Scientific uncertainty in media content: Introduction to this special issue

Abstract

This introduction sets the stage for the special issue on the public communication of scientific uncertainty that follows by sketching the wider landscape of issues related to the communication of uncertainty and showing how the individual contributions fit into that landscape. The first part of the introduction discusses the creation of media content as a process involving journalists, scientific sources, stakeholders, and the responsive audience. The second part then provides an overview of the perception of scientific uncertainty presented by the media and the consequences for the recipients’ own assessments of uncertainty. Finally, we briefly describe the six research articles included in this special issue.

Keywords

scientific uncertainty science communication science journalism media reception media effects

1. Introduction

Uncertainty is central to science, making both journalistic representations and their impacts on audiences critical to our understanding of the science journalism process. While the reader will encounter a growing body of scholarship in this arena, a summer conference on the topic, organized in the framework of the special priority program “Science and the Public: Understanding Fragile and Conflicting Scientific Evidence” of the German Research Foundation (DFG) at the Forschungszentrum Jülich in Germany, offered an opportunity to bring together a set of scholars and current work that we hope will inform the uncertainty-and-media dialogue. This introduction is intended to set the stage for the research that follows.

Definition of “scientific uncertainty”

It is important to differentiate “probabilistic uncertainty,” defined as uncertainty about the occurrence of specific events whose statistical distribution is known, from “epistemic uncertainty,” which is uncertainty about the validity of truth claims. This special issue focuses on epistemic uncertainty of scientific claims, that is, claims based on scientific evidence or theories and/or claims communicated by scientists.

Discussing random error in inferences, possible systematic errors, generalization issues, and competing explanations or interpretations is the rule in science and demanded by the norm of “organized skepticism” that Merton (1957) considered a crucial aspect of the scientific ethos. However, not only do researchers and their peers discuss the (un)certainty of scientific knowledge but also journalists, decision-makers, and audiences may reflect on it. These external actors may base their reflections on epistemic arguments or opinions (e.g. incompatibility with competing everyday knowledge, cf. Rowan, 1999) or on alleged cognitive biases or motivations of researchers, research organizations, or “science” in general. Trust in scientists, scientific organizations, and science in general, thus, becomes interlocked with the perception of scientific uncertainty.

2. Different ways to deal with (un)certainty in the scientific versus public arena

“Reconstructions” of science vary across different arenas that are characterized by specific combinations of audience, channel/medium, and communication purpose. In these reconstructions, consideration of scientific uncertainty may differ, allowing for, among other things, different communication purposes between communicators and audience. While some authors have rightfully pointed to the rich diversity of arenas (Bucchi, 1998; Fleck, 1935; Hilgartner, 1990), many have settled for a dichotomy: a public arena constituted by journalistic media such as TV, radio, newspapers, and general interest magazines (including their online equivalents) and a scientific arena constituted by scholarly communication among scientists in scientific journals and conferences (Post, 2013; Rühl, 1981). Both are ideal types rather than homogeneous settings, of course. In another popular dichotomy, many authors have contrasted the specific norms, quality criteria, and communication preferences of scientists and journalists (Friedman et al., 1986; Peters, 1995; Salomone et al., 1990).

Some authors suggest that scientific claims in the mass-mediated public arena have traditionally been separated from the context of discovery and are, thus, more likely to be reported as certain than would be the case in the scientific arena (Fahnestock, 1986; Fleck, 1935). They find a shift in focus in that public arena from the question “Is it true?” to “What does it mean?” Such a shift may be the result of different relevance judgments as well as a side effect of simplification for a lay audience. If one has in mind a “diffusion of knowledge” model (Fleck) or a science popularization model (Fahnestock), this assumption seems plausible.

However, the public reconstruction of “science” nowadays is more complex and often does not follow a simple dissemination-of-knowledge paradigm. Scientific controversies are covered by the media, scientific experts publicly disagree about facts and conclusions relevant for decision-making, scientific misconduct leading to errors and deceptions is reported, and the interdependencies between science and industry or between science and politics are discussed as possible causes of bias in scientific knowledge. It is, thus, likely that we find a mix of knowledge dissemination that often skips over epistemic uncertainty as expected by Fleck and also critical discussion of competing knowledge claims with references to epistemic uncertainty as an essential part of the discourse. In particular, in the reporting of science as expertise, that is, in the context of social problems and decisions, scientists and/or journalists may highlight scientific uncertainty, or uncertainty may become obvious to the audience through inclusion of contradictory statements from scientific experts.

3. Audiences encounter scientific uncertainty in mediated communication

Singer and Endreny (1993) claimed that probabilistic uncertainty is neglected in the journalistic coverage of risk, creating a sense of certainty that the risk will actually cause damage and, thus, producing a tendency to overestimate low-probability/high-consequence risks that get media attention. However, this widely cited finding should not be misunderstood as proving the absence of references to scientific uncertainty in media coverage. Although much science coverage indeed continues to report scientific results as “facts” without discussing the epistemic basis and assessing the level of certainty, science journalists today quite regularly confront their audiences with different kinds of “scientific uncertainty”:

In some instances, the debate over truth claims within science spills over into the public sphere or may even originate there. Some examples are cold fusion (Bucchi, 1998; Weingart, 2001), the relationship between CO₂ emissions and global warming (Boykoff and Boykoff, 2004), or the reported discovery of bacteria substituting arsenic for phosphorus in their metabolism (Yeo et al., 2016). In other cases, public controversies erupt between science and pseudo-science, such as in the evolution versus creation debate (Bleckmann, 2006).

The problematic generalizability of findings to other species, populations, situations, cultures, or historical times sometimes becomes a focal uncertainty point in media accounts. For example, one of the arguments against the use of lab animals in biomedical research is that results of experiments with mice or rats may not be applicable to humans (Engber, 2011), an issue that has received much media attention.

Media often reflect on the uncertainty of attributing causality to events and developments. For example, the German media reported extensively about the rejection, by an investigatory court, of the expertise of a British mass panic expert regarding the causes of a mass panic at the Love Parade 2010 in Duisburg, Germany, which killed 21 people (Diehl and Gebauer, 2014). Also, epidemiological results covered by the media are often accompanied by cautionary comments about causal interpretation of correlations (e.g. Kolata, 2016).

