Localizing COVID-19 Misinformation: A Case Study of Tracking Twitter Pandemic Narratives in Pennsylvania Using Computational Network Science.

The recent COVID-19 outbreak has highlighted the importance of effective communication strategies to control the spread of the virus and debunk misinformation. By using accurate narratives, both online and offline, we can motivate communities to follow preventive measures and shape attitudes toward them. However, the abundance of misinformation stories can lead to vaccine hesitancy, obstructing the timely implementation of preventive measures, such as vaccination. Therefore, it is crucial to create appropriate and community-centered solutions based on regional data analysis to address mis/disinformation narratives and implement effective countermeasures specific to the particular geographic area.In this case study, we have attempted to create a research pipeline to analyze local narratives on social media, particularly Twitter, to identify misinformation spread locally, using the state of Pennsylvania as an example. Our proposed methodology pipeline identifies main communication trends and misinformation stories for the major cities and counties in southwestern PA, aiming to assist local health officials and public health specialists in instantly addressing pandemic communication issues, including misinformation narratives. Additionally, we investigated anti-vax actors' strategies in promoting harmful narratives. Our pipeline includes data collection, Twitter influencer analysis, Louvain clustering, BEND maneuver analysis, bot identification, and vaccine stance detection. Public health organizations and community-centered entities can implement this data-driven approach to health communication to inform their pandemic strategies.

The effects of evolutionary adaptations on spreading processes in complex networks.

A common theme among previously proposed models for network epidemics is the assumption that the propagating object (e.g., a pathogen [in the context of infectious disease propagation] or a piece of information [in the context of information propagation]) is transferred across network nodes without going through any modification or evolutionary adaptations. However, in real-life spreading processes, pathogens often evolve in response to changing environments and medical interventions, and information is often modified by individuals before being forwarded. In this article, we investigate the effects of evolutionary adaptations on spreading processes in complex networks with the aim of 1) revealing the role of evolutionary adaptations on the threshold, probability, and final size of epidemics and 2) exploring the interplay between the structural properties of the network and the evolutionary adaptations of the spreading process.

Social-Cyber Maneuvers During the COVID-19 Vaccine Initial Rollout: Content Analysis of Tweets.

BACKGROUND: During the time surrounding the approval and initial distribution of Pfizer-BioNTech's COVID-19 vaccine, large numbers of social media users took to using their platforms to voice opinions on the vaccine. They formed pro- and anti-vaccination groups toward the purpose of influencing behaviors to vaccinate or not to vaccinate. The methods of persuasion and manipulation for convincing audiences online can be characterized under a framework for social-cyber maneuvers known as the BEND maneuvers. Previous studies have been conducted on the spread of COVID-19 vaccine disinformation. However, these previous studies lacked comparative analyses over time on both community stances and the competing techniques of manipulating both the narrative and network structure to persuade target audiences. OBJECTIVE: This study aimed to understand community response to vaccination by dividing Twitter data from the initial Pfizer-BioNTech COVID-19 vaccine rollout into pro-vaccine and anti-vaccine stances, identifying key actors and groups, and evaluating how the different communities use social-cyber maneuvers, or BEND maneuvers, to influence their target audiences and the network as a whole. METHODS: COVID-19 Twitter vaccine data were collected using the Twitter application programming interface (API) for 1-week periods before, during, and 6 weeks after the initial Pfizer-BioNTech rollout (December 2020 to January 2021). Bot identifications and linguistic cues were derived for users and tweets, respectively, to use as metrics for evaluating social-cyber maneuvers. Organization Risk Analyzer (ORA)-PRO software was then used to separate the vaccine data into pro-vaccine and anti-vaccine communities and to facilitate identification of key actors, groups, and BEND maneuvers for a comparative analysis between each community and the entire network. RESULTS: Both the pro-vaccine and anti-vaccine communities used combinations of the 16 BEND maneuvers to persuade their target audiences of their particular stances. Our analysis showed how each side attempted to build its own community while simultaneously narrowing and neglecting the opposing community. Pro-vaccine users primarily used positive maneuvers such as excite and explain messages to encourage vaccination and backed leaders within their group. In contrast, anti-vaccine users relied on negative maneuvers to dismay and distort messages with narratives on side effects and death and attempted to neutralize the effectiveness of the leaders within the pro-vaccine community. Furthermore, nuking through platform policies showed to be effective in reducing the size of the anti-vaccine online community and the quantity of anti-vaccine messages. CONCLUSIONS: Social media continues to be a domain for manipulating beliefs and ideas. These conversations can ultimately lead to real-world actions such as to vaccinate or not to vaccinate against COVID-19. Moreover, social media policies should be further explored as an effective means for curbing disinformation and misinformation online.

