At the same time, this complexity poses a significant challenge to communicating climate science to the broader public, which makes the public discourse on climate science more vulnerable to the opinions of contrarians whose prominence in the media is disproportionate to their representation in the scientific community. Indeed, communicating authoritative information about the risks of inaction is crucial for achieving global action.
Yet, sending uniform and authoritative messages is challenging for various reasons. One reason is that CC communication often requires strategically paring down this wicked problem for non-expert audiences A related problem is the diminishing demand for expertise in scientific discourse aimed at the public These problems are further exacerbated by the proliferation of new media, which democratize the production and consumption of information, making it increasingly challenging to identify trustworthy information Unless countered by improvements to quality control management that can match the production scale, this information deluge is likely to overwhelm the traditional safeguards of professional editorial oversight.
Against this background, we contribute to the CC communication literature on authoritative messengers and new media by analyzing two carefully selected groups of prominent individuals who are frequently sourced in CC media articles. Thus, by juxtaposing media visibility and scientific authority for two counter-positional groups, we are able to objectively measure the discrepancy in CC authority between consensus scientists CCS and contrarians CCC. As such, we selected CCC using open registries that clearly document their contrarian positions.
There are several limitations to our data-driven analysis worth first discussing. First, we do not account for the range of professional backgrounds, nor do we account for the different types of skepticism promoted by different CCC By way of example, recent work comparing fundamental skepticism relating to sources and existence of CC to impact skepticism relating to potential impacts of CC reveals that the frequency of the fundamental skepticism has decreased over time, whereas the frequency of impact skepticism has increased over time, possibly signaling a strategic shift within the contrarian movement While distinguishing visibility according to these two skeptic types could explain some variation in media visibility observed across individuals Fig.
A second limitation relates to the sampling of fixed group sizes of individuals from global populations of contrarians and scientists who differ greatly in their size. As such, the disparity in visibility may be affected in part by there being relatively fewer greater number of contrarians scientists combined with journalists seeking balance 8 , 9 , 11 , Another source of variation that limits the interpretation of our comparison is the composition of the CCC group, which includes business people and politicians in addition to skeptic scientists, thereby reflecting the same drivers of variation underlying the primary frames e.
We addressed this compositional difference in several ways. First, we restricted our analysis to the time period before the US presidential election so that media visibility is more reflective of the scientific rather than the political arena; second, we focused our analysis of scientific authority drawn from peer-reviewed research on the subset of CCCs who did appear in the publication data, and compared them with a size-balanced set of CCSs; and third, we compared the two groups using normalized media visibility measures in order to account for variation in scientific authority between the CCC and CCS Fig.
In this way, we explored alternative explanations for observed discrepancies in visibility and authority. One final limitation relates to how individuals appear in media articles, as we do not distinguish whether individuals are sourced as experts or dismissed as illegitimate authorities Thus, as in the case of positive and negative citations, our measures of media visibility are partially conflated by dismissive mentions.
While the size of the contrarian movement may be relatively small, our study reveals the degree to which new media facilitates the production and mass distribution of assertive content by CCC—which intentionally or not, crowds out the authoritative message of real CCS. However, if we condition the article count tallies using select mainstream media sources, i. As such, we objectively demonstrate the discrepancy in the scientific authority and media visibility between these two sets of prominent CCS and CCC.
In order to provide contextual depth, we also analyzed how these prominent individuals are sourced within media articles—are they just mentioned or do they contribute content via quotes or authorship? Our results point to an additional level of discrepancy—CCCs are more likely than their counterparts to be mentioned or contribute via non-scientific quotes, whereas CCSs are more likely to contribute via scientific-oriented quotes or authorship.
Research shows that journalists often quote contrarians either to infuse objectivity or to dismiss their position outright 9 , Yet, these approaches also detract attention from the relevant CC narrative and provide the counterproductive impression that there is something substantial in contrarian arguments to be debated.
