Risk and Risk Perception: Definitions and Dimensions
Risks are pervasive issues both within and across national borders. Noteworthy examples include natural disasters such as hurricanes and earthquakes, human-made disasters such as radiation exposure, and recent instances of global infectious diseases such as Ebola, Middle East Respiratory Syndrome (MERS), and the Zika virus. The concept of risk refers to the probability of experiencing harm or hazards. Hazards refers to threats to people and the things they value. Probability refers to the likelihood of a harm’s or hazard’s occurrence, which will tend to be perceived with some degree of uncertainty. The uncertain aspect of risk is related to people’s disagreements about a given risk’s magnitude and severity. People experience uncertainty when a situation is ambiguous, unpredictable, or probabilistic. Interpretations and other subjective judgments about risks are known as risk perceptions (Slovic, 2000). Risk perceptions are important determinants of health- and risk-related decisions such as adopting healthy behaviors, curtailing unhealthy behaviors, and accepting or rejecting a certain level of risks (e.g., policy decisions on nuclear plants, GMO foods, processed meats).
A common assumption in risk perception research is that people’s knowledge and certainty about a risk determines how they will perceive it. This assumption is based on the rational choice model of decision making, which portrays people as evaluating the possibility of outcomes after they calculate potential costs and benefits. This way of evaluating risks is predominantly ascribed to experts, who are assumed to rely on scientific information and objective assessment. By contrast, laypeople are commonly assumed to evaluate risks by using heuristics and other informal thought processes. For example, when people are more aware of certain risks, they tend to believe that those risks happen more frequently than they actually do. This tendency is known as the availability heuristic (Kahneman, Slovic, & Tversky, 1982). Consider some examples from health contexts: If you have any friends or family members who died of colon cancer, you are more likely than other people to perceive that disease as posing a fatal risk. Also, people who have been heavily exposed to media coverage of an infectious disease such as the H1N1 flu may perceive it to be more prevalent and risky than those who have not.
Other ways of misperceiving the frequency and magnitude of risks can occur due to individual characteristics. A notable one is optimistic bias or unrealistic optimism, the tendency to believe that risks pose a less serious threat to oneself than they do to other people (Weinstein, 1980). For example, smokers who have a strong optimistic bias are likely to believe that smoking may be hazardous to other people’s health but not to their own. Heuristics such as these, as well as other individual tendencies, make people perceive risks in different ways. In addition, because laypeople often don’t have access to detailed information about risks, they tend to perceive them more in conjunction with emotions such as dread and fear. This tendency can lead people to overestimate hazards’ actual frequency and severity.
When risk perception was initially explored, researchers focused on people’s cognitive judgments about the magnitude and likelihood of risks. Eventually, however, they acknowledged the important role that emotions such as dread, fear, and outrage play in evaluating risks. Slovic and his associates called attention to the affect heuristic, which in the context of risk perception refers to people’s tendency to rely on their current emotions when they make judgments about risks. If we feel intense dread when we perceive a risk, we are likely to evaluate it as more threatening and more prevalent. Similarly, the risk-as-feelings hypothesis predicts that emotional reactions to risks are often independent of cognitive assessments of them, and that they are stronger determinants of people’s behavior (Loewenstein, Weber, Hsee, & Welch, 2001).
As these emotional aspects of risk perception were being explored in the field of psychology, a similar view of emotions’ role in risk perception was developed separately in the field of risk communication. Risk communication refers to the process of informing and persuading the public about risks so that they will be able to perceive them accurately and make appropriate decisions about them (Walaski, 2011). As the field of risk communication matured, researchers discovered that laypeople often do not understand risks in the same ways experts do. Laypeople’s responses to risks are determined more by their subjective perceptions, and they tend to have less knowledge about objective risk factors. This was yet another discovery that affirmed the importance of the emotional dimension of risks and risk characteristics. One researcher who paid close attention to how laypeople perceive and respond to risks was Peter Sandman.
Sandman (1989) defines risk as a combination of “hazards and outrage.” Hazards constitute the technical component of risk, and outrage is a nontechnical component that refers to an amalgam of voluntariness, control, responsiveness, trust, dread, and other non-rational responses (Walaski, 2011). Although Sandman’s concept of outrage is idiosyncratic and needs further clarification, it is notable for including emotional components such as dread and fear. In his efforts to understand the roles of media (particularly journalism) in people’s risk perceptions and subsequent behaviors, Sandman particularly highlights the emotional and sensational nature of media coverage, and he suggests risk communication strategies according to different levels of hazards and outrage. (The role of media in risk perceptions is discussed in more detail below.)
