https://orcid.org/0000-0002-6778-5098
Figures
Abstract
Objectives
The objective of this study is to investigate, from a longitudinal perspective, how WHO communicated COVID-19 related information to the public through its press conferences during the first two years of the pandemic.
Methods
The transcripts of 195 WHO COVID-19 press conferences held between January 22, 2020 and February 23, 2022 were collected. All transcripts were syntactically parsed to extract highly frequent noun chunks that were potential topics of the press conferences. First-order autoregression models were fit to identify “hot” and “cold” topics. In addition, sentiments and emotions expressed in the transcripts were analyzed using lexicon-based sentiment/emotion analyses. Mann-Kendall tests were performed to capture the possible trends of sentiments and emotions over time.
Results
First, eleven “hot” topics were identified. These topics were pertinent to anti-pandemic measures, disease surveillance and development, and vaccine-related issues. Second, no significant trend was captured in sentiments. Last, significant downward trends were found in anticipation, surprise, anger, disgust, and fear. However, no significant trends were found in joy, trust, and sadness.
Conclusions
This retrospective study provided new empirical evidence on how WHO communicated issues pertaining to COVID-19 to the general public through its press conferences. With the help of the study, members of the general public, health organizations, and other stake-holders will be able to better understand the way in which WHO has responded to various critical events during the first two years of the pandemic.
Citation: He S, Li D, Liu C-H, Xiong Y, Liu D, Feng J, et al. (2023) Crisis communication in the WHO COVID-19 press conferences: A retrospective analysis. PLoS ONE 18(3): e0282855. https://doi.org/10.1371/journal.pone.0282855
Editor: Christian Napoli, Sapienza University of Rome, ITALY
Received: December 7, 2022; Accepted: February 24, 2023; Published: March 13, 2023
Copyright: © 2023 He et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting information files.
Funding: The authors received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The outbreak of coronavirus disease 2019 (COVID-19) is an unprecedented public health crisis that has posed enormous challenges to the global healthcare systems and the general public [1,2]. These challenges range from the psychological maladjustment to the widespread misinformation on social media, and the confusion pertaining COVID-19 vaccines. In face of these challenges, timely and effective communication is needed to promote information transparency, which in turn, helps keep members of the public informed of the pandemic [1,3]. Therefore, understanding how information is communicated to facilitate the prevention and control of COVID-19 is a key research question.
The research on crisis communication during the COVID-19 pandemic has attracted extensive attention from the research community [4–8]. One line of research focuses on the sentiments or emotions expressed on social media. For example, Li et al. [9] examined changes in emotions and sentiments before and after the declaration of COVID-19 in China on the social media platform Weibo. It was found that negative emotions and sensitivity to social risks increased whereas positive emotions and life satisfaction decreased after the declaration. In addition, the public focused more on health and family than leisure and friends. Similarly, Xue et al. [10] examined the public discourse and emotions related to COVID-19 on Twitter. They found that fear arose when new confirmed cases or deaths were reported.
Another line of research examines the official announcements and statements on COVID-19 made by a local or national government [1,3,11–13]. For example, Antiwi and Nyarkoh [12] examined official statements released by the Ghana government during the COVID-19 pandemic. They found that the government always tried to spread calmness, rather than fear, to help build trust and confidence. Similarly, Schueler and Marx [14] analyzed the speech act in COVID-19 press conferences from the Dutch government. They found that assertive statements were more frequently used in neutral press conferences than in easing or tightening press conferences. Thus, the focus of neutral press conferences was to inform people on the current state of affairs. In addition, emotions of press conferences were closely associated with the development of the pandemic. That is, emotions became more tightening when facing a shortage of medical resources while emotions would be easing when situations were improved.
The aforementioned studies have contributed significantly to our understanding of crisis communication in the pandemic. However, these findings seemed inconclusive. On the one hand, most of these studies focused on opinions expressed by individuals or officials who are not medical professionals. Since the control and prevention of COVID-19 requires highly specialized knowledge in medical sciences, messages from medical professionals or public health institutions are of particular importance. On the other hand, as one of the top public health organizations, the World Health Organization (WHO) plays an critical role in supporting countries to prepare for, respond to, and recover from the pandemic [15]. The WHO COVID-19 press conference is one of the most important channels for WHO to communicate with the public during the pandemic. The press conferences covered all important issues pertaining to COVID-19 and were widely reported by international media. Despite the immense influence, how did WHO use the press conferences as a means to coordinate its responses to the prevention and control of disease and inform the public remains largely unknown.