Tentative or controversial interpretations of research results can lead to a media focus on uncertainty. For example, the suggested implications of neuroscientific experiments on unconscious decision-making by Libet (1985) for the assumption of human “free will” have been criticized not only within academia but also in public (Ananthaswamy, 2012). Another example is the discovery of fossils of Homo naledi in South Africa and the scientific discussion, mirrored in popular websites and media, about the interpretation of the site where the fossils were found as a burial place (Wong, 2016).

Public controversies over science and technology and about risks and benefits of policy decisions have received much attention in the past (Kepplinger, 1989; Mazur, 2004; Nelkin, 1992; Nowotny, 1982; Rothman, 1990). In these issues, the evidence for risks and benefits is frequently challenged, with experts active on both sides. Examples range from issues of public health (e.g. MMR (measles, mumps, and rubella) vaccination, fluoridation) and technological innovation (e.g. nuclear power, genetic engineering) to environmental issues (e.g. the ozone hole, climate change).

Communicators and mediators have developed many ways to characterize the uncertainty of knowledge claims through the use of specific words, expressions, or rhetorical figures (see Simmerling and Janich, 2015). They can discuss limitations of the research used to justify the claims, mention dissenting views, or point to a need for further research. Finally, they can discuss the trustworthiness of researchers or communicators, questioning their competence and suggesting biases motivated by the researchers’ own interests or by organizational or sponsor pressure.

A systematic overview of the prevalence of “scientific uncertainty” in media accounts remains to be accomplished. A recent content analysis of media coverage of neuroscience in a sample of German and US print and online media by Peters et al. (2013) found that, in both countries, about 30% of the articles mentioned scientific uncertainty in some way: 18% of the articles explicitly acknowledged scientific uncertainty or missing evidence, 10% acknowledged research demand, and 15% mentioned a controversy between researchers (multiple coding possible). Longer articles more often included references to scientific uncertainty than shorter articles. In a sample of 207 German TV programs on molecular medicine, Ruhrmann et al. (2015) found that 63% of them dealt with scientific uncertainty. Most frequently mentioned were “missing research results,” “doubt over the application of experimental findings to humans,” or “different interpretations of the data.”

4. Uncertainty and actors in the public communication process

Mediated science is typically the product of multiple actors: journalists, their sources, and members of the audience, whose online comments often supplement articles on websites. Scientists, scientific publications, and press releases are important sources for journalists. But stakeholders whose interests are affected by public reactions to knowledge claims often weigh in on these claims in ways that attract journalists’ attention. As Anderson et al. (2014) have shown, comments by audience members following online stories can alter the perceived meaning of those stories. Discussions of uncertainty, as well as explicit or implicit characterizations of the (un)certainty of scientific knowledge claims in mediated communication, reflect the combined influences of all these actors.

Understanding why and in which situations uncertainty information is included or excluded in media content, thus, requires an understanding of the communication approaches of the involved actors. Scholars have explored a broad spectrum of possible constraints, professional practices, and motivations that may play a role while acknowledging that public communication of scientific uncertainty is not just motivated by accuracy goals but, in many instances, serves rhetorical and strategic functions (Simmerling and Janich, 2015; Stocking and Holstein, 2009; Zehr, 1999). Below, we discuss some factors that influence how scientific uncertainty is included in public communication with respect to journalists, scientific sources, (other) stakeholders, and responsive audience.

Journalists

Journalists can become alerted to the uncertainty of scientific claims through the communication of scientists or other scientific sources who may mention issues of statistical significance, systematic error, or competing interpretations in scientific publications, press releases, or interviews. Furthermore, journalists themselves may be aware of contradictory claims, or they may find scientists’ claims implausible on the basis of their own understanding of scientific processes, common-sense reasoning, or the incongruity of these claims with the journalists’ own beliefs. Finally, they may distrust their scientific sources based on previous experience or the perceived interests of these sources (e.g. because of their relationship with industry or government), leading to doubts about the validity of claims.

If journalists perceive scientific uncertainty in the claims they want to report on, they have several options: they can ignore the uncertainty and report on the scientific claims as if these claims were certain, they can abstain from publishing the story or part of the story, or they can account for the perceived uncertainty via various rhetorical forms or by applying the balance norm and including competing views in their coverage (Boykoff and Boykoff, 2004; Kohl et al., 2015). Stocking (1999) mentions several factors supporting the tendency of the media to “present science as more solid and certain than it, in fact, is” such as using only one source in a story, setting aside context, and emphasizing scientists’ efforts to reduce uncertainty (science as a “triumphant quest”).

Journalists may have a preference for “certain” scientific results as they—probably correctly—assume that their audience is looking for valid and concrete information about the world when they turn to the media. Although it is difficult to show conclusively, potential stories may remain untold because journalists consider the news value of a story low if the core information is uncertain (Lehmkuhl and Peters, 2016).

However, in other cases, a story may be perceived as newsworthy by journalists even if the claim on which the story is based is uncertain, such as in the communication of risk, or if scientific speculation is fascinating, although a vague possibility (e.g. time travel, particles moving faster as light, parallel universes). And in still other cases, the story may be newsworthy because of the uncertainty it creates if that uncertainty challenges established knowledge, relates to a social conflict, or criticizes a scientific–technical project (Lehmkuhl and Peters, 2016; Simmerling and Janich, 2015).

Guenther and Ruhrmann (2016) showed that science journalists’ intentions to discuss scientific uncertainty in stories depended on their perception of the presence of uncertainty in their field of coverage. Furthermore, anticipated audience expectations and the expected strategy of other media toward inclusion of uncertainty played roles as well. According to Lehmkuhl and Peters (2016), a reporter’s specialization and reputation, as well as the media environment in which he or she works, influence strategies for dealing with scientific uncertainty. This leads us to conclude that how journalism deals with scientific uncertainty is extremely context-sensitive.

Scientific sources

Scientists seek to reduce the ambiguities of what we know about the world around us. But in the pursuit of that goal, they need elaborate strategies to manage uncertainties within science and in their communications across the borders of science (Zehr, 1999). Uncertainty is a major motivator for research. To justify research and mobilize resources, for example, in a research grant proposal, scientists need to identify the presence of sufficient uncertainty and make the case that knowledge gained will substantially reduce that uncertainty. Scientists are rewarded, of course, for making original and well-supported truth claims (Cole and Cole, 1967; Gaston, 1970). Thus, it is in their interest to stress scientific uncertainty or lack of knowledge as a motivator of their research and then to downplay the remaining uncertainty thereafter. Furthermore, as psychological research on “overconfidence” has shown, scientific experts are not free from biased perceptions of the validity of their own findings and knowledge (Lin and Bier, 2008).