Risk Characteristics and Relevant Models
One theoretical framework that incorporates both the cognitive and emotional dimensions of risk perceptions is the psychometric paradigm developed by Slovic and his associates (Slovic, 2000). According to the psychometric paradigm, people judge the riskiness of a hazard based on the combination of a range of (perceived) risk characteristics, which include the following: the severity of the risk is not controllable; the risk makes people feel dread; the risk could be globally catastrophic; the risk is certain to be fatal; people will experience the risk in unequal ways; many people are exposed to the risk; the risk could threaten future generations; the risk is increasing; exposure to the risk is involuntary; the risk affects us personally; the risk is not observable; people do not know whether they are exposed to the risk; the risk’s effects are immediate; the risk is new and unfamiliar; the risk is unknown to science (Slovic, Fischhoff, & Lichtenstein, 2000). The psychometric paradigm classifies this range of risk characteristics under two broad labels, dread risk and unknown risk. Dread risk includes “perceived lack of control, dread, catastrophic potential, fatal consequences and the inequitable distribution of risks and benefits” (Slovic, 2000, p. 225). Unknown risk includes “hazards judged to be unobservable, unknown, new, and delayed in their manifestation of harm” (Slovic, 2000, p. 226).
Critics of the psychometric paradigm claim that these labels are ambiguous. Some have proposed that dread and unknown risk should instead be viewed as two dimensions of risk judgments, cognitive and emotional (Coleman, 1993; Dunwoody & Neuwirth, 1991). But despite such criticisms, the psychometric paradigm has increased our understanding of the complex psychology behind people’s risk perceptions. It has also helped explain why certain risk issues (e.g., radiation from nuclear plants) are perceived to be more serious than others—even when they in fact are not (Paek, 2014).
Using several of these risk characteristics in the context of risk communication, Covello proposed four theoretical models that explain how people perceive risks, how they process risk information, and how they make decisions accordingly. First, the risk perception model identifies a wide variety of factors that influence people’s risk perceptions. They include voluntariness, controllability, familiarity, equity, benefits, understanding, uncertainty, dread, trust in institutions, reversibility, personal stake, ethical/moral nature, human versus natural origin, and catastrophic potential. For example, if people perceive a risk to be voluntarily incurred, they are more likely to accept it because they understand their role in experiencing the implications of the risk. By contrast, if people have less intense and less fearful emotions toward a risk, they are more likely to accept it. These factors have been used to inform risk and crisis communication strategies (Walaski, 2011). Second, the mental noise model posits that events producing a higher level of mental noise (or stress) reduce people’s ability to process risk-related information. Factors that cause a high level of mental noise include controllability, voluntariness, familiarity, cause of the disaster (human-made versus natural), dread, uncertainty, and the victim’s vulnerability (e.g., child, pregnant woman). These factors closely resemble those identified in the risk perception model. Third, the negative dominance model predicts that situations producing risks and subsequent emotions such as fear, dread, and anxiety create an environment where people are more likely to focus on negative messages. Fourth, the trust determination model highlights the importance of perceived trust of the communicator in people’s perceptions of and reactions to given risks. It highlights several trust determination factors that help build the communicator’s trust, such as caring and empathy, competence and expertise, honesty and openness. Of these four models, the risk perception model has been used most widely.
The psychometric paradigm and Covello’s four models focus on how individuals’ psychological characteristics affect risk perceptions. Other approaches have indicated a variety of cultural and social influences on people’s perceptions and responses to risks. For example, the social amplification of risk framework (SARF) attempts to show the relations among the technical analysis of risk and the cultural, social, and individual response structures that shape people’s experience of risk (Kasperson et al., 1988). SARF presumes that risk events interact with psychological, social, and cultural processes in ways that can heighten or attenuate public perceptions of risk and related risk behaviors. An important feature of SARF is that it highlights the roles played by communication channels in risk amplification and attenuation. One channel is informal and interpersonal communication networks. Friends, family, and co-workers may amplify or attenuate risk perceptions by giving one another information or reinforcing habitual perceptions and cultural biases. The other channel is the news media, which can determine which risks receive public attention. The media tend to pay more attention to (and thereby amplify) unusual or dramatically striking risks, and they pay less attention to well-known or dramatically uninteresting risks, even though such risks may continue to be severe.
