pandemic, nor was the goal of the study to address any particular claim including the origins of the SARS-CoV-2 pandemic. The study also does not focus on issues of vaccine hesitancy or resistance, primarily owing to the significant complexity of the underlying reasons for these views. However, the outcomes of the study may be broadly helpful to scientists seeking to counter misinformation about vaccines.

The primary audience for the study and its outputs, specifically the engagement strategy report and how-to guide, are scientists and scientific institutions from Southeast Asia. Because mis- and disinformation are global problems, scientists from other regions also may benefit from the study outcomes. For this study, the term “scientist” includes laboratory and field life scientists; clinicians (human and veterinary); public health scientists; social scientists; and scientists from a variety of other natural, physical, and computer sciences and organizations (academia, industry, nongovernmental organizations, government laboratories, and community or unconventional laboratories). Potential broader audiences of corrective messages are policymakers, journalists, and lay and religious leaders. Although the primary focus of the study is to develop the trusted network of scientists rather than to refute claims in broader public discourse or in designed informational or behavioral change campaigns, the outcomes provide informative knowledge to regional scientists who may choose to engage policymakers and other non-scientific stakeholders in refuting misinformation.

The regional focus of the study, which is specified in the Statement of Task, allowed the committee to include structural and cultural considerations and challenges of the region, and develop actionable recommendations based on existing regional resources and scientific networks. Although the focus and framing of the report and recommendations are within the lens of Southeast Asia, the scholarship of network design, misinformation, and science communication may inform efforts by scientists in other parts of the world to counter inaccurate and misleading information given the international nature of mis- and disinformation. In addition, international networks may offer certain advantages to engaging scientists with different expertise and accessing resources from international organizations. However, opportunities for involving scientists from Southeast Asia and for building understanding and trust among network members may be greater at the regional level where the cultures and societal structures may be more similar. Therefore, this strategy report and accompanying how-to guide are tailored to provide regionally appropriate and actionable advice, rather than high-level options that are not specific to any region or country in the world.

To address the Statement of Task, the committee produced the following consensus products: (1) this report, which describes the committee’s recommended strategy for developing a trusted network of qualified scientists who address inaccurate and misleading information that might lead to mis- or disinformation; (2) a how-to guide for scientists to determine whether and how to address a claim and, subsequently, how to correct the scientific information fueling the claim; and (3) a relational mapping of selected online collaboration platforms.¹ The strategy report and guide do not provide recommendations that request limitless capacity

___________________

¹ See https://www.nap.edu/catalog/26466.

building, workforce training, and funding for a number of reasons. These types of recommendations are less actionable, do not empower partners, do not help scientists navigate potentially challenging policy and practice, and put the focus on long-term funding as opposed to providing the tools and knowledge needed to implement the engagement strategy and guide. The strategy report and guide highlight actions that could have high impact, focus on local solutions, and build on local resources so they can be implemented feasibly by scientists in Southeast Asia. Given the complexity of mis- and disinformation, many stakeholders and efforts are necessary to prevent and otherwise counter these claims. This study involves one set of stakeholders—scientists—in helping to address mis- and disinformation that are fueled by inaccurate and misleading information. Furthermore, the landscape and risks associated with mis- and disinformation, though not new, are evolving rapidly because of the volume of information available, the availability of new online platforms (e.g., social media) and tools (e.g., artificial intelligence [AI]), and changing pandemic and broader societal context. Therefore, critically analyzing all available (often new) resources for addressing mis- and disinformation for their effectiveness is not feasible. The outcomes of this study are intended to present current scholarship and practical knowledge about how to engage scientists in working collaboratively to address inaccurate and misleading information, which could be provided to broader audiences (e.g., policymakers, journalists, lay and religious leaders, and other members of the public) with evidence-supported, robust scientific information that effectively can counter mis- and disinformation.

This report starts with a detailed description of the problem of compounding threats involving inaccurate and misleading information that leads to or perpetuates mis- or disinformation, and their associated consequences to public health and national security. This discussion is followed by a description of cases where life, social, and/or computer scientists have helped to address these problems and challenges. Subsequently, the strategy for a trusted network of scientists from Southeast Asia is described. The strategy builds on scholarship about scientific networks and network influence, countering misinformation, and understanding uncertainty, which are detailed in the last several sections of the report. Drawing on the scholarship presented in this report, the how-to guide² provides clear instructions for scientists to determine whether and how to address particular claims.

COMPOUNDING THREATS OF FALSE INFORMATION AND INFECTIOUS DISEASES AND OTHER BIOLOGICAL THREATS

Misinformation is defined as the unintentional spread of false or misleading information that is shared by mistake or under the presumption of truth (Sedova 2021). Disinformation is defined as false, misleading, distorted, or isolated factual information that is spread deliberately with the intention to cause harm or damage.

___________________

² See https://www.nap.edu/catalog/26466.

The lines between accidental and deliberate spread of inaccurate information may be blurry. For example, false claims that may have been introduced deliberately may be spread by actors who believe them to be true and do not intend to spread false information. Similarly, actors who intend to cause harm may spread misinformation deliberately, including inaccurate scientific information that fits their purpose. Further distinguishing among misinformation, disinformation, and other related concepts, such as conspiratorial thinking or a state of being uninformed (Scheufele and Krause 2019), is beyond the scope of this report. Furthermore, the actors and approaches involved in addressing targeted disinformation often are different than those involved in correcting inaccurate and misleading information that contribute to misinformation. Scholarship on countering targeted disinformation is extremely limited, but increasingly social, computer, political, and natural scientists are studying the creation, spread, and correction of misinformation. Therefore, this report and associated guide draw on the available scholarship on misinformation to inform the recommended approach for countering inaccurate and misleading information, regardless of the intention behind its spread.

Consequences of the spread of inaccurate, misleading, or even hyped scientific information include loss of trust in the public health system and/or ineffective public health responses during epidemics, international conflict about the source and responsibility of epidemics, or the targeting of individual scientists and research institutions associated with particular claims. This section provides examples of inaccurate and misleading claims about biological threats.

