.e., anecdotal vs. statistical) over others (Slovic, 1987). Although unpopular, I took the stance that it is our ethical responsibility to (a) determine whether or not we think radon is a significant health risk warranting action (we were the "experts"), (b) develop clear recommendations (i.e., goals) to reduce any harmful health-related effects for the residents of Michigan if needed, and (c) craft messages that promote these goals (e.g., residents in some areas of Michigan should have their basements tested). I pointed out that if we simply disseminated all available information then we ran the risk of having our messages produce unintended harmful outcomes due to message framing, amount of information released, and order effects. For example, if we held a press conference publicising "information" about radon, it is likely that the media would pick-up only the most sensational, threatening, scary information, and only partially report the "complete" set of facts due to their time and/or space constraints.[1] Messages that emphasise only one aspect of the radon risk over another are a disservice.
   
   Overall, a message will produce one of four outcomes -- no response, a positive response, a negative response, or a mixed response (partly positive and partly negative). In the radon example above, messages can produce no response to the risk of radon, a positive response (people have their homes tested, precautions are taken), a negative response (people become frightened, property values plummet, anxiety increases), or a mixed response (people test their homes but are very anxious about the situation). Because we have the theories and technology needed to craft messages that produce positive outcomes, we should use them. If messages are crafted arbitrarily, then one's chances for inducing negative outcomes are greatly increased. To avoid a "big brother" mentality, however, messages must be crafted to promote the common good as determined by a community standard.
2. Health communications should promote the "common good." The ethical standard that prevails throughout the world is that of "the greatest good for the greatest number of people," formally called utilitarianism. The Bible, the Koran, Vedic writings, and the Torah all support the stance that if a choice is to be made between competing attractive alternatives, the alternative benefitting the greatest number of people should be chosen. I believe the same standard should apply to health communication messages. Messages that promote the greatest good for the greatest number of people should be crafted and disseminated.
   
   The common good can be determined by assessing the total number of persons affected by a certain act in terms of intensity, duration, and propinquity (Miller, 1987, p. 530). In provider-patient encounters, the Hippocratic oath demands that the patient's needs and interests come first.
   
   Similarly, the World Health Organization's charter advocates promoting community and social health and well-being, as does the new definition of health communication offered by the Centers for Disease Control and Prevention (Roper, 1993). Most health communicators define the "common good" as that which promotes health and well-being for the individual, the family, and the community. Commensurate with the Hippocratic oath, WHO, and the CDC's guidelines, the patient or community's rights and needs must come first. And, when competing rights and needs conflict, a "common good" standard should be used to promote the health and well-being of the greatest number of people.
3. A community standard should be used to determine the common good. Salmon (1989) raises a good point when he questions whether the "common good" exists. He asks, who determines the common good and which common good for which group of people is most important?
   
   Traditionally, the power to define the common good and make decisions regarding important issues has resided "disproportionately with government, corporations, and other institutions possessing legitimacy, social power, and resources and access to the mass media" (Salmon, 1989, p. 25). However, decisions made by the government and corporations may serve the "good of the agency" over the "good of the people."
   
   To truly address each community's health-related concerns and priorities, community representatives (representing different interest groups) should decide what the common good is for that community. While safety and well-being are identified by Salmon (1989) as universal humanitarian values, each community is likely to define safety, health, and well-being in a different manner. In some communities, the prevention of infectious diseases takes precedence over cancer or heart disease, in others environmental concerns outweigh substance abuse treatment.
   
   In short, a community panel with either elected or appointed individuals representative of each interest group is recommended to determine what health-related messages should be disseminated and how those messages should be presented (i.e., framing, content, order). The community panel should be a free-standing panel that acts in an advisory capacity to local departments of public health. Ideally, the panel should be composed of people representing the demographic and psychographic profile of each community (e.g., 10-12 people). Specifically, this community panel should be charged with the following tasks:
   1. To identify health needs in conjunction with experts (what health threats are significant enough to warrant attention?).
   2. Identify appropriate recommended responses in conjunction with experts (which recommended responses are effective in averting the health threat?).
   3. Prioritise the health needs (which is the most important health threat that must be addressed immediately? which is the next? and so on...).
   4. Prioritise the recommended responses (which recommended responses will work to avert the health threat? which are most acceptable with the community's psycho-social-religious composition?).
   5. Develop a list of recommendations and goals (e.g., goal #1 -- awareness of breast cancer as a deadly but preventable disease, goal #2 -- increase by 50% the number of mammograms obtained by low-income women).
   6. Have communication experts develop messages, based on available theories, that promote the recommendations and goals.
   7. Develop campaign strategy that reaches the target population.