Taken together, the psychometric paradigm, Covello’s four theoretical models, and SARF highlight how people’s risk perceptions are determined by various risk characteristics and factors of individual psychology, societal institutions, and communication channels.
Theoretical Perspectives in Risk Perceptions
Risk perceptions are important precursors to behaviors that experts recommend for either dealing with or preventing risks, for example vaccination, hand washing, wearing seat belts, and early screening for diseases. Several theories of behavior change have incorporated risk perception variables. Some of these theories—the Health Belief Model, Protection Motivation Theory, the Extended Parallel Process Model, and the Risk Perception Attitude framework—try to predict behaviors. Other theories try to explain how risk perceptions are formed and changed, and they pay special attention to the roles that media play in these processes. Because of this volume’s focus on health and risk message design and processing, this second type of theory is explored in more detail in the later parts of this chapter.
The Health Belief Model (HBM) assumes that people want to avoid illness and that they will adopt behaviors which they believe will protect them from illness. The HBM identifies four types of risk perceptions as determinants of health behavior: perceived susceptibility, perceived severity, perceived benefits, and perceived barriers. Perceived susceptibility refers to people’s subjective beliefs about how vulnerable and susceptible they are to a disease or other health risk (Janz & Becker, 1984). Perceived severity refers to how serious people believe a health risk to be, and whether it will have adverse physical consequences such as death, disability, and pain, and adverse social consequences such as ostracism, stigma, and shame. Perceived benefits refers to people’s beliefs about whether a health behavior will enable them to manage a health risk. Perceived barriers refers to people’s beliefs about whether the costs or negative aspects of adopting a health behavior will prevent them from doing so.
Perceived susceptibility and severity also play roles in Protection Motivation Theory (PMT; Rogers, 1983) and the Extended Parallel Process Model (EPPM; Witte, 1992). Both theories explain that perceived susceptibility and severity constitute people’s perceived threat, which is a precursor to adopting a recommended health behavior. A high level of perceived threat is a necessary but not sufficient condition for adopting such a behavior. If people are not confident that they carry out recommended actions (self-efficacy) or if they doubt those actions can control the threat (response efficacy), they will not adopt the recommended behavior. Similar to EPPM, the Risk Perception Attitude (RPA) framework also presumes that risk perceptions (both perceived susceptibility and severity) play key roles in motivating people’s changes in health behaviors (Rimal & Real, 2003). While EPPM tries to understand the underlying mechanism of how fear appeals work, RPA is useful for predicting individuals’ motivations and self-protective behaviors and for segmenting audience characteristics accordingly. Altogether, these health behavior theories and theoretical models highlight the importance of understanding risk perceptions as determinants of preventive and protective health behaviors.
These theories are relevant to health and risk communication because they can guide message design. To take the case of vaccination, if people do not adopt behaviors such as getting shots for themselves or their children, formative research could be conducted to identify what risk perceptions in HBM are relatively low and which of them are most closely related to people’s behavioral intentions. If people perceive that the benefits of getting vaccinations are most closely related to their vaccination behavior, vaccination campaign messages could be designed to amplify the perceived benefits of vaccination.
Other major theories of risk perception address the issue of how risk perceptions are formed and changed. We learn about many risk issues indirectly—either from other people or from the media. Media play critical roles in forming and affecting risk perceptions, and researchers have identified a variety of media factors that affect the general public’s risk perceptions (McCarthy, Brennan, Boer, & Ritson, 2008). These factors include the following: (1) amount of media coverage; (2) frames used for presenting risks; (3) valence and tone of media coverage; (4) types and trustworthiness of risk information sources; (5) media message formats; and (6) types of media. To varying degrees, the relevance to risk perception of each of these media factors has received theoretical and empirical support.
1.
Media coverage. If the media devote a lot of coverage to a risk issue, it will become more salient to the public. In turn, the public will regard the issue as important. Media research on agenda setting has explicated this link between the media agenda (the issues that journalists and other media professionals consider to be worth covering) and the public agenda (the issues that members of the general public care about). For risk communicators, this link has important implications. If increased media coverage can heighten the public’s perceptions of risk issues, risk communicators should take special care to provide credible risk-related information and to identify specific actions that the public should take.
2.