Mis- and Disinformation Involving Infectious Diseases

During the height of the Cold War, the Soviet Union deliberately spread false claims about the United States conducting biological activities with non-peaceful intent, seeking falsely to identify the United States as a violator of the Biological and Toxins Convention, which bans the development, stockpiling, production, acquisition, and retention of microbial or other biological agents or toxins that have no peaceful or prophylactic purposes (Gamberini and Moodie 2020). Such disinformation did not disappear at the end of the Cold War. During the past 15 years, Russia has continued to spread disinformation about U.S. biological investments in Central Asia, and specifically about the activities of the Richard Lugar Center for Public Health in Tbilisi, Georgia, which Russia claims is a secret biological weapons facility run by the United States that could be responsible for the release of viruses (Fidler 2019; Gamberini and Moodie 2020; Isachenkov 2018; Lentzos 2018). During the COVID-19 pandemic, as in previous epidemics, Russia and China have spread disinformation to discredit the United States and several European countries, which already has adversely affected the public health system and the government during the COVID-19 pandemic (Dubow et al. 2021). These claims are spread deliberately, are motivated politically, and have real consequences in international diplomacy and/or response to public health emergencies.

Inaccurate and misleading claims associated with biological events also can undermine response activities in affected areas and discredit the current

international leaders in biosecurity. For example, during the 2014–2016 West Africa Ebola Virus Disease (EVD) outbreak, distrust of health care and international aid workers by West Africans affected by the virus adversely altered disease dynamics and response activities. The consequences of this distrust and skepticism presented significant challenges to infection control, particularly in identifying, isolating, and treating exposed or newly infected individuals.

Although this study did not focus on vaccine hesitancy, the ongoing coronavirus outbreak (2019–present) has revealed significant vulnerabilities from misinformation about the safety and effectiveness of vaccines and natural products in Southeast Asia. In Thailand, public health officials initially provided incorrect information about the two vaccines (Sinovac-CoronaVac and AstraZeneca COVID-19 vaccine) that the country has acquired, stating their efficacies as significantly higher than experimentally supported (Chulalongkorn University 2021; KIN () – Rehabilitation & Homecare n.d.; Mallapaty 2021). Once the incorrect information was discovered, the Ministry of Health made corrections to its materials and communications, but the inaccurate information continues to be accessible online. In Indonesia, numerous claims have been made by non-scientists about the efficacy of natural products in protecting against SARS-CoV-2 infection and the sources of the pandemic (Fitrianingrum 2021; Rahmawati et al. 2021). However, no scientific evidence exists supporting these false claims (Adelayanti 2020) whereas other claims have been corrected (Indonesia COVID-19 Task Force n.d.; Majelis Ulama Indonesia 2021). In Malaysia and Indonesia, false claims about vaccines not being halal (Ministry of Health for Malaysia 2021), specifically that they are made in aborted fetal cells or include cells from pigs (Kassim 2020; Ruzki and Ismail 2021), have adversely affected vaccine acceptance. In addition, translation of information into local languages has contributed to the spread of inaccurate information. In other parts of the world, claims spread about the SARS-CoV-2 vaccine include the following: (1) a means for covert implantation of a microchip, (2) the cause of sterility in women, and (3) a scientific hoax (Goodman and Carmichael 2020; Henley and McIntyre 2020; Imhoff and Lamberty 2020; Mayo Clinic Health System 2021).

Another significant focus of disinformation claims during the SARS-CoV-2 pandemic is related to its origin, which has yet to be determined. Numerous articles have been published attributing the introduction of SARS-CoV-2 into humans to deliberate release as a biological weapon by China, the United States, Israel, or the United Kingdom (Gertz 2020; MEMRI 2020; Pamuk and Brunnstrom 2020; Rogin 2020; Stevenson 2020). These claims can affect diplomatic efforts among countries adversely, expose researchers to various personal and security harms (e.g., death threats), and/or limit efforts to enhance safety and security of field and laboratory-based research of emerging infectious diseases.

The volume and variability of scientific information communicated have led the World Health Organization (WHO) to coin the term “infodemics,” which it defines as “a tsunami of information—some accurate, some not—that spreads alongside an epidemic” (WHO 2020). Limited to no evidence exists that suggests the problem with infodemics has gotten worse over time (Simon and Camargo 2021). Limited evidence also exists suggesting that access to more

accurate information is connected to more informed choices by citizens and policymakers. For example, a recent U.S. National Academies’ report stated that science alone rarely changes attitudes and behaviors regardless of whether the audience understands the correct information or suggests their minds would be changed with new information (NASEM 2020). Nevertheless, WHO has maintained the idea that unique challenges exist in addressing the COVID-19 pandemic because of the volume and variable quality and accuracy of available scientific information (WHO 2020, n.d.). According to WHO, these challenges are exacerbated during emergency situations when little information is available, public health and/or security questions are unanswered, and decisions for response are made quickly.

Scientific Information Overload: A New Threat Landscape

Digital communication technologies, in particular the internet and social media, profoundly have changed the way scientific information is produced, consumed, and disseminated and have contributed to access to and spread of mis- and disinformation (Okeleke and Robinson 2021). In the 1990s, scientists were among the earliest to adopt internet technologies primarily to communicate with their peers. Today, scientists can choose to disseminate their work to many different audiences and at many different stages of the scientific process. In addition, scientists now can share data, crowdsource peer review, conduct analyses, request funding (i.e., crowdsourced funding), and establish and maintain collaborations online.

These developments have been complicated by changing norms within the scientific community related to promoting one’s own work and sharing data and results prior to peer review. The choice about whether to use online channels for communication may be influenced by the stage of a scientist’s research and the intended audience of communications about their research. Prior to writing a new scientific article, email, listservs, and various online platforms can be used to connect with colleagues to exchange comments, ideas, and data. After completion of a manuscript but prior to its peer review, preprint servers can be used to disseminate early manuscript drafts to other scientists outside of one’s direct collaboration circles. Two primary preprint servers exist for biological research, bioRxiv and medRxiv, both of which add disclaimers about whether individual papers have been peer reviewed and published. Both preprint servers have internal processes for checking for plagiarism, completeness, non-scientific content, and biosecurity risks. This review process involves trying to identify papers that could cause harm, such as those countering established scientific knowledge (e.g., claims such as “smoking does not cause cancer”). During the past few years, the number of submissions and downloads to bioRxiv has significantly increased, and in 2017 only two-thirds of the papers submitted to bioRxiv were published in peer-reviewed journals (Abdill and Blekhman 2019). As demonstrated during the COVID-19 pandemic, the significant number of papers available to scientists and non-scientists alike, many of which have not been peer reviewed, provide ample opportunities for any inaccurate or misleading

information to be identified and used or shared by interested individuals. Blogs and social media can be used to advertise a newly published manuscript to the broader scientific community or even the public (Yeo et al. 2017).