Duped by Bots: Why Some are Better than Others at Detecting Fake Social Media Personas.

OBJECTIVE: We examine individuals' ability to detect social bots among Twitter personas, along with participant and persona features associated with that ability. BACKGROUND: Social media users need to distinguish bots from human users. We develop and demonstrate a methodology for assessing those abilities, with a simulated social media task. METHOD: We analyze performance from a signal detection theory perspective, using a task that asked lay participants whether each of 50 Twitter personas was a human or social bot. We used the agreement of two machine learning models to estimate the probability of each persona being a bot. We estimated the probability of participants indicating that a persona was a bot with a generalized linear mixed-effects model using participant characteristics (social media experience, analytical reasoning, and political views) and stimulus characteristics (bot indicator score and political tone) as regressors. RESULTS: On average, participants had modest sensitivity (d') and a criterion that favored responding "human." Exploratory analyses found greater sensitivity for participants (a) with less self-reported social media experience, (b) greater analytical reasoning ability, and (c) who were evaluating personas with opposing political views. Some patterns varied with participants' political identity. CONCLUSIONS: Individuals have limited ability to detect social bots, with greater aversion to mistaking bots for humans than vice versa. Greater social media experience and myside bias appeared to reduce performance, as did less analytical reasoning ability. APPLICATION: These patterns suggest the need for interventions, especially when users feel most familiar with social media.

Coordinating Narratives Framework for cross-platform analysis in the 2021 US Capitol riots.

Coordinated disinformation campaigns are used to influence social media users, potentially leading to offline violence. In this study, we introduce a general methodology to uncover coordinated messaging through an analysis of user posts on Parler. The proposed Coordinating Narratives Framework constructs a user-to-user coordination graph, which is induced by a user-to-text graph and a text-to-text similarity graph. The text-to-text graph is constructed based on the textual similarity of Parler and Twitter posts. We study three influential groups of users in the 6 January 2020 Capitol riots and detect networks of coordinated user clusters that post similar textual content in support of disinformation narratives related to the U.S. 2020 elections. We further extend our methodology to Twitter tweets to identify authors that share the same disinformation messaging as the aforementioned Parler user groups.

"The coronavirus is a bioweapon": classifying coronavirus stories on fact-checking sites.

The 2020 coronavirus pandemic has heightened the need to flag coronavirus-related misinformation, and fact-checking groups have taken to verifying misinformation on the Internet. We explore stories reported by fact-checking groups PolitiFact, Poynter and Snopes from January to June 2020. We characterise these stories into six clusters, then analyse temporal trends of story validity and the level of agreement across sites. The sites present the same stories 78% of the time, with the highest agreement between Poynter and PolitiFact. We further break down the story clusters into more granular story types by proposing a unique automated method, which can be used to classify diverse story sources in both fact-checked stories and tweets. Our results show story type classification performs best when trained on the same medium, with contextualised BERT vector representations outperforming a Bag-Of-Words classifier.

Active, aggressive, but to little avail: characterizing bot activity during the 2020 Singaporean elections.

Digital disinformation presents a challenging problem for democracies worldwide, especially in times of crisis like the COVID-19 pandemic. In countries like Singapore, legislative efforts to quell fake news constitute relatively new and understudied contexts for understanding local information operations. This paper presents a social cybersecurity analysis of the 2020 Singaporean elections, which took place at the height of the pandemic and after the recent passage of an anti-fake news law. Harnessing a dataset of 240,000 tweets about the elections, we found that 26.99% of participating accounts were likely to be bots, responsible for a larger proportion of bot tweets than the election in 2015. Textual analysis further showed that the detected bots used simpler and more abusive second-person language, as well as hashtags related to COVID-19 and voter activity-pointing to aggressive tactics potentially fuelling online hostility and questioning the legitimacy of the polls. Finally, bots were associated with larger, less dense, and less echo chamber-like communities, suggesting efforts to participate in larger, mainstream conversations. However, despite their distinct narrative and network maneuvers, bots generally did not hold significant influence throughout the social network. Hence, although intersecting concerns of political conflict during a global pandemic may promptly raise the possibility of online interference, we quantify both the efforts and limits of bot-fueled disinformation in the 2020 Singaporean elections. We conclude with several implications for digital disinformation in times of crisis, in the Asia-Pacific and beyond.