Because of the risk that “knowledge” might be accepted as valid too readily, scientific norms require scientists to beware of possible errors and the limited accuracy of measurements or statistical inferences. Procedures that minimize errors and unjustified conclusions are part of the methodology of any academic discipline, although these procedures vary with the kind of research conducted. Sociologists consider dealing with uncertainty as part of the epistemic culture of any scientific discipline (Wehling, 2015).

When scientists communicate their results and knowledge in public arenas—when they talk to journalists, for example—they have to decide whether they should also communicate scientific uncertainty in that public setting. As a general principle, scientists seem to be prepared to include scientific uncertainty and disagreement in public communication, but not without reservations. In a survey of 1354 biomedical researchers from Europe, United States, and Japan, respondents moderately agreed (mean value = 0.9 on a rating scale ranging from −2 to +2) that “the public should be informed when scientists disagree about relevant issues” and slightly disagreed (mean value = −.5) that “discussions of uncertainties about facts and models should be kept within the scientific community” (Peters et al., 2009).

Post (2016) found, in a survey of German climate researchers, that the readiness of scientists to admit “that many questions about climate change are still unresolved” correlated negatively with their involvement in mediated communication. It is unclear whether this correlation results from journalists’ preferences for clear messages or from scientists’ consideration of the consequences of admitting uncertainties, for example, for the public perception of climate change or for their own credibility. Besides the question of whether the scientific norm of addressing uncertainty in scholarly communication should also be applied to public communication, a number of further considerations come into play:

Adaption to non-scientific audiences. Scientists may (correctly) assume that non-scientific audiences are less interested in the epistemology of research than in the outcomes and that addressing issues of epistemic uncertainty would increase the complexity of the message. However, when it comes to knowledge relevant to decisions about policy or health-related behavior, the audience may expect to encounter information about uncertainties, as this information is crucial for efficient decision-making (Fischhoff and Davis, 2014). Withholding information about uncertainty may then lead to massive public criticism or even to legal prosecution, as was the case of the communication of reassuring messages before the earthquake disaster in L’Aquila, Italy (Nosengo, 2012).

Concern about public image. In functionally differentiated societies, the task of science is to create relevant and reliable knowledge. Public trust in science and public support for research probably depend on the perception that science fulfills that task well. Science increasingly worries about its public image (Weingart, 2012). Individual scientists, research organizations, and scientific associations are, thus, tempted to emphasize a good fit between their research and public expectations, and they may perceive communication of uncertainty as counter-productive to that goal. On the other hand, if scientific uncertainties are obvious to the public, ignoring them may lead to mistrust, and communicators can improve their public image if they address uncertainties in a professional way (Jensen, 2008; Zehr, 1999).

Anticipation of persuasive effects. Scientists and science organizations often communicate to persuade, such as influencing risk perceptions, public opinion toward technologies, or public policy. If communicators assume that uncertainty information will alter the effects of their messages, they may be tempted to “optimize” inclusion or exclusion of uncertainty information rather than view uncertainty as a component of scientific accuracy. One can spot such motivations in efforts to increase compliance with vaccination campaigns, raising support for climate protection policy, promoting acceptance of a technology, or reassuring the public about a risk, for example.

Dunwoody (1999) has argued that the impact of scientific sources on media coverage of science is strong, partly because of the authority science enjoys in our societies, and that they “will maintain the upper hand in determining the characteristics of scientific uncertainty and will play an increasing role in deciding when and how those elements appear in mass media accounts” (p. 76). Analyzing the misrepresentation of the certainty of conclusions in scientific publications dealing with attention deficit hyperactivity disorder (ADHD) and their impact on media representations, Gonon et al. (2011) point to the responsibility of science for the accurate reporting of uncertainties. The tendency to mask scientific uncertainty is an important issue in the current debate on science hyping (Caulfield et al., 2016). Several scientific associations and academies have issued guidelines aiming at more accurate public presentation of research results, explicitly including the truthful communication of scientific uncertainties (International Society for Stem Cell Research (ISSCR), 2016; On Designing Communication between the Scientific Community, the Public and the Media, 2014). Similar to our conclusion on how journalists report on scientific uncertainty, scientists’ reflections on scientific uncertainty are extremely context-sensitive, too.

Stakeholders

For the purpose of this article, we define stakeholders as groups or organizations that have an interest in a specific outcome of a communication process, consonant with particular goals. The preferred outcome may be a certain public framing of an issue and, often, a persuasive effect.

Our distinction between “science” and “stakeholders” as actors in the public communication of science is not without problems, since science is a stakeholder, too. The goals of science may include such things as public funding and science-friendly regulation of research. But a stakeholder role also comes into play if scientists or science organizations want to support the utilization of scientific knowledge in technical or medical innovations or if they want to improve the scientific “rationality” of policy decisions (e.g. climate policy) or individual behavior (e.g. non-smoking).

Scientific claims often have implications for the goals of stakeholders. For example, a claim that carbon dioxide emissions increase global warming challenges the business of oil companies and of companies operating coal-fired power plants. These stakeholders may, therefore, have an interest in focusing on the uncertainties of global warming. But the same claim may legitimize a policy aiming at a transition of the energy system from fossil to renewable energy sources. Stakeholders such as non-governmental organizations (NGOs) or manufacturers of wind energy plants and solar panels may, thus, have an interest in downplaying uncertainties related to the claim of human-made climate change.

Post and Maier (2016) analyzed the factors influencing stakeholders’ intention to mention scientific uncertainty about biotechnological research in talks with journalists. They were indeed able to show that stakeholder intentions depended on their respective persuasive goals. For example, scientists and representatives of industry were more prepared to talk about uncertainties if they believed that this would increase research funding. Government agencies were less prepared to talk about uncertainties if they believed that this would make the public more critical, an outcome that increased the uncertainty communication intentions of public interest groups.

Lay audience participation in uncertainty messaging

Unfortunately, little is known about the patterns of audience comments to online newspaper stories or other online content. A study of reactions of individuals reading newspaper stories or viewing TV magazine stories about genetic engineering showed that many thoughts induced by the stories were critical of the views of scientists quoted in them (Peters, 2000). Similar results emerged in a study of reactions to newspaper articles about climate change (Peters and Heinrichs, 2005). Even if a story did not include uncertainty cues or dissent, recipients compared the story content with their personal knowledge and attitudes and generated counter-arguments to the claims of the scientists quoted or mentioned feelings of distrust.