With so many channels and opportunities for dissemination, risks may exist from an overabundance of information that cannot be assessed effectively for quality and accuracy before broad dissemination. Information overload leads to new threats to the integrity of science and the scientific workflow. Sometimes, errors may occur from misuse of a channel. For example, during the SARSCoV-2 pandemic, individuals, including scientists, discussed the virus, illnesses, and other health- and science-related issues on social media platforms, which included sharing unconfirmed and misleading information (Cinelli et al. 2020). As another example, inaccurate information posted on preprint servers has been amplified by the press or by social media influencers before they could be vetted as part of the regular peer-review process. Other times, errors may occur even as part of the “regular” process of scientific dissemination—for example, when correct, peer-reviewed information is distorted or exaggerated in institutional press releases (Li et al. 2017). As social media increasingly becomes a “hype machine,” developing strategies to prevent dissemination of biased and distorted claims will become crucial (West and Bergstrom 2021).

Types of Scientific Inaccuracies and Misinformation

Producing accurate scientific information in crisis situations is more difficult than in non-crisis situations. Often, a limited amount of information is available, data are sparing, and the situation is evolving rapidly in crisis situations. At the same time, crises demand urgent policy actions, which rely on emerging science, often conducted under immense public scrutiny. Still, scientists have been able to lend their expertise during crises in positive ways. However, the SARS-CoV-2 pandemic has presented a cautious tale of generating and/or sharing sound and evidence-supported information, particularly when the outbreak-causing virus is not well understood. The volume of information produced and communicated through peer-reviewed journals, non-peer-reviewed preprint servers, predatory journals, online platforms and websites, and social media create significant challenges for scientists and other consumers of scientific information to evaluate the accuracy and quality of publicly shared scientific information. Articles that are poorly conducted, involve small sample sizes, and/or demonstrate other significant weaknesses have fueled misinterpretation and misinformation by the media or other groups with particular agendas. For example, the results of a 2020 Danish study on the effectiveness of masks for protecting against SARS-CoV-2 infection were interpreted by the media as “questioning the efficacy of masks” and used by antimasking campaigns to promote its agenda (O’Grady 2021). Because evidence-informed decision-making involves acquiring and aggregating available information, inaccuracies resulting from weak study design, low sample size, and poor data analysis also run the risk of developing ineffective and misinformed policy and public health actions.

infections and effective control measures, and development of new vaccines and medicines among many other activities. For example, the 2011 Escherichia coli O104:H4 outbreak initially was characterized using bioinformatics analyses of bacterial genomic sequences following an unusual infection pattern—24 infections in France and more than 3,816 cases in Germany (Guy et al. 2012; IOM 2012). The sequence analysis enabled food safety experts to use epidemiological methods to track the source, which was contaminated sprout seeds (Grad et al. 2012), but not until after the outbreak was correlated inaccurately with lettuce, cucumbers, and tomatoes (Foley et al. 2013). These analyses, which were conducted relatively quickly after initial infections were documented, enabled public health officials to prevent further spread of the bacteria and addressed any concern about the deliberate nature of the outbreak, which resulted from the unusual infection patterns. Scientists from other disciplines, including various social science disciplines and computer science, have enabled effective control of biological incidents. The 2014–2016 EVD outbreak in West Africa illustrates the important role of social and computer scientists. Cultural anthropologists shed light on why white protective suits and removal of ill or deceased individuals caused distrust, leading health agencies to change the color of the suit (because white signifies death within the local culture) and develop safe practices for traditional burials (Nyhan 2014). These changes allowed health workers to gain the trust they needed among the local population to control the spread of EVD. Similarly, advances in big data analysis and modeling of infections informed international awareness about and response to the outbreak (Nieddu et al. 2017; Wein 2014). During the SARS-CoV-2 pandemic, computer scientists also bring to infectious disease outbreaks their knowledge of information transfer via social media and other online platforms, and advanced data analytics such as AI for quickly identifying effective existing therapies and near-real-time surveillance during outbreaks (Basu 2021; Begley 2021; Bridgman et al. 2021; Budd et al. 2020; Himelein-Wachowiak et al. 2021; Kostkova et al. 2021; Mohsin et al. 2020; Zhou et al. 2020).

Conclusion 1: Life, social, and computer scientists play critical roles in ensuring their science and the science produced and shared by others within their communities is accurate and supported by well-designed and implemented studies. They can work collaboratively to leverage appropriate expertise to produce and/or disseminate accurate scientific information, peer review and correct inaccurate scientific information, and ensure that scientific information is communicated in an unbiased, objective, and culturally informed manner.

Conclusion 2: Scientists are one of several stakeholders in addressing misinformation. Other stakeholders include policymakers, journalists, and members of the public. Engaging these other stakeholders positively is critical to building trust, communicating corrective information clearly and effectively, and addressing misinformation in a timely manner if and when necessary.

(Horowitz 2021; Meta AI 2020). However, these tools have demonstrated challenges in identifying misinformation related to the COVID-19 pandemic (Siriwardana n.d.). Therefore, understanding the capabilities and limitations of AI, especially in helping to identify false claims, is critical if national and/or nongovernmental entities plan to use AI as part of their fact-checking efforts and publishers and scientists plan to use AI to identify relevant articles for review and analysis. AI refers to the ability of computers to perform tasks that normally require human intelligence and has been studied for decades, primarily focusing on routine software to codify knowledge of human experts in specifically programmed rules (i.e., “if given a specific input, then provide a specific output”) (Sedova 2021). Recent interest in AI focuses mostly on machine learning, a type of AI that is driven by data abundance, innovation in algorithms, and computing power. Machine learning systems learn from data rather than human expertise and involve the development of a model based on a training dataset and defined algorithm architecture that can compensate for bias in the data. Four types of machine learning exist: (1) supervised learning that uses example data that have been curated and labeled by humans; (2) unsupervised learning that uses data that do not require labels but rather identifies patterns in the data, clusters similar groups of data, and can detect new behaviors from previously unidentified patterns; (3) semi-supervised learning that uses labeled and unlabeled data; and (4) reinforcement learning that uses an autonomous AI agent that gathers its own data as it performs goal-oriented tasks to maximize rewards in a learning environment and improve based on trial and error in fictitious or real-world situations.