Exploring the Stability of Communication Network Metrics in a Dynamic Nursing Context.

Network stability is of increasing interest to researchers as they try to understand the dynamic processes by which social networks form and evolve. Because hospital patient care units (PCUs) need flexibility to adapt to environmental changes (Vardaman, Cornell, & Clancy, 2012), their networks are unlikely to be uniformly stable and will evolve over time. This study aimed to identify a metric (or set of metrics) sufficiently stable to apply to PCU staff information sharing and advice seeking communication networks over time. Using Coefficient of Variation, we assessed both Across Time Stability (ATS) and Global Stability over four data collection times (Baseline and 1, 4, and 7 months later). When metrics were stable using both methods, we considered them "super stable." Nine metrics met that criterion (Node Set Size, Average Distance, Clustering Coefficient, Density, Weighted Density, Diffusion, Total Degree Centrality, Betweenness Centrality, and Eigenvector Centrality). Unstable metrics included Hierarchy, Fragmentation, Isolate Count, and Clique Count. We also examined the effect of staff members' confidence in the information obtained from other staff members. When confidence was high, the "super stable" metrics remained "super stable," but when low, none of the "super stable" metrics persisted as "super stable." Our results suggest that nursing units represent what Barker (1968) termed dynamic behavior settings in which, as is typical, multiple nursing staff must constantly adjust to various circumstances, primarily through communication (e.g., discussing patient care or requesting advice on providing patient care), to preserve the functional integrity (i.e., ability to meet patient care goals) of the units, thus producing the observed stability over time of nine network metrics. The observed metric stability provides support for using network analysis to study communication patterns in dynamic behavior settings such as PCUs.

Social cybersecurity: an emerging science.

With the rise of online platforms where individuals could gather and spread information came the rise of online cybercrimes aimed at taking advantage of not just single individuals but collectives. In response, researchers and practitioners began trying to understand this digital playground and the way in which individuals who were socially and digitally embedded could be manipulated. What is emerging is a new scientific and engineering discipline-social cybersecurity. This paper defines this emerging area, provides case examples of the research issues and types of tools needed, and lays out a program of research in this area.

Association between medical academic genealogy and publication outcome: impact of unconscious bias on scientific objectivity.

BACKGROUND: Our previous studies suggest that the training history of an investigator, termed "medical academic genealogy", influences the outcomes of that investigator's research. Here, we use meta-analysis and quantitative statistical modeling to determine whether such effects contribute to systematic bias in published conclusions. METHODS: A total of 108 articles were identified through a comprehensive search of the high-grade glioma (HGG) surgical resection literature. Analysis was performed on the 70 articles with sufficient data for meta-analysis. Pooled estimates were generated for key academic genealogies. Monte Carlo simulations were performed to determine whether the effects attributed to genealogy alone can arise due to chance alone. RESULTS: Meta-analysis of the HGG literature without consideration for academic medical genealogy revealed that gross total resection (GTR) was associated with a significant decrease in the odds ratio (OR) for the hazard of death after surgery for both anaplastic astrocytoma (AA) and glioblastoma (AA: log [OR] = - 0.04, 95% CI [- 0.07 to - 0.01]; glioblastoma log [OR] = - 0.36, 95% CI [- 0.44 to - 0.29]). For the glioblastoma literature, meta-analysis of articles contributed by members of a genealogy consisting of mostly radiation oncologists revealed no reduction in the hazard of death after GTR [log [OR] = - 0.16, 95% CI [- 0.41 to 0.09]. In contrast, meta-analysis of published articles contributed by members of a genealogy consisting of mostly neurosurgeons revealed that GTR was associated with a significant reduction in the hazard of death [log [OR] = - 0.29, 95% CI [- 0.40 to 0.18]. Monte Carlo simulation revealed that the observed discrepancy between the articles contributed by the members of these two genealogies was unlikely to arise by chance alone (p < 0.006). CONCLUSIONS: Meta-analysis of articles contributed by authors belonging to the different medical academic genealogies yielded distinct and contradictory pooled point-estimates, suggesting that genealogy contributes to systematic bias in the published literature.

Relationship of Staff Information Sharing and Advice Networks to Patient Safety Outcomes.