Media framing of risk issues. Compared to the amount of coverage that the media devote to a risk issue, what is often more important is the way they present it. One important aspect of presentation is media framing, a topic widely studied in communication research. Framing refers to the process of “selecting and highlighting some facets of events or issues, and making connections among them so as to promote a particular interpretation, evaluation, and/or solution” (Entman, 2004, p. 5). Researchers have studied many of the relations between the way the public perceives risk issues and the way the mass media frame them. Media frames, which are also called news frames, consist of “the words, images, phrases, and presentation styles that a speaker (e.g., a politician, a media outlet) uses when relaying information about an issue or event to an audience” (Chong & Druckman, 2007, p. 100).
When the media cover a scientific issue that may be related to a risk, they tend to use frames that emphasize the issue’s dramatic characteristics. News stories on risk tend to emphasize who is responsible for causing or solving the risk, what actions people can take to deal with it, and what information should make them feel reassured (Oh et al., 2012). Many studies have content-analyzed media coverage to understand how journalists frame risk issues and how frames appear in certain patterns in media coverage. However, few studies have determined which exact types of media frames most strongly affect people’s risk perceptions.
3.
Valence and tone of media coverage. News media tend to pay special attention to the emotional aspects of risk issues and to select issues that generate strong emotions. Emotions felt by the public with respect to risk issues typically include dread, worry, anger, distrust, and distress (Sandman et al., 1993). Journalists may highlight such emotions over information and statistics about risks. Sometimes they even create spurious phenomena such as “media pandemics” (Gainor & Menefee, 2006; Paek et al., 2008). Journalists also tend to focus more on human interest topics, highlight worst-case scenarios, and describe risks with sensationalistic and emotionally charged language. For example, a content analysis of the environmental risk news stories that were chosen as best articles by newspaper editors found that 68% of the stories featured conflicts and emotionally charged opinions and did not include any risk information (Sandman et al., 1987). In another study, Johnson, Sandman, and Miller (1992) found that information about people’s emotional reactions to risk had a substantial effect on risk perceptions, while technical details about the risk had no effect. Through these and other empirical studies, Sandman and his colleagues have affirmed the importance of emotional content in shaping public perception of risk.
4.
Risk information sources. The types of sources that are used in media coverage on risk issues can also influence people’s risk perceptions. Journalists may favor sources who have strong opinions and can generate exciting debates, or they may seek out sources who help them balance publicly expressed views on controversial issues. They tend to use government, industry, and expert sources to represent the “safe” side of risk debates, and they tend to use activists and laypeople to represent the “risky” side (Sandman, 1997). People’s risk perceptions may also be affected by their perceptions of sources’ trustworthiness. In risk communication literature, trust has been found to play a significant role in predicting people’s risk perceptions, risk-preventive behaviors, and support for government. For uncertain health risk issues such as Ebola, MERS, and the Zika Virus, or for abrupt and unexpected natural or human-made risks, people may rely on the scientists, experts, or government officials who appear as sources in media coverage. However, if people distrust any of these sources, they will doubt the information they provide, and this doubt will in turn affect their risk perceptions. The more people trust the institutions that deal with risk issues, the more likely they will accept certain risks (Peters, Covello, & McCallum, 1997). By contrast, when risk communication efforts prove ineffective, lack of trust may be the cause. According to the asymmetry principle, people tend to notice negative and trust-destroying events more than positive and trust-building events, and they tend to consider sources of bad and trust-destroying news to be more credible than sources of good news (Slovic, 1993). While building trust in risk situations is important, such psychological tendencies provide additional challenges to health and risk communicators.
5.
Risk presentation formats. The ways in which media present risk information can also affect people’s risk perceptions. Different message formats may convey uncertainty differently. Uncertainty is a central issue of risk perception because it affects how people perceive the risk itself, how they interpret risk information, and how motivated they will be to seek additional information about it (Powell et al., 2007). The two basic formats for presenting risk information are verbal and numerical estimates (Wardekker et al., 2008). Verbal estimates present risks without numbers, and the words in them tend to be vague, such as likely, unlikely, probably, possibly, etc. (Hove et al., 2015; Wallsten et al., 1986). Numeric estimates present risk information with numbers that either stand alone, or appear in ranges, or are accompanied by verbal qualifiers.
Empirical studies have shown that these different risk presentation formats may have varying effects on audiences’ perceptions of and reactions to risks. On the issue of which format is more effective, there is still no consensus, and relatively little research has examined how media stories using the different formats affect people’s risk perceptions. One exception is an experimental study gauging audience reactions to media stories on H1N1, mad cow disease, and carcinogenic hazards in South Korean contexts (Hove & Paek, 2015). Findings were not consistent across all these topics, but the numeric presentation format generally yielded a higher level of risk perception than the verbal presentation format.