Conclusion 3: Technical solutions alone are not sufficient to identifying and addressing inaccurate and misleading claims. Scientists from a diversity of disciplines are necessary to address inaccurate and misleading claims about biological threats. What is needed is the establishment of a structure for human interactions, favorable policies for data sharing and analysis, and a process for collaboration to correct inaccurate and misleading information before it has the potential to fuel misinformation.

STRATEGY OF THE TRUSTED NETWORK OF SCIENTISTS

A strategy for engaging scientists in addressing inaccurate and misleading information builds on the role that scientists play in curbing the spread of misinformation of biological threats, benefits from leveraging local and international networks in a dynamic and distributed network, and enables scientists to interact on an ongoing basis to improve scientific accuracy and excellence. A distributed network is a network of networks that enables active involvement of scientists from many disciplines and institutions to assist with correcting inaccurate and misleading information, and draws on existing discipline- or sector-specific networks of scientists. The underlying scholarship and regional considerations for the recommended strategy and considerations for scientists

involved in countering misinformation are described after the description of the recommended strategy.

Recommendation 1: Leaders of established scientific networks in Southeast Asia jointly should create a distributed network of individuals and organizations (i.e., a network of networks) that draws on a diversity of scientific disciplines and sectors needed to correct inaccurate and misleading scientific information about infectious diseases and other biological threats. The network should be regional and have a leadership structure that includes scientists from countries in the regional network. The network itself should be virtual only, leveraging recently developed online collaboration tools, but should be based in a host nation within Southeast Asia to support key operations (e.g., website, email addresses, and resource repositories) and gain credibility by regional and national authorities.

The following sub-recommendations define the scope and boundaries of the network.

Recommendation 1.1: The network should establish (1) a governing board, comprised of knowledgeable scientists who are responsible for making strategic decisions for the network; and (2) an executive team, comprised of scientists who have a proven track record of developing and sustaining scientific networks. The executive team should manage, motivate, and expand the membership; foster diversity among the membership (i.e., gender, age, experience level, country, ethnicity, scientific discipline, type of institution); ensure continuous services are provided to members; and activate the network to address inaccurate and misleading information about biological threats when such claims arise. The executive team should work with individual members and external stakeholders to identify claims and determine which claims to address before activating relevant expert members of the network.

Recommendation 1.2: The network should include scientists from all relevant disciplines—specifically a variety of life and other natural sciences, computer and information science, social sciences including science communication experts, and political science including researchers of misinformation—to ensure that the most relevant expertise is leveraged in addressing inaccurate and misleading information and can be communicated effectively to the appropriate stakeholders, whether within or outside the network. Furthermore, the network should include scientists and country-level networks from academia, industry, and other nongovernmental and government organizations.

to (1) empower talent; (2) motivate scientists; (3) engage international, regional, and national efforts; and (4) leverage existing initiatives in responsible science.

Empowering Local and Regional Talent

A critical element for the trusted network is having a diversity of scientific experts who work collaboratively to address inaccurate and misleading information. Many countries in Southeast Asia have invested in developing scientific talent and scientific experts in various biological sciences disciplines. Current measures of scientific output (e.g., number of graduates in specific disciplines and number of papers published) do not characterize local scientific capacity adequately and may present difficulties in generating or recruiting future talent in various scientific fields. Many countries in Southeast Asia have science, technology, and innovation talent development roadmaps, but many do not include similar initiatives to support the development of social scientists because the social sciences are perceived as not being associated with economic gain from innovation (Scott-Kemmis et al. 2021). The pandemic has revealed a significant lack of a robust community of social science experts capable of supporting the laboratory or field-based natural science community in Southeast Asia to address

a variety of societal problems, including capabilities to address inaccurate and misleading claims effectively.

In addition, equipping life and computer scientists from Southeast Asia with science communication and stakeholder engagement skills can empower them to address inaccurate and misleading information. Furthermore, providing skills to foster inclusive leadership and promote trust and sustainability within scientific networks can enable cross-disciplinary, cross-sector, regional collaboration to address these claims. Identifying the talent gaps in specific scientific disciplines in Southeast Asia is critical to ensure local relevance and sufficiency. Furthermore, an inclusive network of scientists can enable all groups to work toward a common goal of addressing misinformation about emerging infectious diseases and biological threats by leveraging each other’s expertise and diverse perspectives.

Finding 2: Training scientists on science communication, public engagement, and science policy is critical to countering misinformation.

Finding 3: In Southeast Asia, research and education in these and other social science disciplines are limited. But they are necessary for ensuring that cross-disciplinary scientific teams have all relevant skills and expertise available when addressing inaccurate information that could lead to or perpetuate misinformation.

Recommendation 2: National competent authorities, whether the Ministries of Science or other Ministries, in Southeast Asian countries should develop undergraduate, graduate, or postgraduate training programs on various social science disciplines, including science communication and public engagement. In addition, national competent authorities should develop informal training programs in science communication and public engagement for life, computer, and other natural scientists and engineers to enhance their communication skills and their ability to build public trust. Countries that already have established education or training programs should share their curricula, best practices, and lessons learned.

Motivating Service for the Purpose of the Network

Another critical element of the trusted network is motivating local scientists to contribute actively in promoting scientific excellence and integrity, building the defensible and accurate scientific knowledge base to correct inaccurate and misleading information, and identifying and addressing misinformation. Examples of ways that scientists may be motivated to contribute to the network include the following: (1) appealing to concepts of scientific responsibility, associating the practice of countering inaccurate information with scientific integrity and excellence; (2) highlighting the potential consequences of misinformation

such as the erosion of trust in science and scientists by members of the public and negative effects on the research and public health ecosystem; (3) recognizing scientists for their contributions to countering inaccuracies; and (4) providing financial support for time spent addressing inaccurate information.