OBJECTIVE: The aim of this study was to compare information sharing and advice networks' relationships with patient safety outcomes. BACKGROUND: Communication contributes to medical errors, but rarely is it clear what elements of communication are key. METHODS: We investigated relationships of information-sharing and advice networks to patient safety outcomes in 24 patient care units from 3 hospitals over 7 months. Web-based questionnaires completed via Android tablets provided data to create 2 networks using ORA, a social network analysis application. Each hospital provided nurse-sensitive patient safety outcomes. RESULTS: In both networks, medication errors correlated positively with node count and average distance and negatively with clustering coefficient. Density and weighted density negatively correlated with medication errors and falls in both networks. Eigenvector and total degree centrality correlated negatively with both safety outcomes, whereas betweenness centrality positively related to falls in the information-sharing network. CONCLUSION: Technology-enabled social network analysis data collection is feasible and can provide managers actionable system-level information.

Impact of medical academic genealogy on publication patterns: An analysis of the literature for surgical resection in brain tumor patients.

"Academic genealogy" refers to the linking of scientists and scholars based on their dissertation supervisors. We propose that this concept can be applied to medical training and that this "medical academic genealogy" may influence the landscape of the peer-reviewed literature. We performed a comprehensive PubMed search to identify US authors who have contributed peer-reviewed articles on a neurosurgery topic that remains controversial: the value of maximal resection for high-grade gliomas (HGGs). Training information for each key author (defined as the first or last author of an article) was collected (eg, author's medical school, residency, and fellowship training). Authors were recursively linked to faculty mentors to form genealogies. Correlations between genealogy and publication result were examined. Our search identified 108 articles with 160 unique key authors. Authors who were members of 2 genealogies (14% of key authors) contributed to 38% of all articles. If an article contained an authorship contribution from the first genealogy, its results were more likely to support maximal resection (log odds ratio = 2.74, p < 0.028) relative to articles without such contribution. In contrast, if an article contained an authorship contribution from the second genealogy, it was less likely to support maximal resection (log odds ratio = -1.74, p < 0.026). We conclude that the literature on surgical resection for HGGs is influenced by medical academic genealogies, and that articles contributed by authors of select genealogies share common results. These findings have important implications for the interpretation of scientific literature, design of medical training, and health care policy.

'Journal Bias' in peer-reviewed literature: an analysis of the surgical high-grade glioma literature.

The core premise of evidence-based medicine is that clinical decisions are informed by the peer-reviewed literature. To extract meaningful conclusions from this literature, one must first understand the various forms of biases inherent within the process of peer review. We performed an exhaustive search that identified articles exploring the question of whether survival benefit was associated with maximal high-grade glioma (HGG) resection and analysed this literature for patterns of publication. We found that the distribution of these 108 articles among the 26 journals to be non-random (p<0.01), with 75 of the 108 published articles (69%) appearing in 6 of the 26 journals (25%). Moreover, certain journals were likely to publish a large number of articles from the same medical academic genealogy (authors with shared training history and/or mentor). We term the tendency of certain types of articles to be published in select journals 'journal bias' and discuss the implication of this form of bias as it pertains to evidence-based medicine.

Using *ORA, a network analysis tool, to assess the relationship of handoffs to quality and safety outcomes.

Communication during patient handoffs has been widely implicated in patient safety issues. However, few studies have actually been able to quantify the relationship between handoffs and patient outcomes. We used *ORA, a dynamic network analysis tool, to examine handoffs between day and night shifts on seven units in three hospitals in the Southwest. Using *ORA's visualization and analysis capabilities, we examined the relationships between the handoff communication network metrics and a variety of patient safety quality and satisfaction outcomes. Unique network patterns were observed for different types of outcome variable (eg, safety, symptom management, self-care, and patient satisfaction). This exploratory project demonstrates the power of *ORA to identify communication patterns for large groups, such as patient care units. *ORA's network metrics can then be related to specific patient outcomes.

The effects of legacy organization culture on post-merger integration.

We explore the relationship between the characteristics of pre-existing organization cultures and post-merger integration dynamics; this study involves examining data produced by computer simulation. Two characteristics of organization culture, its characteristic complexity and its propensity for members' to share information, are controlled in computational experiments. To characterize post-merger integration dynamics, we measure the transfer of information with respect to two types: (a) that which is necessary in performing work tasks, and (b) that which underlies the features of a group's culture. The extent to which this information is common in a group is indicative of task performance and the cultural cohesiveness of its members; leading to the level of performance for the group. We consider cultural knowledge as it pertains to both that of the entire organization and at the work-team level; often times, these can be dissimilar. We find that cultural complexity and exchange motivation vary in their influence on the diffusion of task and cultural knowledge: the more complex the culture, the longer for post-merger integration to complete, while simultaneously task performance suffers. However, the inclination for an organization to energetically share their culture with another group does not immensely impact the diffusion of cultural or task knowledge; moreover, high levels of task focus in a culture can hinder cultural diffusion, though performance is positively correlated with this characteristic. This study has relevance to post-merger integration research and practice by providing a theoretically grounded, quantitative model useful for estimating the post-merger dynamics of cultural awareness and knowledge diffusion for a specific merger situation.