6.
Genres and types of media. The genres and types of media in which risk messages and information appear can also affect risk perceptions. Two basic genres of media—news and entertainment—have been found to influence people’s risk perceptions, sometimes in different ways. Compared to other media factors, types of media have been more systematically analyzed for how they affect risk perceptions. Two competing theoretical hypotheses have emerged—the impersonal impact hypothesis and the differential impact hypothesis.
The impersonal impact hypothesis focuses on the effects of news media, and it makes a distinction between people’s personal-level and societal-level judgments about risk. Personal-level judgments refer to individuals’ beliefs about how much a risk threatens themselves, while societal-level judgments refer to their beliefs about how much a risk threatens collectives such as a city, a nation, or the world population (Tyler & Cook, 1984). This distinction is important because personal-level risk perception may directly lead to preventive behaviors, while societal-level risk perception may not have such a direct influence. The impersonal impact hypothesis predicts that news media exert more powerful impact on societal-level than on personal-level risk judgments. The reason may be that when the news media feature a risk issue, journalists are more likely to describe it as a threat posed to generalized others whom audiences do not imagine as being similar to themselves.
Different media types (television, print) and media genres (news, entertainment) may play different roles in influencing risk perceptions. For example, one study found that television exposure predicted personal-level risk perception, while frequent use of newspaper news predicted societal-risk perceptions for voluntary health and risk issues (heart disease, AIDS, smoking) (Coleman, 1993). Based on such findings, an alternative hypothesis has been proposed to acknowledge the potentially different roles played by different types of media.
The differential impact hypothesis predicts that entertainment media are more likely to influence people’s personal-level risk perceptions while news media are more likely to influence their societal-level risk perceptions. Entertainment media tend to present risks in dramatic and emotional ways. Compared to news, dramas and movies may make a given health threat seem more salient and personally relevant. For example, a study on the portrayal of AIDS in the media found that movies and situation comedies were significantly related to personal-level risk perception (Snyder & Rouse, 1995). It is possible that exposure to news media affects the cognitive dimension of people’s risk judgments, while exposure to entertainment media affects the emotional dimension of risk judgments and personal-level risk perceptions. More research needs to explore how new and hybrid media genres such as health infotainment and neomedical documentaries affect people’s risk perceptions compared to traditional news and entertainment genres. A study comparing portrayals of smoking in Korean media found the following: exposure to news programs predicted smokers’ personal-level risk perceptions; exposure to entertainment programs predicted nonsmokers’ personal-level risk perceptions; and exposure to infotainment programs predicted both smokers’ and nonsmokers’ societal-level risk perceptions (So et al., 2011). These findings indicate that media type/channel/genre and audience characteristics (e.g., personal relevance, behavioral status, motivations) may also interact to affect people’s types of risk perceptions.
Discussion of the Literature
There has been growing recognition that risk communicators need to understand the many dimensions of people’s risk perceptions in order to do their jobs effectively. However, more efforts need to be devoted to understanding the determinants of risk perceptions and the underlying mechanisms through which risk perceptions affect subsequent behaviors. Some scholars have argued that the concept of risk perception is overly complex and vague, and that it should instead be called risk judgment because it has not only perceptual but cognitive, affective, and behavioral dimensions (Dunwoody & Neuwirth, 1991). However, the concept of risk judgment seems to focus more on the cognitive and rational aspects of risk perceptions and to overlook the various emotional ways in which people respond to risks. This emphasis on the rational and cognitive aspects of risk perceptions is also reflected in health behavior theories such as the HBM, PMT, EPPM, and RPA framework. These theoretical models commonly include perceived susceptibility and perceived severity, and they consider these risk perceptions as precursors to health behaviors. Even on the cognitive side of risk perceptions, some researchers have argued that more dimensions need to be explored. For example, the authors of a meta-analysis proposed adding perceived likelihood, the probability that one will be harmed by a risk (Brewer et al., 2007). Based on 34 studies that these authors reviewed, perceived likelihood seems to be a distinct component of risk perceptions that have consistently been found to be related to health behaviors.