Intrinsic motivation-building strategies include promotion of inclusivity and an understanding of the local culture, trust, and openness. Teams involving Southeast Asian scientists need to be created in a manner that allows for inclusion and productive contribution of scientists who are early in their careers and are from a variety of disciplines and sectors. Promoting trust through a culture of transparency and openness is critical to building a trusted network both for internal interactions among members and external interactions among various stakeholders (e.g., scientists, members of the public, policymakers, and journalists). This transparency is a critical factor in addressing effectively the growing mistrust, skepticism, and various other challenges associated with inaccurate and misleading information that fuels misinformation by supporting greater accountability, quality, and safety, and by promoting ethical attitudes and practices, all of which contribute to scientific excellence. The value of transparency in fostering trust is highly dependent on the external and internal communication processes that are mediated by several relational and contextual factors.

Motivating the scientific community to engage in efforts such as addressing misinformation involves social responsibility and integrity. For example, the Young Scientists Network-Academy of Sciences Malaysia and the Association of Southeast Asian Nations-Young Scientists Network (ASEAN-YSN) promote Responsible Conduct of Research to encourage greater awareness among scientists in upholding the principles of research integrity and addressing contemporary needs of various external stakeholders in Southeast Asia. These existing networks promote multi-sector engagement among academia, industry, government, and broader society and embrace responsible team science among interdisciplinary teams of scientists and others within the scientific ecosystem. Through these efforts, researchers in the region are encouraged to reflect on their own actions to ensure they align with ethical principles associated with correcting scientific errors and addressing misinformation.

Finally, when misinformation involves scientific issues, scientists may be concerned about the presumed harm that these claims have on the reputation of other scientists and the public, which in turn can activate scientists’ support for education and legislation for correcting misinformation. These considerations can become key enablers in encouraging domain experts, such as life, computer, and social scientists, to work together to safeguard the interests of the scientific community and the broader society.

As scientists become motivated to contribute to the network, enhance scientific excellence, and address misinformation, continuous engagement and involvement in the network can be promoted. Continuing engagement can be measured via the ability of scientists to receive continuous professional development opportunities for improving their skills in addressing inaccurate and misleading claims, and enhancement of existing and innovative capacities for addressing inaccurate information at institutional and national levels.

new initiatives and programs, often have a greater cultural understanding of issues and regional context, and may receive financial and political support by regional (e.g., ASEAN) or national authorities. Examples of country-level societies are national academies and national young academies. These networks may be connected to regional and international networks. Networks of scientists also could be formed on a disciplinary- or field-specific basis such as the American Society of Microbiology, the International Society for Infectious Diseases’ ProMed, the American Biological Safety Association International, the International Life Science Institute, the International Union of Biological Sciences, the International Union of Biochemistry and Molecular Biology, the International Network for Government Science Advice, and the Southeast Asia One Health University Network and national one health university networks in Southeast Asian countries.

Trust is built on scientific excellence, clear and concise communications, and active and respectful dialogue with various stakeholders. Although the need for addressing misinformation about biological threats has been recognized by several international organizations (e.g., WHO, the United Nations Interregional Crime and Justice Research Institute), prioritizing regional and national efforts that involve scientists as part of the solution toward addressing inaccurate and misleading information that leads to or otherwise enhances misinformation would provide high-level support for the trusted network and other related activities. For example, if ASEAN prioritizes initiatives for addressing misinformation, members of the trusted network could provide accurate, authoritative, and defensible scientific information that could debunk those claims and also could provide high-level support for the trusted network. In The State of Southeast Asia: 2021 Survey Report, published by the ASEAN Studies Centre at the Institute of Southeast Asian Studies-Yusof Ishak Institute, respondents (from academia, think-tanks or research institutions, government entities, civil society organizations, nongovernmental organizations, business or finance institutions, and regional or international organizations) from ASEAN, in particular Singapore and Brunei, urged scientists and medical doctors to engage in public policy discussions and scientific advice to better address the SARS-CoV-2 pandemic (Seah et al. 2021). Similarly, partnerships between this trusted network and respected international organizations such as the International Science Council, IAP, and national academies (e.g., the U.S. National Academies) further enhance the network’s ability to draw on scientific leaders in a variety of disciplines, including those not yet well represented in Southeast Asia.

Recommendation 3: The distributed network (see Recommendation 1) should leverage existing scientific networks in Southeast Asia, catalyzing collaboration among their members, especially those with needed expertise, to address specific claims. Regional scientists involved in developing this distributed network should develop open and trusted lines of communication and partnership with relevant Ministries to gain high-level support for the network and its outputs.

gain support among the scientific and policy communities in Southeast Asia, adding value to members on an ongoing basis (e.g., through training of new skills), developing and executing the process for adjudicating claims that need to be addressed by the network and for enabling cross-disciplinary and cross-sector collaboration to address those claims, and identifying and implementing improvements of the distributed network. Furthermore, the implementation plan can specify the means and frequency of interaction among network members; the means of interaction with other networks and regional leaders; the process for developing common messages about corrected information that are accessible in multiple languages; and a unified platform through which members and broader stakeholders can communicate and access authoritative, evidence-informed, and defensible scientific information. Finally, the implementation plan can include the development of key operational actions, including (1) a process for maintaining connectivity and interaction among members, stakeholders, and other networks; (2) an action plan for sustaining the network beyond initial funding (i.e., through the development of a fundraising or business plan that identifies potential funders, articulates the value proposition to potential funders, and documents measures of success of the network); (3) a clear definition of the membership types, levels, and costs, if relevant; (4) incentives for member and stakeholder engagement; and (5) consideration of the nation that hosts the network, even if implemented as virtual-only, and associated policies for operating the recommended network.