Assuming that thoughts evoked during consumption of media content motivate audience comments, we suggest that critical audience comments can add uncertainty elements to a media story that itself does not mention uncertainty. We might even hypothesize that the motivation to comment on online content is greater if people disagree with the main claim of a story than if they agree. Although not specifically dealing with the perception of uncertainty, the Anderson et al. (2014) study of the effects of comments on the perceived meaning of a media story by subsequent readers suggests that audience comments may indeed influence impressions of the uncertainty of the claims made in that story.

5. Mediated-message effects on lay perceptions of uncertainty

Messages can certainly influence perceptions of uncertainty. Exploring this literature, however, may benefit by first making a conceptual distinction between external and internal uncertainty. This external/internal delineation is explicated in the work of Kahneman and Tversky (1982), who argued that perceived uncertainty is often misrepresented as a unidimensional construct (an argument countered by the discussion above as well). In addition to reflecting on an array of different “states of mind,” they note, uncertainty can be situated in two loci. On one hand, external uncertainty derives from actions in the world around us; we can perceive these causal processes but have no control over them. Internal uncertainty, in contrast, reflects judgments made in one’s own mind, decisions that presumably are under one’s control (see, also, Løhre and Teigen, 2015).

Recast in a way relevant to the narrative at hand, external uncertainty resides in attributes of the science that is being communicated, while internal uncertainty reflects a judgmental state of the receiver. Studies of audience interactions with the former typically focus on the level of accuracy achieved in the message communication process (in more colloquial terms, an assessment of the “goodness of fit” between the uncertainty message and audience representations of it). In contrast, studies of the latter assess the impact of uncertainty information on judgmental states and behavioral reactions.

External uncertainty

The perception that non-scientists often misperceive uncertainty, resulting in a bad fit between those perceptions and the way science expresses uncertainty, has served as a catalyst for a sustained focus on strategies for communicating accurately about external uncertainties. In a survey of scientific experts drawn from universities, industry, and government, for example, Frewer et al. (2003) uncovered a widespread belief that the general public is unable to conceptualize uncertainties. Respondents even worried that providing information about uncertainty would sow panic and confusion, as well as encourage distrust of the scientific establishment. To handle these perceived deficits, studies of external uncertainty impacts typically tweak uncertainty representations (ratios vs. natural frequencies, quantitative probability statements vs. verbal “higher/lower” statements) and then evaluate the “accuracy” of resulting perceptions.

For example, Gigerenzer and Hoffrage (1995, 1999) find that expressing uncertainty in terms of natural frequencies, rather than conditional probabilities, makes it easier for non-experts to grasp “probabilistic uncertainty” (Hoffrage et al., 2002: 343), while others (see, for example, Girotto and Gonzalez, 2002) find more parity across strategies.

Using ranges can be challenging. Dieckmann et al. (2015) found that, without explanatory help, some lay individuals interpreted the likelihood of harm as uniform across a range of uncertainties (rather than varying in a roughly normal distribution) and that misinterpretation was less likely in individuals with higher levels of numeracy. When predicting the path of a hurricane, forecasters often display the range of probable paths using a “cone of uncertainty.” Broad et al. (2007) concluded that most visual representations of the cone (e.g. in weather forecasts) do not provide much explanatory detail, leading to variance in audience interpretations. For example, some individuals regard the central line of the cone as the only locations within the cone that will be hit by the hurricane, others declare all areas within the cone as equally likely to be damaged, while still others declare the areas immediately outside the cone to be immune to damage.

Internal uncertainty

While much of the literature examining lay individuals’ statistical reasoning skills, in service to external uncertainty analysis, is critical of those abilities, scholars also acknowledge that non-scientists usually bring to the table a rough understanding of probability. For example, a study of lay individuals’ perceptions of the uncertainty of weather forecasts (Joslyn and Savelli, 2010) found that respondents expected some uncertainty in forecasts and anticipated greater uncertainty as forecasts looked out over a greater length of time. Most people who buy lottery tickets acknowledge that the likelihood of holding a winning ticket is very small. And when flipping a fair coin, someone who gets a run of “heads” typically declares that outcome unexpected, even disconcerting.

But that kind of innate understanding of probability may not be helpful in the face of a specific issue. And mediated accounts of issues have been shown to influence readers’ uncertainty perceptions through a variety of reporting and narrative choices. Put another way, story construction decisions that, in the mind of the reporter, have nothing to do with uncertainty have been shown to affect audience members’ internal representations of uncertainty.

As noted earlier in this article, Stocking (1999) introduced this idea by arguing that some story features (use of single sources, the loss of research caveats, an emphasis on what a study finds rather than how it goes about its investigation) may make a probabilistic outcome seem more certain than it really is, while others (giving equal space to competing truth claims in a single story, separating contested truth claims into separate stories) can make an outcome seem less certain than it really is.

Journalistic norms, particularly efforts to avoid judging the validity of truth claims by maintaining “objectivity” and, instead, sharing with the audience information about the varied truth claims orbiting an issue (a process known as “balance”), have become popular foci of researchers seeking to understand how sometimes inadvertent media representations of uncertainty affect individuals’ personal uncertainty judgments (Binder et al., 2016; Clarke et al., 2015; Corbett and Durfee, 2004; Kohl et al., 2015; Kortenkamp and Basten, 2015). In much—but not all—of this literature, the inclusion of varying and sometimes orthogonal truth claims does indeed strengthen public perceptions of uncertainty.

The main interpretive framework of a narrative—also known as a frame—also has been shown to affect perceptions of uncertainty, as well as to interact with uncertainty representations. Tversky and Kahneman’s (1992) well known prospect theory finds that narrative frames that emphasize loss versus gain dramatically influence risk attitudes. Morton et al. (2011) applied this to climate change risks and found that a loss frame that was highly uncertain (i.e. a 20% likelihood of abrupt and severe changes) led to decreased intentions to engage in environmental behaviors while a gain frame that was paired with high uncertainty (e.g. a 20% likelihood that abrupt and severe changes would not occur as a result of climate change) led to greater behavioral intentions.