The guidelines outlined here assume that each member of a community panel will adopt the "common good" ethic and put their personal biases and beliefs aside in order to promote the health and well-being of their community. A good example of this is former Surgeon General Koop, a conservative man against pre-marital sexual intercourse. When faced with the reality and extent of AIDS infection across the United States, he put aside his personal beliefs and promoted the common good by advocating the use of condoms to prevent HIV transmission. Community representatives must be willing to face political and social pressures from competing interest groups in order to promote the common good of the entire community. However, because health problems or issues often cluster within certain communities (e.g., environmental concerns in Brownsville, Texas; AIDS/HIV prevention in San Francisco), the panel is likely to receive support from a community that believes its unique health problems are being addressed. The local public health department would take recommendations from the panel in order to best promote the health and needs of the local community.

Summary

It is time for health communicators to acknowledge the manipulative nature of much of their work and develop ethical guidelines within which to operate. Health communicators should consider the impact of their messages before presenting them, adopt a "common good" credo, and let their theoretically-based messages be guided by a community standard.

Conclusion

According to much empirical evidence, one cannot not manipulate when communicating about health. Thus, although the term manipulation holds a negative connotation, this essay argues that much of what health communicators do is to manipulate behavior in order to promote health and well-being. It is time health communicators develop ethical guidelines for manipulating -- or as some prefer, influencing -- individuals' health-related behaviors through their messages. The present work offers an initial starting point for this endeavor.

References

Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191-215.

Barnlund, D.C. (1993). The mystification of meaning: Doctor-patient encounters. In B.C. Thornton & G.L. Kreps (Eds.), Perspectives on health communication (pp. 30-41). Prospect Heights, IL: Waveland Press.

Bettinghaus, E., & Cody, M.J. (1994). Persuasive communication (5th edition). New York: Harcourt Brace.

Bostrom, R.N. (1983). Persuasion. Englewood Cliffs, NJ: Prentice-Hall.

Cialdini, R.B. (1988). Influence: Science and practice (2nd ed.). New York: Harper Collins.

Graeff, J.A., Elder, J.P., & Booth, E.M. (1993). Communication for health and behavior change: A developing country perspective. San Francisco: Jossey-Bass Publishers.

Janz, N., & Becker, M. (1984). The health belief model: A decade later. Health Education Quarterly, 11, 1-47.

McNeil, B.J., Pauker, S.G., Sox, H.C., & Tversky, A. (1982). On the elicitation of preferences for alternative therapies. The New England Journal of Medicine, 306, 1259-1262.

McNeil, B.J., Pauker, S.G., & Tversky, A. (1988). On the framing of medical decisions. In Bell, D.E., Raiffa, H., & Tversky, A. (Eds.), Decision Making: Descriptive, Normative, and Prescriptive Interactions (pp. 562-568). Cambridge: Cambridge University Press.

McNeil, B.J., Weichselbaum, R., & Pauker, S.G. (1978). Fallacy of the five-year survival in lung cancer. The New England Journal of Medicine, 299, 1397-1401.

Miller, D. (1987). The Blackwell Encyclopaedia of Political Thought. Oxford: Blackwell Reference.

Miller, G.R., & Steinberg, M. (1975). Between people: A new analysis of interpersonal communication. Chicago: Science Research Associates.

Motley, M.T. (1990). On whether one can(not) not communicate: An examination via traditional communication postulates. Western Journal of Speech Communication, 54, 1-20.

Pratkanis, A., & Aronson, E. (1991). Age of propaganda: The everyday use and abuse of persuasion. New York: Freeman.

Prochaska, J. (1979). Systems of psychotherapy: A transtheoretical analysis. Belmont, CA: Dorsey Press.

Ray, E.B. (1993). Case studies in health communication. Hillsdale, NJ: Lawrence Erlbaum.

Rice, R.E., & Atkin, C.K. (1989). Public Communication Campaigns (2nd ed.). Newbury Park, CA: Sage.

Roper, W.L. (1993). Health communication takes on new dimensions at CDC. Public Health Reports, 108, 179-183.

Salmon, C.T. (1989). Information Campaigns: Balancing Social Values and Social Change. Newbury Park, CA: Sage.

Skinner, B.F. (1948). In J. Bartlett (Ed.), Bartlett's Familiar Quotations (p. 862). Boston, MA: Little, Brown & Co.

Slovic, P. 1987. Perception of risk. Science, 236, 280-285.

Slovic, P., Fischoff, B., & Lichtenstein, S. (1984). Behavioral decision theory perspectives on risk and safety. Acta Psychologica, 56, 183-203.

Thompson, T.L. (1990). Patient health care: Issues in interpersonal communication. In E.B. Ray & L. Donohew (Eds.), Communication and Health: Systems and Applications (pp. 27-50). Hillsdale, NJ: Lawrence Erlbaum.

Watzlawick, P., Beavin, J., & Jackson, D.D. (1967). Pragmatics of human communication: A study of interaction patterns, pathologies, and paradoxes. New York: Norton.

Witte, K. (1992). Putting the fear back into fear appeals: The Extended Parallel Process Model. Communication Monographs, 59, 329-349.

1 Other constraints on message dissemination include cost and public sophistication and/or knowledge on a topic. Some topics may be so technical as to be confusing to laypersons and journalists who pick out only those pieces of information they understand. It is for these technical topics that the greatest misunderstanding can come about.

Revised August 6, 2002.