Recommendation 4: The governing body and executive team of the distributed network (Recommendation 1.1) should establish an implementation plan for the creation of the distributed network prior to its formation. This plan should specify important aspects of the network’s operation such as (1) sustainability of the network once established; (2) interaction with scientific and non-technical stakeholders; (3) ongoing benefits to members (e.g., training, mentorship and coaching opportunities, professional networking, production of publications); and (4) processes for identifying, adjudicating, and, if deemed necessary, addressing inaccurate information relating to biological threats. The plan should build on the key considerations of the strategy, lessons learned about effective correction of inaccurate information that are described in this report (section on Interventions Against Inaccurate and Misleading Information), and the associated how-to guide for identifying and potentially addressing inaccurate information that leads to or perpetuates misinformation. Furthermore, the plan should define the role that scientists play in addressing inaccurate and misleading information and scientific responsibility. This role should be considered within the broader context of other societal and international actors who may be addressing mis- and disinformation.

Recommendation 5: The implementation plan should specify how the network will interact with stakeholders in government, media, and the broader

public to establish itself as a “go-to” regional focal point for addressing inaccurate and misleading information with evidence-supported and clearly communicated scientific information. The implementation plan also should describe the relationships among the network’s leadership, members, external scientists, and other external stakeholders to provide a clear understanding of who makes inquiries about addressing inaccurate statements and how these statements will be evaluated and addressed.

Recommendation 6: The implementation plan should draw on the guide, which is a companion to this strategy report, for addressing inaccurate and misleading statements. Network leaders and members should work with experts in the life sciences, science communication, team science, computer science, and research studying the flow, spread, and correction of misinformation to train network members in using the guide effectively.

NETWORK STRUCTURES CONSIDERED IN THE STUDY

Different approaches to building a network to combat misinformation about biological threats were considered. One approach is a network that involves formal or informal associations of individuals as members. Another approach is a consortium that involves formal association of institutions as members. In this section, both models, practicalities of their creation and maintenance, and limitations and capabilities of each are described along with common considerations that apply to any scientific network model.

Selecting which model is appropriate involves analysis of several key considerations including sustainability, which is a significant concern associated with the longevity and durability of any network, and the ability to be dynamic. Continuous engagement of members in the network’s primary activities and emphasis on the benefits of membership to the network promote its sustainability. Dynamic networks require more active participation of members and specialized domain expertise, especially for events that occur periodically. For example, although misinformation is a constant concern, misinformation about infectious disease research and epidemics may occur periodically and only when specific concerns or events emerge. Therefore, designing the network such that it can activate different members based on changing needs is important for addressing inaccurate information about emerging infectious diseases and other biothreats. Both models described here incorporate these considerations in network design.

Trust is a critical component of any type of network. Creating a stable and trusted network involves the provision of opportunities for affiliation and social exchange, development and mentoring, and support for strategic collective action. Collective learning and collaboration can enable a community to construct moral authority. Known as collective intelligence, both networks of individuals and institutions can foster innovation more effectively to solve highly complex challenges and marshal resources and knowledge fast (Riedl et al. 2021; Suran et al. 2020).

which already have stable membership of individuals and trust within their communities. A particularly effective example of a formal network of organizations is the Societies Consortium on Sexual Harassment in STEMM, based on recommendations by the U.S. National Academies (NASEM 2018). The mission of this Societies Consortium is to “support academic and professional disciplinary societies in fulfilling their mission-driven roles as standard bearers and standard setters for excellence in science, technology, engineering, mathematics, and medical (STEMM) fields, addressing sexual harassment in all its forms and intersectionalities.” As an organization, the consortium sought to create a strong and influential, collective voice, “backed by action, of a diverse and large group of committed societies (of many sizes and areas of focus) to set standards of excellence in STEMM fields.” In short, the Societies Consortium was created to bring together scientific networks to address a challenge in science and society, and to help create and provide resources and tools for researchers to increase ethical behavior in science, specifically on sexual harassment (Societies Consortium on Sexual Harassment in STEMM n.d.).

The benefit of a consortium of societies (or institutions) is in the strength of a collective and unified voice that can more easily and rapidly unite thousands of scientists. Member societies can get resources and information out rapidly if needed to address immediate, external issues. The focus on the consortium can be more about creating useful resources and leadership guidance on addressing issues, setting standards for ethics in science, or providing other common resources, rather than evolving more slowly as a society of individual members in specific scientific disciplines. Another benefit of a consortium is the potential sustainability of the network. A consortium creating resources and tools for societies that can then be passed down as resources for individual smaller network members is creating value for all through the shared cost and the non-duplication of efforts. Finally, consortia may provide institutional protection to members who contribute to addressing inaccurate information, helping to protect individuals’ privacy and security for particularly sensitive topics.

The limitation of a consortium is that individuals can be lost in the structure. Although the internal personnel may be structured similarly to an association or institution with an executive board and leadership group, the individual voices of members might be more difficult to hear, as membership to the consortium is through a smaller number of representatives of other networks and/or institutions. This structure may have more of a singular leadership role or source of tools rather than a society made up of individual members. Although the benefits of a consortium trickle down to individual society members, the resources and tools developed are vetted by the consortium leadership and developed to address the specific challenge.

Hybrid of Consortium and Individual Models

The committee also considered a hybrid model in which a network of societies (e.g., consortium) also invites individual membership. Such a model would have many of the same benefits as a consortium. An additional benefit

is that individuals not represented by a society or institution are able to participate. However, the institutional verification and validation of members would be diminished in the hybrid model. Although connectivity is of central importance to connective knowledge, connectivity also can come from individual champions or influencers. Rarely will a scientist not belong to either an association or a research institution, so valuable resources of a consortium would be available for their use. Some scientists and fields are impacted more by mis- and disinformation than others, and could use the support and resources straight from the source.

Another advantage of a hybrid model is that it could take advantage of existing social ties among scientists participating in the same societies or with shared institutional affiliation. Membership in and influence of such a hybrid network would thus spread over the existing “social network” of institutional affiliations, taking full advantage of indirect and direct network effects of this underlying network, in a way reminiscent of how social networks like Facebook took advantage of existing ties among college students in its early growth (Sundararajan 2006).

REGIONAL CONSIDERATIONS

Existing regional networks in Southeast Asia already may have resources to support the translation of scientific information into multiple local languages, communication and interaction among scientists from different countries, access to information and expertise in challenging environments, and planning around important cultural traditions and events. Numerous local languages are used by scientists throughout Southeast Asia. For example, Malaysia is a multiracial country with many languages and dialects (e.g., Mandarin, Malay, Tamil, Bjaus, Iban, Cantonese, and Hakka). Building trust and open lines of communication with scientists and local communities requires the use of local languages for communicating science in crises.