Consequences

Studies of mediated information’s impact on uncertainty judgments find that the configuration of messages (see section above on “influence factors”) can influence a variety of “internal” outcomes, including both cognitive and affective reactions, which in turn affect uncertainty perceptions. Additionally, those factors can then lead to changes in decision-making at both individual and policy levels. For example, Powell et al. (2007) found that emotional reactions (worry, anger) to consuming potentially contaminated fish from the Great Lakes, as well as assigning the issue a higher level of personal salience, were associated with greater levels of perceived uncertainty, while knowledge about the potential risk influenced uncertainty perceptions only indirectly. This suggests that media risk stories that generate stronger emotional reactions in readers and viewers will affect perceived uncertainty more than will stories that emphasize information gain at the expense of emotion.

But how might that variance affect “downstream” decision-making? In one study, Kienhues et al. (2011) exposed participants to either conflicting or consensual information about treating high levels of cholesterol and then asked the participants to recommend a course of action to a fictitious “other” who had been diagnosed with high cholesterol levels. The group that received consensual information advised medication significantly more often than they had done before encountering the information, while those who were exposed to conflicting information about cholesterol treatment recommended refusing medication more often.

Scientists and policy-makers worry that sharing the uncertainties embedded in scientific discoveries with lay audiences will lead to a loss of credibility of both science in general and scientists in particular. But studies exploring this possibility typically have found just the opposite: narratives that reflect on caveats and the uncertainties that accompany findings lead lay audiences to confer higher levels of credibility on scientists and their work. For example, in one experiment across five cancer messages, both scientists and journalists were viewed as more trustworthy when the story made an effort to explain limitations of the research and, importantly, when those reflecting on those limitations were the scientists who conducted the research, not scientists unaffiliated with the study (Jensen, 2008).

But that conferral of credibility/trustworthiness in the face of uncertainty messaging may be contingent on a number of factors. For example, Jensen and Hurley (2012) found that story topic mattered. Participants in an experiment who were exposed to conflicting views in a story about wolf reintroduction accorded less credibility to the scientists who served as sources than did individuals who read a story that presented conflicting views on dioxin. The authors speculate that the uncertainty induced by the two stories may have led to different emotional reactions—negative in the case of the wolf story and positive with respect to the dioxin story—which in turn may have prompted those differential credibility evaluations.

The role of uncertainty in larger scale policy making is a complex one, made even more difficult to parse by the extensive use of uncertainty as a rhetorical tool in policy battles. Social scientists have long known about this tendency to politicize uncertainty (see, for example, Campbell, 1985; Zehr, 1999), but it is only in recent years that careful case studies have demonstrated the incredible power of uncertainty arguments to paralyze policy efforts. For example, Heazle (2006) illuminated the ways in which whaling nations employed uncertainty to stall efforts by the International Whaling Commission to regulate the whaling industry; the ploy was so successful that only the catastrophic collapse of whaling stocks forced governments’ hands and resulted in a ban on commercial whaling that remains in place today.

In one of the best known studies, Naomi Oreskes and Erik Conway combed through historical records to assemble the timeline of a longstanding effort by extremely conservative factions in the United States—some well-respected scientists among them—to stall governmental efforts to regulate everything from smoking to CO₂ emissions by emphasizing scientific uncertainty (Oreskes and Conway, 2010). Again, these efforts have been so successful that the US federal government has so far failed to enact emissions reduction policies in order to try to mitigate a warming climate.

If stakeholder interests might be impaired by a mainstream consensus view, stakeholders may be tempted to align with scientific outsiders or maverick scientists in order to emphasize the uncertainty of the mainstream view. Climate change denial in the United States is an extreme example of this strategy (Dunlap and McCright, 2011). Because of the newsworthiness of conflict and the prevalence of a balance norm in journalism, non-mainstream views may get a disproportionate amount of attention in the media, thus creating the false impression of a scientific controversy (Boykoff and Boykoff, 2004; Dearing, 1995; Kohl et al., 2015; Rothman, 1990).

6. Overview of the special issue

Our introductory article seeks to sketch a landscape of uncertainty issues that follow from the complex interplay among scientists, journalists, and audiences. These issues relate, first, to the inclusion of scientific certainty/uncertainty in media coverage and, second, to public reception of media messages and its implications for attitudes and behavior of media audiences. Of the six research articles that compose this special issue, four deal with reconstruction of uncertainty in the media, focusing on journalists (Guenther and Ruhrmann, 2016; Lehmkuhl and Peters, 2016), stakeholders (Post and Maier, 2016), and the coverage itself (Simmerling and Janich, 2015). The other two articles explore the impact of different kinds of journalistic reporting strategies on audience perceptions of uncertainty (Kohl et al., 2015) and the implications for trust in scientists if ethical issues related to their research are mentioned (Hendriks et al., 2016).

With a case study approach, Markus Lehmkuhl and Hans Peter Peters analyze how journalists produce stories, which factors influence whether they perceive scientific uncertainty or not, and how they deal with scientific uncertainty if they perceive it. Based on their in-depth analysis of the genesis of 21 media stories about neuroscientific topics, they conclude that journalists rarely downplay scientific uncertainty but often simply do not perceive it, especially if the journalist is not specialized on science coverage.

In a phone survey of 201 German science journalists covering life sciences, Lars Guenther and Georg Ruhrmann ask about the respondents’ intentions to include uncertainty information in their articles and also measure potential predictors of that intention. They find that the journalists’ intention to include uncertainty information was higher if they perceived uncertainty in the research field they covered, if they believed that the audience expected it, and if they assumed that other media would also include uncertainty information. Gender and past behavior also played a role.

With a similar approach, Senja Post and Michaela Maier conducted 234 phone interviews with stakeholders, that is, representatives from industry, government agencies, and public interest groups, and with scientists. Their analysis shows that the predictors of the intention to mention uncertainties of biotechnological research varied according to the respective interests of these actors. For example, believing that mentioning uncertainties would increase research funds made scientists and representatives of companies more inclined to talk about uncertainties. Believing that uncertainties would make people more critical toward biotechnology made public interest groups more inclined but government agencies less inclined to talk about uncertainties.

Anne Simmerling and Nina Janich analyze in detail references to scientific uncertainty in a single newspaper article dealing with geo-engineering. In this analysis, they illuminate the variety of explicit and subtle ways in which uncertainty is expressed by journalists, including the use of metaphors (“game”), references to science fiction and limited knowledge, or use of modal words (“possibly”). The authors also show that references to uncertainty in the story have important rhetorical functions by contributing to characterizing the geo-engineering project and its proponents as irresponsible.