The emotional side of risk perceptions has been extensively studied by Slovic and his associates as part of their work on the affect heuristic and the risk-as-feeling hypothesis. Their theoretical arguments overlap with Sandman’s argument defining risks as a combination of hazards and outrage. Covello’s identification of risk perception factors also overlaps with these concepts. However, researchers still need to determine how and to what extent discrete emotions (e.g., anger, fear, worry) affect risk perceptions and subsequent behaviors. Although a variety of risk characteristics have been identified under the headings of unknown and dread risk, little research has explored which exact ones have stronger or weaker impacts on risk perceptions and subsequent behaviors.
Such detailed explorations would also need to be more fully integrated into research that examines how interpersonal and mediated communication affect perceived risk characteristics and risk perceptions. A recent study found evidence for the differential roles of the cognitive and emotional dimensions of people’s perceived risk characteristics in risk perceptions, as well as for a more promising role of entertainment media in the process (Oh et al., 2015). Explicating the underlying mechanisms through which news and entertainment media affect risk perceptions in the context of H1N1 in South Korea, it found the following: (1) exposure to news media is positively correlated with the cognitive dimension of risk characteristics, while exposure to entertainment media is positively correlated with both their cognitive and the emotional dimensions; (2) the emotional, but not the cognitive, dimension of risk characteristics is positively related to both personal- and societal-level risk perceptions; and (3) exposure to entertainment media affects personal-level risk perceptions—but only indirectly, through the emotional dimension of risk characteristics.
Media continue to play a critical role in communicating risks to the public, and scholars continue to try identifying media factors that affect risk perceptions. While some media factors, particularly media frames, have been extensively researched, others such as risk presentation formats need more attention. Efforts to understand how various verbal and numerical risk presentation formats can affect risk perceptions and subsequent behavior could help health and risk communicators in developing more effective messages and campaigns. However, such efforts would also need to take account of audience characteristics such as numeracy and relevant personal traits (e.g., uncertainty avoidance, risk seeking tendency, optimistic bias).
Research on media channels and genres has generated interesting theoretical hypotheses. Distinguishing the personal and the societal levels of risk perceptions has moved risk perception research one step further. This distinction is important because these two levels of risk perception have differential impacts on subsequent behaviors. For example, if people think infectious diseases are more likely to affect only other people or society in general, they may not take preventive actions or follow government recommendations such as quarantining and stockpiling.
Researchers have also made promising new discoveries regarding the different ways in which the media genres of news and entertainment influence risk perceptions. Recently, several studies have explored how and why people use hybrid genres of media such as edutainment, infotainment, and genre-specific media, and how these new media types affect risk perceptions. However, this area of research is limited because it still relies on survey measurements of media exposure (e.g., frequency, amount of media use). Such methods cannot capture the fact that, even within each type of informative or entertainment media, the way risk information is presented and framed could affect risk perceptions and subsequent behaviors. Researchers have also not yet adequately explicated the mechanisms through which media exposure affects risk perceptions. However, some attempts have been made to address these limitations and to examine how emotionally charged news media affect risk perceptions. One recent study’s findings suggest that the distinction between informative and entertainment media has become blurrier, and that people’s emotional reactions (i.e., fear) could also have different effects on their personal- and societal-level risk perceptions (Paek et al., 2016). Future research should try to replicate such findings and examine the differential effects of discrete emotions such as shame and anger on risk perceptions and subsequent behaviors.
Finally, research that examines how media types/genres/channels/platforms affect risk perceptions needs to catch up with developments in social media. On a variety of new media platforms, people who were formerly passive receivers of risk information from traditional media have now become active producers and disseminators. During outbreaks of infectious diseases such as Ebola in the United States and MERS in South Korea, communicators on social media such as Twitter played critical roles in the rapid production, sharing, and dissemination of information, but often without much regard for accuracy. In some cases, rumors and misinformation about MERS that were originally disseminated through social media managed to influence the traditional and public media agendas. Because of the emerging role of social media in risk and crisis situations, the World Health Organization (WHO) has now issued outbreak communication guidelines that recommend how to handle rumors on social media. However, academic research has not yet fully addressed such issues, with the exception of some studies in public relations and crisis communication (e.g., Schultz, Utz, & Göritz, 2011; Utz, Shultz, & Glocka, 2013). More serious theoretical and empirical efforts should be made to integrate research on social media across disciplines in order to have a better understanding of how source, medium, message, risk/crisis type, and audience characteristics interact to affect the public’s risk perceptions and subsequent behaviors. Such understanding could enhance risk communication by making it more effective in giving people appropriate risk perceptions and motivating them to carry out recommended actions.