As with many regions around the world, each Southeast Asian country has different laws governing scientific activities and data sharing and access, which are helpful to know when determining whether and how to counter inaccurate information. At the heart of some of these policies are different perspectives on trust, respect, and reciprocal benefit of regional countries and scientists and international scientists (Merson et al. 2015). In 2019, Indonesia strengthened its laws governing access to its biodiversity samples by international scientists, nearly banning export of physical or digital information about samples without a material transfer agreement and increasing the legal penalties of violating these requirements (Rochmyaningsih 2019). Simultaneously, Indonesia began supporting open access to public health articles, an action that accelerated during the COVID-19 pandemic (IAKMI n.d.; Universitas Airlangga n.d.). In Malaysia, legislation related to conservation of biological diversity is mainly sector-based and often is implemented by federal or state level entities (sometimes, implementation is shared by federal and state entities) (MyBIS 2015). Malaysia supports open access to scientific data and information, primarily to reap the benefits of research (MOSP n.d.). Since 1992, Thailand has implemented relevant policies and measures to comply with the Convention on Biological Diversity (CBD)

(Anti-Fake News Center Thailand n.d.; Jalli 2020; Office of Environmental Policy and Planning 2000; Schuldt 2021; Tanasugarn et al. 2000). Vietnam has implemented its Law on Biodiversity and strengthened its legal structures for the management of access to genetic resources and sharing of benefits arising from their use to meet its obligations under the CBD and associated Nagoya Protocol on Access to Genetic Resources and Fair and Equitable Sharing of the Benefits Arising from their Utilization (Socialist Republic of Vietnam 2008, 2017, 2019, 2020a,b). Several countries, including Vietnam, have developed open or limited access to digital sequence information (DSI) to promote scientific pursuits, while the countries party to the CBD are exploring whether DSI should be included in the Nagoya Protocol (Strategic Framework for Capacity-building 2017).

Several Southeast Asian countries also have established laws and partnerships to address misinformation that emerged during the COVID-19 pandemic (Jalli 2020). The Indonesian government works with fact-checking organizations to address misinformation, and the Indonesian Anti-Defamation Society (Masyarakat Anti Fitnah Indonesia) was created before the pandemic to counter hoaxes and other false information (MAFINDO n.d.). The Malaysian government created a platform to cross-check information spread through social media, and Malaysia’s Ministry of Health Malaysia (2021) launched a new website, COVIDNOW, for more transparent COVID-19 data sharing with the public to increase public trust. Thailand created an Anti-Fake News Center and Singapore created Factually,³ both of which are designed to correct false information spread through the media (Anti-Fake News Center Thailand 2021; Schuldt 2021).

Finding 6: Many countries in Southeast Asia have established programs to address false claims, but few academic articles have been published to characterize the scale, scope, and enablers of mis- and disinformation in the region. In addition, studies evaluating effective measures to counter false claims in Southeast Asia have not been published.

Recommendation 7: The executive team of the network should develop resources for scientists to understand country-specific laws governing data sharing, scientific research, and combating misinformation. Furthermore, information describing similarities and differences in trust among scientists and with other stakeholders, respect among scientists, and mutual benefit of scientific activities should be provided to network members to enable productive and positive collaboration.

Recommendation 8: The governance body and executive team of the proposed network should begin building trusted connections with country-specific programs aimed at combating misinformation and in local languages to assist in identifying and communicating corrective information.

___________________

³ See https://www.gov.sg/factually.

beyond scientific information alone including past experiences, values, beliefs, and other associations (von Winterfeldt 2013). All models will have their merits and limitations, so evaluating the circumstance in which scientific information is being communicated and the audiences to which information is communicated is critical for determining which models are going to be most effective.

Recommendation 9: When communicating complex, technical information, scientists and science communicators should use communication models that meaningfully connect new information to audiences’ existing cognitive schemas or mental models.

Creating Counter-Misinformation Messages

When creating messages to counter misinformation among diverse audiences, several actions are critical for consideration (Goldstein et al. 2021; The Royal Society 2022):

1. Expertise in a variety of disciplines is needed to address misinformation.
2. Use of language and context that is relatable can be used to communicate and connect with diverse audience(s).

3. Developing these connections, demonstrating empathy, actively listening, demonstrating integrity, and avoiding hype can help build trust with various audiences, which is critical for countering misinformation where scientific opinion is shared prolifically and not necessarily consistent.
4. Scientists need to listen and be attentive to their audience’s responses and underlying views when developing corrective messages. In addition, learning about audience perspectives, context, and values also helps to develop messages and use language and images that resonate with and are tailored to specific audiences.
5. When correcting inaccuracies, scientists need to provide the corrective messages clearly and with audience-appropriate context, information, and language at the start of any communication.

Communicating Uncertainty

During emerging infectious disease outbreaks or other incidents involving biological threats, the scientific foundation may be inconclusive or in flux because of lack of available data, availability of relevant and verifiable scientific literature to the particular events, the newness of the events, and/or new information being produced as events unfold over time. These uncertainties present significant challenges to individual and national-level policymaking. Policy decisions may require scientific input, regardless of how settled a particular area of science might be at the time (Scheufele et al. 2020), and scientists might be faced with the dilemma of wanting to combat claims that they believe to be incorrect with science that itself might turn out to be incorrect or evolving (Krause et al. 2022). Uncertainty is described in greater detail later in this report.