Presenting conflicting scientific claims in a media story may lead to exaggerated audience perceptions of uncertainty even if one of the claims is supported by many more scientists than the other. A research team consisting of Patrice Ann Kohl, Soo Yun Kim, Yilang Peng, Heather Akin, Eun Jeong Koh, Allison Howell, and Sharon Dunwoody tested the impact of “weight-of-evidence strategies” in reporting controversial claims. Results of their experiment with 352 test subjects confirm that weight-of-evidence information influences the perception of the level of scientific (un)certainty expressed in media stories and, thus, could be a means to attenuate perceptions of unwarranted uncertainty among audiences.

If people doubt the trustworthiness of researchers, they will be uncertain whether to believe the claims the researchers make. Using blog posts by scientists as examples, Friederike Hendriks, Dorothe Kienhues, and Rainer Bromme present results of psychological experiments exploring the impact of information about ethical issues in research on trust ascribed to the researcher. They found that raising ethical issues did indeed lower trust, but only if the ethical issues were mentioned by another scientist or as prior information. If the researcher as blog author mentioned these issues him/herself, epistemic trust remained stable.

Footnotes

Acknowledgements

Compiling this special issue has been challenging and inspiring. We owe many thanks to Professor Robert J. Griffin, who generously agreed to serve as anchor reviewer and also offered a brief reaction to the issue at its end. We also thank the authors of the articles for their cooperation with us and for their contributions, which made this special issue possible.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author biographies

Hans Peter Peters is a Senior Researcher at the Research Center Jülich, Germany, and Adjunct Professor of Science Journalism at the Free University of Berlin. His research deals with public communication of science and technology and the formation of public opinion on science, technology and science-related issues in a media society.

Sharon Dunwoody is Evjue-Bascom Professor Emerita in the School of Journalism and Mass Communication at the University of Wisconsin -Madison. Research interests include journalistic constructions of science news as well as the effects of such messages on audiences, particularly in the face of contested scientific claims.

References

Ananthaswamy

(2012) Brain might not stand in the way of free will. New Scientist, 6 August. Available at: https://www.newscientist.com/article/dn22144-brain-might-not-stand-in-the-way-of-free-will/

Anderson

Brossard

Scheufele

Xenos

Ladwig

(2014) The “nasty effect”: Online incivility and risk perceptions of emerging technologies. Journal of Computer-Mediated Communication 19(3): 373–387.

Binder

Hillback

Brossard

(2016) Conflict or caveats? Effects of media portrayals of scientific uncertainty on audience perceptions of new technologies. Risk Analysis 36(4): 831–846.

Bleckmann

(2006) Evolution and creationism in science: 1880–2000. BioScience 56(2): 151–158.

Boykoff

(2004) Balance as bias: Global warming and the US prestige press. Global Environmental Change-Human and Policy Dimensions 14(2): 125–136.

Broad

Leiserowitz

Weinkle

Steketee

(2007) Misinterpretations of the “cone of uncertainty” in Florida during the 2004 hurricane season. Bulletin of the American Meteorological Society 88: 651–667.

Bucchi

(1998) Science and the Media: Alternative Routes in Scientific Communication. London: Routledge.

Campbell

(1985) Uncertainty as symbolic action in disputes among experts. Social Studies of Science 15(3): 429–453.

Caulfield

Sipp

Murry

Daley

Kimmelman

(2016) Confronting stem cell hype. Science 352(6287): 776–777.

10.

Clarke

Weberling McKeever

Holton

Dixon

(2015) The influence of weight-of-evidence messages on (vaccine) attitudes: A sequential mediation model. Journal of Health Communication 20(11): 1302–1309.

11.

Cole

(1967) Scientific output and recognition: A study in the operation of the reward system in science. American Sociological Review 32(3): 377–390.

12.

Corbett

Durfee

(2004) Testing public (un)certainty of science media representations of global warming. Science Communication 26(2): 129–151.

13.

Dearing

(1995) Newspaper coverage of maverick science: Creating controversy through balancing. Public Understanding of Science 4(4): 341–361.

14.

Dieckmann

Peters

Gregory

(2015) At home on the range? Lay interpretations of numerical uncertainty ranges. Risk Analysis 35(7): 1281–1295.

15.

Diehl

Gebauer

(2014) Love-Parade-Verfahren: Die Pannen des Panikforschers [Love-parade trial: The slip-ups of the panic researcher]. Spiegel Online, 5 April. Available at: http://www.spiegel.de/panorama/justiz/love-parade-warum-das-gericht-die-anklage-ablehnt-a-1085518.html

16.

Dunlap

McCright

(2011) Organized climate change denial. In: Dryzek

Norgaard

Schlosberg

(eds) The Oxford Handbook of Climate Change and Society. Oxford: Oxford University Press, pp. 144–160.

17.

Dunwoody

(1999) Scientists, journalists, and the meaning of uncertainty. In: Friedman

Dunwoody

Rogers

(eds) Communicating Uncertainty: Media Coverage of New and Controversial Science. Mahwah, NJ: Lawrence Erlbaum, pp. 59–79.

18.

Engber

(2011) The mouse trap. Slate Magazine, 15 November. Available at: http://www.slate.com/articles/health_and_science/the_mouse_trap/2011/11/the_mouse_trap.html

19.

Fahnestock

(1986) Accommodating science: The rhetorical life of scientific facts. Written Communication 3(3): 275–296.

20.

Fischhoff

Davis

(2014) Communicating scientific uncertainty. Proceedings of the National Academy of Sciences 111(Suppl. 4): 13664–13671.

21.

Fleck

(1935) Entstehung und Entwicklung einer wissenschaftlichen Tatsache: Einführung in die Lehre vom Denkstil und Denkkollektiv [Genesis and Development of a Scientific Fact: Introduction into the Theory of Thought Style and Thought Collective]. Basel: Schwabe.

22.

Frewer

Hunt

Brennan

Kuznesof

Ness

Ritson

(2003) The views of scientific experts on how the public conceptualize uncertainty. Journal of Risk Research 6(1): 75–85.

23.

Friedman

Dunwoody

Rogers

(1986) Scientists and Journalists: Reporting Science as News. New York, NY: The Free Press.

24.

Gaston

(1970) Communication and the reward system of science: A study of a national “invisible college.” The Sociological Review 18(S1): 25–41.