Thus the time has arrived for professional journalists and editors to ameliorate the disproportionate attention given to CCCs by focusing instead on career experts and relevant calls to action. Our individual-centric approach also provides insight into the structural properties of the consensus—contrarian interface.
Even accounting for productivity differences, we find that CCC cite CCS 20 times more frequently than in reverse. In summary, our work contributes to recent computational social science efforts 7 , 34 , 38 , 49 , 56 by leveraging new opportunities in large-scale data collection and analysis 37 to clarify both the individual and collective properties of complex socio-technical systems Related research on the emergence of polarization around critical yet controversial socio-political issues 51 , 58 , the impact of new media on the public 59 , and the spread of inaccurate information 16 , 17 , 43 will together provide important guidance on how to improve the effectiveness of CC communication 14 , 15 , 21 , 32 , 60 , Indeed, despite the challenges posed by new media, there are also new opportunities.
One relevant example that borrows from the post-publication peer-review system in science is the new media tool climatefeedback. See Supplementary Note 2 for additional details on the data collection. We chose to set the upper bound for the MC data collection as October so to avoid confounding the deluge of CC articles related to US elections and subsequent cabinet and other government administration appointments. One clear limitation of our MC dataset is that it is biased toward English-language content We then refined the dataset by applying the following media article disambiguation method.
Upon close inspection of individual MC article metadata, we found that a significant number of articles indexed by MC refer to the same media article. We found this to be particularly common among the select media sources included in our content analysis, in which the same article may have multiple different URLs, representing an array of hyperlinks from different facets of their website—e.
As such, the initial set of , MC articles requires an additional merging procedure in order to avoid overcounting, resulting in a final dataset of , unique MC articles. While two or more MC articles may have unique MC article identifiers, inspection of the URL and article title indicate that they are indeed the same.
We addressed this redundancy by merging articles with the same MC media source and similar title into a single article instance so that article counts for sources and individuals are not systematically inflated. We determine whether two titles are similar by calculating the Damerau—Levenshtein edit distance D jk between the title T j and T k. Titles that are similar have small D jk , meaning that a small number of character insertions, deletions, and swaps can convert T j and T k , or vice versa since it is a symmetric measure.
We compiled a list of contrarians by merging three overlapping name lists obtained from three public sources. Another source of error occurs in the attribution of publications to authors commonly known as the author name disambiguation problem. We used a tested initial-based author name disambiguation approach 45 , which nevertheless tends to overestimate the number of publications for a given author, thereby corresponding to a positive misattribution or clumping error.
See Supplementary Note 1 for additional discussion of our author name disambiguation strategy. Analysis of the individual media sources producing CC content also revealed a wide range of production volume. To account for this variation, in what follows we distinguish media visibility within 30 prominent mainstream sources. Supplementary Table 1 lists additional information about the type, age, MC unique identifier, and number of unique MC articles collected from each media source. We manually located the full-text of each article using the hyperlink provided by MC and then tallied how individuals were sourced according to a typology comprised of five categories.
The first class corresponds to individuals who are mentioned—but not actively involved in the article content. The second corresponds to individuals quoted—and including scientific content.
The third corresponds to individuals quoted—but not including scientific content. The fourth corresponds to a scenario similar to the previous category, in which individuals are quoted—including adversarial content aimed at the counterposition.
And the fifth corresponds to when the individual authored the media article. The first category represents a reference to an individual, often related to a documented event such as a report or conference, that does not include any attributable content.
As such, it is likely that individuals sourced in this way did not actively contribute to the production of the media article, and so these mentions do not necessarily confer any privileged expert status upon the individual.
However, the remaining four categories reflect some level of contribution by the individual—either active in the case that they were interviewed or passive in the case that statements were quoted from external text, e. In the case of quoted categories 2—4 , we do not distinguish between quotes derived from external sources e. We then group sourcing categories 2—5 together into a combined contributed category to facilitate direct comparison with the mentioned category.