Media and Public Engagement Training

Basic training in media and public engagement may include the following: (1) approaches for implementing public engagement activities, (2) potential pitfalls in conducting public engagement, (3) approaches for resolving these issues, (4) public speaking, (5) science communication writing (e.g., editorials and newspaper articles), (6) use of new media platforms, and (7) media landscape and approaches for collaborating with media practitioners. This training may include guidance on navigating the political landscape while correcting misinformation, which can help to address any concerns or fears of political repercussions when engaging various audiences. Scientists may avoid public engagement if they are concerned about potential detrimental effects on their reputations and/or careers (Ho et al. 2020a,b). To limit these consequences, institutions can develop and provide clear rules and guidelines to help scientists navigate the intricacies and sensitivities in their cultural and political contexts. Furthermore, institutions can support and provide recognition to scientists who seek to address inaccurate and misleading claims (Ho et al. 2020c).

decisions, such as closure of schools and/or businesses (or lockdowns) and other practices (e.g., mask and quarantine mandates), divided support among the broader public, which may fall on partisan lines (Cauchemez et al. 2014; CIDRAP 2009a,b; Davis et al. 2015; Harvard T.H. Chan School of Public Health 2009; Krupa 2009; McNeil 2009; Pallarito 2009; Toffelmire n.d.). Science derives its cultural authority as society’s creator and curator of knowledge from the perception that it operates systematically, objectively, and free from partisan politics (Brossard and Nisbet 2007). Therefore, individual scientists, who may be domain experts in relevant fields, may choose to work with science policy experts to translate existing and objective scientific knowledge to inform policy and practice. However, avoiding endorsement of particular policies without all available information in hand may create new, unanticipated problems.

Should and How Should Scientists Engage the Public?

Engaging various public and policy audiences is a two-way street, with scientists understanding the key questions these audiences have before providing them defensible, accurate, and objective information, and with these audiences supporting the generation of new knowledge or updating existing scientific knowledge. To do this effectively, collaboration is needed among domain experts, including life, computer, and social scientists with expertise in public engagement and science communication, to monitor public opinion(s) and media content and to conduct sentiment analysis within various communities that are invested in or are working to address the issue at hand.

Another consideration when engaging members of the public is that individuals who trust science may be more likely to believe and share inaccurate information containing scientific information than individuals who do not trust science (O’Brien et al. 2021). Individuals who trust science implicitly believe claims made by scientists or include scientific information. Similarly, individuals who are primed to critically evaluate information are less likely to believe a particular claim regardless of whether scientific information is referenced than individuals who are primed to trust the science. These findings suggest that public engagement strategies focusing on encouraging critical evaluation of information may be more effective than encouraging a general trust in science at reducing belief in misinformation.

How Can Uncertainties Be Captured and Communicated Transparently?

Uncertainty is a particularly challenging topic to address, particularly when very little information is available or situations are unfolding. In these situations, the “best available science at the moment” may be necessary to consider and avoid long-term damage to science as a trusted institution (Scheufele et al. 2020, 2021a). The COVID-19 pandemic illustrated the difficulty in navigating uncertainty within the constantly changing global and national policy, public health, and pandemic-relevant scientific landscape (Marcus and Oransky 2020).

Within this context, poor communication of what actually is known scientifically and what still remains unknown creates significant challenges, from scientists overstating preexisting knowledge of response measures used for other infectious diseases (e.g., effectiveness of mask use, particularly masks made of different materials, and distance of airborne spread of viral particles) to journalists using results from single studies, including studies that were disseminated through preprint servers (Bok et al. 2021; Escandón et al. 2021; Hornik et al. 2021; West and Bergstrom 2021).

Mistakes or miscalculations in the published scientific literature can enhance these complexities further. For example, the Chief of Thai Traditional and Alternative Medicine self-retracted a study titled “Efficacy and safety of Andrographis paniculata extract in patients with mild COVID-19: A randomized controlled trial,” which was uploaded to the online preprint server medRxiv in July 2021, because of a statistical miscalculation, which led the authors to conclude the study’s hypothesis was more valid than it actually was (Wanaratna et al. 2021).

UNDERSTANDING UNCERTAINTY IN CORRECTING INACCURATE AND MISLEADING INFORMATION

Scientific information and conclusions rarely are absolutely certain at any given time, and scientific knowledge always is open to revision. Expressing the degree of certainty around scientific knowledge, conclusions, and scientifically informed recommendations is necessary but complex for several reasons.

Degrees of certainty exist around scientific conclusions and degrees of certainty exist around scientific recommendations for practical decisions, and these types of certainty are different (Fischhoff and Davis 2014; The Royal Society 2022). Royall (2000) offers a tripartite distinction of certainty about what is known, certainty about what to believe, and certainty about what to do. With respect to scientific conclusion-making, standards of evidence do (or should) not logically change as a function of the needs of individuals, organizations, or humanity at large or the difficulty of doing studies (Travis 2006; Vernooij et al. 2021). In contrast, the degree of certainty needed for prudent decision-making and action-taking that are informed by science may indeed change as a function of the circumstances, and the degree of evidence needed in such situations is itself an issue that cannot be decided by science alone (Richardson et al. 2017; The Royal Society 2022).

Describing the uncertainty around scientific information, conclusions, or scientifically informed decisions is difficult because uncertainty is not a simple dichotomy of “yes things are certain” or “no, things are not certain”—uncertainty exists in degrees (Weiss and LaPorte 2018). Only in a small minority of cases can individuals objectively and scientifically quantify the degree of scientific certainty or uncertainty. In most cases, only qualitative statements can be made about the degree of certainty and the factors influencing certainty (NRC 1996). Efforts to make certainty quantitative and some of the inherent subjectivity involved have been recognized previously (Fischhoff and Davis 2014; van der Bles et al. 2019).

___________________

⁴ See https://www.nap.edu/catalog/26466.

the claim in an authoritative and defensible manner; and (4) the potential for amplification of the claim rather than its correction. Claims that may result in significant harms (e.g., presenting significant barriers to public health response during an epidemic or accusation of deliberately malicious action) and that are caused or perpetuated by inaccurate or misleading information could be addressed through the development and/or communication of accurate and evidence-supported science. If insufficient data exist or the knowledge basis is weak, scientists may choose either to not address the inaccurate information or to undertake new, peer-reviewed analysis to produce the needed knowledge to correct the inaccuracies. If scientists choose to address the claims, considerations about the spread of information, uncertainty associated with the corrective message, methods and primary audience for communication, and public engagement will determine their overall effectiveness at countering inaccuracies. This collective information has been translated into a how-to guide⁵ for scientists to determine whether and how to address particular inaccurate information and broader misinformation claims, and to whom and how to communicate the corrective actions.

REFERENCES

___________________

⁵ See https://www.nap.edu/catalog/26466.

This page intentionally left blank.