25.

Gigerenzer

Hoffrage

(1995) How to improve Bayesian reasoning without instruction: Frequency formats. Psychological Review 102: 684–704.

26.

Gigerenzer

Hoffrage

(1999) Helping people overcome difficulties in Bayesian reasoning: A reply to Lewis and Keren (1999) and Mellers and McGraw (1999). Psychological Review 106: 425–430.

27.

Girotto

Gonzalez

(2002) Chances and frequencies in probabilistic reasoning: Rejoinder to Hoffrage, Gigerenzer, Krauss, and Martignon. Cognition 84: 353–359.

28.

Gonon

Bezard

Boraud

(2011) Misrepresentation of neuroscience data might give rise to misleading conclusions in the media: The case of attention deficit hyperactivity disorder. PLoS ONE 6(1): 1–8.

29.

Guenther

Ruhrmann

(2016) Scientific evidence and mass media: Investigating the journalistic intention to represent scientific uncertainty. Public Understanding of Science 25(8): 927–943.

30.

Heazle

(2006) Scientific Uncertainty and the Politics of Whaling. Seattle, WA: University of Washington Press.

31.

Hendriks

Kienhues

Bromme

(2016) Evoking vigilance: Would you (dis)trust a scientist who discusses ethical implications of research in a science blog? Public Understanding of Science 25(8): 992–1008.

32.

Hilgartner

(1990) The dominant view of popularization: Conceptual problems, political uses. Social Studies of Science 20(3): 519–539.

33.

Hoffrage

Gigerenzer

Krauss

Martignon

(2002) Representation facilitates reasoning: What natural frequencies are and what they are not. Cognition 84: 343–352.

34.

International Society for Stem Cell Research (ISSCR) (2016) Guidelines for stem cell research and clinical translation (Report). Available at: http://www.isscr.org/docs/default-source/guidelines/isscr-guidelines-for-stem-cell-research-and-clinical-translation.pdf

35.

Jensen

(2008) Scientific uncertainty in news coverage of cancer research: Effects of hedging on scientists’ and journalists’ credibility. Human Communication Research 34(3): 347–369.

36.

Jensen

Hurley

(2012) Conflicting stories about public scientific controversies: Effects of news convergence and divergence on scientists’ credibility. Public Understanding of Science 21(6): 689–704.

37.

Joslyn

Savelli

(2010) Communicating forecast uncertainty: Public perception of weather forecast uncertainty. Meteorological Applications 17(2): 180–195.

38.

Kahneman

Tversky

(1982) Variants of uncertainty. Cognition 11(2): 143–157.

39.

Kepplinger

(1989) Künstliche Horizonte: Folgen, Darstellung und Akzeptanz von Technik in der Bundesrepublik [Artificial Horizons: Consequences, Presentation and Acceptance of Technology in Germany]. Frankfurt am Main: Campus.

40.

Kienhues

Stadtler

Bromme

(2011) Dealing with conflicting or consistent medical information on the web: When expert information breeds laypersons’ doubts about experts. Learning and Instruction 21(2): 193–204.

41.

Kohl

Kim

Peng

Akin

Koh

Howell

. (2016) The influence of weight-of-evidence strategies on audience perceptions of (un)certainty when media cover contested science. Public Understanding of Science 25(8): 976–991.

42.

Kolata

(2016) We’re so confused: The problems with food and exercise studies. The New York Times, 11 August. Available at: http://www.nytimes.com/2016/08/11/upshot/were-so-confused-the-problems-with-food-and-exercise-studies.html

43.

Kortenkamp

Basten

(2015) Environmental science in the media effects of opposing viewpoints on risk and uncertainty perceptions. Science Communication 37(3): 287–313.

44.

Lehmkuhl

Peters

(2016) Constructing (un-)certainty: An exploration of journalistic decision-making in the reporting of neuroscience. Public Understanding of Science 25(8): 909–926.

45.

Libet

(1985) Unconscious cerebral initiative and the role of conscious will in voluntary action. Behavioral and Brain Sciences 8(4): 529–566.

46.

Lin

S-W

Bier

(2008) A study of expert overconfidence. Reliability Engineering & System Safety 93(5): 711–721.

47.

Løhre

Teigen

(2015) There is a 60% probability, but I am 70% certain: Communicative consequences of external and internal expressions of uncertainty. Thinking & Reasoning 22(4): 369–396.

48.

Mazur

(2004) True Warnings and False Alarms: Evaluating Fears about the Health Risks of Technology, 1948–1971. Washington, DC: Resources for the Future.

49.

Merton

(1957) Social Theory and Social Structure. Glencoe, IL: Free Press.

50.

Morton

Rabinovich

Marshall

Bretschneider

(2011) The future that may (or may not) come: How framing changes responses to uncertainty in climate change communications. Global Environmental Change 21(1): 103–109.

51.

Nelkin

(1992) Controversy: The Politics of Technical Decisions, 3rd edn. Newbury Park, CA: SAGE.

52.

Nosengo

(2012) Italian court finds seismologists guilty of manslaughter: Six scientists and one official face six years in prison over L’Aquila earthquake. Nature News, 22 October. Available at: http://www.nature.com/news/italian-court-finds-seismologists-guilty-of-manslaughter-1.11640

53.

Nowotny

(1982) Experten und ihre Expertise: Zum Verhältnis der Experten zur Öffentlichkeit. Zeitschrift für Wissenschaftsforschung [Experts and their expertise: On the relationship between experts and public] 2(3): 611–617.

54.

On Designing Communication, between the Scientific Community the Public and the Media (2014) acatech-National Academy of Science and Engineering, Union of the German Academies of Science and Humanities. German National Academy of Sciences Leopoldina, (Report). Available at: http://www.leopoldina.org/uploads/tx_leopublication/2014_06_Stellungnahme_WOeM_EN.pdf

55.

Oreskes

Conway

(2010) Merchants of Doubt: How a Handful of Scientists Obscured the Truth on Issues from Tobacco Smoke to Global Warming. New York, NY: Bloomsbury Press.

56.

Peters

(1995) The interaction of journalists and scientific experts: Co-operation and conflict between two professional cultures. Media Culture & Society 17(1): 31–48.

57.

Peters

(2000) The committed are hard to persuade. Recipients’ thoughts during exposure to newspaper and TV stories on genetic engineering and their effect on attitudes. New Genetics and Society 19(3): 365–381.