This approach follows a framework for evaluating references to the IPCC and non-select contrarians in the media We also chose this schema because it is readily identifiable, generalizable to different media source types, e. All data analyzed here are openly available from Web of Science and the Media Cloud project. Code used to carry out media source analysis in this manuscript is available along with the data in ref.
Reasonable additional requests and questions about code can be directed to A. Editorial Note: This is an update of an editorial note issued on August Readers are alerted that the editors are aware of a number of criticisms related to this work. These criticisms are being considered by the editors. The Supplementary Information for this Article is currently unavailable due to concerns regarding the identification of individuals.
We will publish an update once our investigation is complete. An amendment to this paper has been published and can be accessed via a link at the top of the paper. National Research Council, et al. Oreskes, N. The scientific consensus on climate change. Science , — Cook, J. Quantifying the consensus on anthropogenic global warming in the scientific literature. ADS Google Scholar.
Consensus on consensus: a synthesis of consensus estimates on human-caused global warming. Doran, P. Examining the scientific consensus on climate change. Eos 90 , 22—23 Farrell, J. Network structure and influence of the climate change counter-movement. Change 6 , Boykoff, M. Balance as bias: global warming and the US prestige press. Change 14 , — Google Scholar. Antilla, L. Climate of scepticism: US newspaper coverage of the science of climate change.
Change 15 , — Anderegg, W. Expert credibility in climate change. Koehler, D. PubMed Google Scholar. Bruggemann, M. Beyond false balance: how interpretive journalism shapes media coverage of climate change. Change 42 , 58—67 Lazer, D. The science of fake news. Evidence-based strategies to combat scientific misinformation.
Change 9 , — Iyengar, S. Scientific communication in a post-truth society. Vosoughi, S. The spread of true and false news online. Shao, C. The spread of low-credibility content by social bots. Stella, M. Bots increase exposure to negative and inflammatory content in online social systems. Kahan, D. The polarizing impact of science literacy and numeracy on perceived climate change risks. Change 2 , Guilbeault, D. Social learning and partisan bias in the interpretation of climate trends.
Scheufele, D. Science audiences, misinformation, and fake news. Brulle, R. Change , — McCright, A. Newman, T. Climate change, cultural cognition, and media effects: worldviews drive news selectivity, biased processing, and polarized attitudes.
Public Underst. Climate-science communication and the measurement problem. Political Psychol. Egan, P. Politics 74 , — On the relationship between personal experience, affect and risk perception: The case of climate change. Public enemy no. Understanding media representations of outlier views on climate change. Janko, F. Is climate change controversy good for science? IPCC and contrarian reports in the light of bibliometrics. Scientometrics , — By providing your email, you agree to the Quartz Privacy Policy.
Skip to navigation Skip to content. Discover Membership. Editions Quartz. More from Quartz About Quartz. Follow Quartz. These are some of our most ambitious editorial projects. From our Obsession.
Skip to navigation. The Basics of Climate Science. Separating Fact from Fantasy. Scientific Consensus on Global Warming. While polls of scientists actively working in the filed of climate science indicate strong general agreement that Earth is warming and human activity is a significant factor, 31, scientists say there is "no convincing evidence" that humans can or will cause "catastrophic" heating of the atmosphere.
This claim originates from the Oregon Institute of Science and Medicine , which has an online petition petitionproject. To participate in the petition one only needs to mark a check box to show that one has a Ph. Unfortunately, that means that anyone can sign the petition, whether they have a degree or not. Since the results are not verifiable, there is no way to know how many signers have actually earned a degree. Do '31, scientists say global warming is not real'?
But more importantly what is the significance of these signatures? The majority of signatures are engineers 10, Without formal training in climate science the level of understanding remains unknown among those that signed the petition. A key question is not how many of those that signed the petition know climate physics in sufficient depth, but rather how many of those that signed the petition work directly in the field of climate science.
But there is no indication how many work in the field of climate science? What is notable is that the polls indicate there is a perspective difference between working climate scientists and scientists not working in the field of climate.
When one examines the experts in the field, one sees a significant divergence from the general view.
0コメント