Gender dynamics on Twitter during the 2020 U.S. Democratic presidential primary.

The Twitter social network for each of the top five U.S. Democratic presidential candidates in 2020 was analyzed to determine if there were any differences in the treatment of the candidates. This data set was collected from discussions of the presidential primary between December 2019 through April 2020. It was then separated into five sets,  one for each candidate. We found that the most discussed candidates, President Biden and Senator Sanders, received by far the most engagement from verified users and news agencies even before the Iowa caucuses, which was ultimately won by Mayor Buttigieg. The most popular candidates were also generally targeted more frequently by bots, trolls, and other aggressive users. However, the abusive language targeting the top two female candidates, Senators Warren and Klobuchar, included slightly more gendered and sexist language compared with the other candidates. Additionally, sexist slurs that ordinarily describe women were used more frequently than male slurs in all candidate data sets. Our results indicate that there may still be an undercurrent of sexist stereotypes permeating the social media conversation surrounding female U.S. presidential candidates.

Investigating coordinated account creation using burst detection and network analysis.

Democracies around the world face the threat of manipulation of their electorates via coordinated online influence campaigns. Researchers have responded by developing valuable methods for finding automated accounts and identifying false information, but these valiant efforts often fall into a cat-and-mouse game with perpetrators who constantly change their behavior. This has forced several researchers to go beyond the detection of individual malicious actors by instead identifying the coordinated activity that propels potent information operations. In this vein, we provide rigorous quantitative evidence for the notion that sudden increases in Twitter account creations may provide early warnings of online information operations. Analysis of fourteen months of tweets discussing the 2020 U.S. elections revealed that accounts created during bursts exhibited more similar behavior, showed more agreement on mail-in voting and mask wearing, and were more likely to be bots and share links to low-credibility sites. In concert with other techniques for detecting nefarious activity, social media platforms could temporarily limit the influence of accounts created during these bursts. Given the advantages of combining multiple anti-misinformation methods, we join others in presenting a case for the need to develop more integrable methods for countering online influence campaigns. Supplementary Information: The online version contains supplementary material available at 10.1186/s40537-023-00695-7.

A combined synchronization index for evaluating collective action social media.

Social media has provided a citizen voice, giving rise to grassroots collective action, where users deploy a concerted effort to disseminate online narratives and even carry out offline protests. Sometimes these collective action are aided by inorganic synchronization, which arise from bot actors. It is thus important to identify the synchronicity of emerging discourse on social media and the indications of organic/inorganic activity within the conversations. This provides a way of profiling an event for possibility of offline protests and violence. In this study, we build on past definitions of synchronous activity on social media- simultaneous user action-and develop a Combined Synchronization Index (CSI) which adopts a hierarchical approach in measuring user synchronicity. We apply this index on six political and social activism events on Twitter and analyzed three action types: synchronicity by hashtag, URL and @mentions.The CSI provides an overall quantification of synchronization across all action types within an event, which allows ranking of a spectrum of synchronicity across the six events. Human users have higher synchronous scores than bot users in most events; and bots and humans exhibits the most synchronized activities across all events as compared to other pairs (i.e., bot-bot and human-human). We further rely on the harmony and dissonance of CSI-Network scores with network centrality metrics to observe the presence of organic/inorganic synchronization. We hope this work aids in investigating synchronized action within social media in a collective manner.

Do you hear the people sing? Comparison of synchronized URL and narrative themes in 2020 and 2023 French protests.

Introduction: France has seen two key protests within the term of President Emmanuel Macron: one in 2020 against Islamophobia, and another in 2023 against the pension reform. During these protests, there is much chatter on online social media platforms like Twitter. Methods: In this study, we aim to analyze the differences between the online chatter of the 2 years through a network-centric view, and in particular the synchrony of users. This study begins by identifying groups of accounts that work together through two methods: temporal synchronicity and narrative similarity. We also apply a bot detection algorithm to identify bots within these networks and analyze the extent of inorganic synchronization within the discourse of these events. Results: Overall, our findings suggest that the synchrony of users in 2020 on Twitter is much higher than that of 2023, and there are more bot activity in 2020 compared to 2023.