58.

Peters

Allgaier

Dunwoody

Y-Y

Brossard

Jung

(2013) Medialisierung der Neurowissenschaften: Bedeutung journalistischer Medien für die Wissenschafts-Governance [Medialization of the neurosciences: Relevance of journalistic media for science governance]. In: Grande

Jansen

Jarren

. (eds) Neue Governance der Wissenschaft: Reorganisation—externe Anforderungen—Medialisierung [New Governance of Science: Reorganization–External Demands–Medialization]. Bielefeld: Transcript, pp. 311–335.

59.

Peters

Heinrichs

(2005) Öffentliche Kommunikation über Klimawandel und Sturmflutrisiken: Bedeutungskonstruktion durch Experten, Journalisten und Bürger [Public Communication about Climate Change and Storm Flood Risks: Meaning Construction by Experts, Journalists and Citizens]. Jülich: Forschungszentrum Jülich.

60.

Peters

Brossard

De Cheveigné

Dunwoody

Kallfass

Miller

. (2009) Kontakte biomedizinischer Forscher mit Journalisten und Öffentlichkeit: internationaler Vergleich von Erfahrungen und Einstellungen in Deutschland, Frankreich, Großbritannien, Japan und den USA [Contacts of biomedical researchers with journalists and the public: International comparison of experiences and attitudes in Germany, France, Great Britain, Japan and the USA]. In: Peters

(ed.) Medienorientierung biomedizinischer Forscher im internationalen Vergleich. Die Schnittstelle von Wissenschaft & Journalismus und ihre politische Relevanz [International Comparison of Biomedical Researchers’ Media Orientations. The Interface of Science & Journalism and its Political Relevance]. Jülich: Forschungszentrum Jülich, pp. 45–99.

61.

Post

(2013) Wahrheitskriterien von Journalisten und Wissenschaftlern [Truth Criteria of Journalists and Scientists]. Baden-Baden: Nomos.

62.

Post

(2016) Communicating science in public controversies: Strategic considerations of the German climate scientists. Public Understanding of Science 25(1): 61–70.

63.

Post

Maier

(2016) Stakeholders’ rationales for representing uncertainties of biotechnological research. Public Understanding of Science 25(8): 944–960.

64.

Powell

Dunwoody

Griffin

Neuwirth

(2007) Exploring lay uncertainty about an environmental health risk. Public Understanding of Science 16(3): 323–343.

65.

Rothman

(1990) Journalists, broadcasters, scientific experts and public opinion. Minerva 28(2): 117–133.

66.

Rowan

(1999) Effective explanation of uncertain and complex science. In: Friedman

Dunwoody

Rogers

(eds) Communicating Uncertainty: Media Coverage of New and Controversial Science. Mahwah, NJ: Lawrence Erlbaum, pp. 201–223.

67.

Rühl

(1981) Journalismus und Wissenschaft—Anmerkungen zu ihrem Wirklichkeitsverständnis. Rundfunk und Fernsehen [journalism and science–remarks on their concept of reality] 29(2–3): 211–222.

68.

Ruhrmann

Guenther

Kessler

Milde

(2015) Frames of scientific evidence: How journalists represent the (un)certainty of molecular medicine in science television programs. Public Understanding of Science 24(6): 681–696.

69.

Salomone

Greenberg

Sandman

Sachsman

(1990) A question of quality: How journalists and news sources evaluate coverage of environmental risk. Journal of Communication 40(4): 117–131.

70.

Simmerling

Janich

(2016) Rhetorical functions of a ‘language of uncertainty’ in the mass media. Public Understanding of Science 25(8): 961–975.

71.

Singer

Endreny

(1993) Reporting on Risk: How the Mass Media Portray Accidents, Diseases, Disasters, and Other Hazards. New York, NY: Russell Sage Foundation.

72.

Stocking

(1999) How journalists deal with scientific uncertainty. In: Friedman

Dunwoody

Rogers

(eds) Communicating Uncertainty: Media Coverage of New and Controversial Science. Mahwah, NJ: Lawrence Erlbaum, pp. 23–41.

73.

Stocking

Holstein

(2009) Manufacturing doubt: Journalists’ roles and the construction of ignorance in a scientific controversy. Public Understanding of Science 18(1): 23–42.

74.

Tversky

Kahneman

(1992) Advances in prospect theory: Cumulative representation of uncertainty. Journal of Risk and Uncertainty 5(4): 297–323.

75.

Wehling

(2015) Nichtwissenskulturen—Theoretische Konturen eines neuen Konzepts der Wissenschaftsforschung. In: Wehling

Böschen

(eds) Nichtwissenskulturen und Nichtwissensdiskurse [Cultures of ignorance–theoretical outline of a new concept in the Social Studies of Science]. Über den Umgang mit Nichtwissen in Wissenschaft und Öffentlichkeit [Cultures of Ignorance and Discourses on Ignorance. How Science and the Public Deal with Ignorance]. Baden-Baden: Nomos, pp. 23–66.

76.

Weingart

(2001) Die Stunde der Wahrheit? Zum Verhältnis der Wissenschaft zu Politik, Wirtschaft und Medien in der Wissensgesellschaft [The Hour of Truth? On the Relationship of Science with Politics, Economy and Media in the Knowledge Society]. Weilerswist: Velbrück Wissenschaft.

77.

Weingart

(2012) The lure of the mass media and its repercussions on science. In: Rödder

Franzen

Weingart

(eds) The Sciences’ Media Connection—Public Communication and its Repercussions. Dordrecht: Springer, pp. 17–32.

78.

Wong

(2016) Debate erupts over strange new human species: Skeptic challenges notion that small-brained Homo naledi deliberately disposed of its dead. Scientific American, 8 April. Available at: http://www.scientificamerican.com/article/debate-erupts-over-strange-new-human-species/

79.

Yeo

Liang

Brossard

Rose

Korzekwa

Scheufele

. (2016) The case of #arseniclife: Blogs and Twitter in informal peer review. Public Understanding of Science. Epub ahead of print 26 May. DOI: 10.1177/0963662516649806.

80.

Zehr

(1999) Scientists’ representation of uncertainty. In: Friedman

Dunwoody

Rogers

(eds) Communicating Uncertainty: Media Coverage of New and Controversial Science. Mahwah, NJ: Lawrence Erlbaum, pp. 3–21.