The objective of this study is to understand, from a longitudinal perspective, how WHO communicated COVID-19 related information through its press conferences during the first two years of the pandemic. To achieve that objective, we set out to investigate the topics, the expression of sentiments and emotions [16] in the WHO COVID-19 press conferences. The following research questions guided the study.
RQ1: What were “hot” and “cold” topics in WHO COVID-19 press conferences during the first two years of the pandemic?
RQ2: How were sentiments and emotions expressed in WHO COVID-19 press conferences evolve during the first two years of the pandemic?
Methods
Data collection
The dataset used in this study was a corpus of WHO COVID-19 press conferences transcripts curated in the first week of March, 2022. We manually downloaded transcripts of the press conferences held between January 22, 2020 and February 23, 2022 from the official website of WHO. Since March 2, 2022, the press conferences covered both coronavirus and different events (e.g., the war in Ukraine). As the focus of the present study was coronavirus, transcripts released after that date were not included in the corpus. In other words, we have collected all the available samples of the press conference transcripts at the time of data collection. Because we aimed to trace the possible diachronic change of topics, sentiments, and emotions by month, all transcripts in each month were merged into one text file (see Table 1).
Topic extraction
First, all transcripts in the dataset were lower-cased and lemmatized. An initial close reading on the original transcript texts revealed two types of inconsistencies in spelling (e.g., ‘covid—19’ versus ‘covid-19’, and ‘lock down’ versus ‘lock-down’). To resolve the differences, all occurrences of ‘covid—19’ and ‘lock down’ were replaced with ‘covid-19’ and ‘lock-down’ respectively.
Second, following the methods proposed in [17], the open-source Natural Language Processing (NLP) library spacy [18] was used to parse the syntactic dependency relations [19] of each sentence in the transcripts. Then, all noun chunks, together with their frequencies and range counts, were automatically extracted from the parsed results.
Third, the parsed results were further filtered based on the repetition criteria (i.e., frequency and range) [17,20,21]. After several rounds of experiment, the final threshold of the repetition criteria was set at 16 for both frequency and range. That is, a noun chunk was considered as a potential topic if it occurred at least 16 times in the entire parsed dataset from at least 16 different transcripts. The repetition criteria were set based on the consideration of frequency and practicality. To be specific, a noun chunk should be deemed as a candidate topic when it had frequent occurrence and spread across a wide range of transcripts. More importantly, the size of the candidate noun chunks should be manageable. Noun chunks that failed to meet the repetition criteria were removed. Two researchers independently checked each of the remaining noun chunks to decide whether a noun chunk was a possible topic. When discrepancy occurred, the researchers discussed until a full agreement was reached. Note that only noun chunks that were considered as candidate topics were kept for further analyses.
Last, the normalized frequency of each noun chunk by month was computed using Formula 1 to make the comparison of noun chunks across each month more legitimate.
Sentiment and emotion analyses
Lexicon-based sentiment and emotion analyses were performed on transcripts with the R package sentimentr [22,23]. For sentiment analysis, the sentimentr package utilizes the Jockers-Rinker sentiment lexicon [24] to determine whether the overall opinion expressed in a transcript was positive (sentiment score > 0), negative (sentiment score < 0), or neutral (sentiment score = 0). For emotion analysis, the sentimentr package relies on the NRC emotion lexicon [25] to compute scores of eight categories emotions in Plutchik’s model of emotions [26] which includes anger, anticipation, disgust, fear, joy, sadness, surprise, and trust.
Note that although both sentiment analysis and emotion analysis are concerned with the affective dimension of human language, they work at different granularity and are designed to gauge different psycholinguistic features [27]. In addition, since sentiment and emotion analyses were conducted separately in the present study, using different lexicons would not have impact on the results.
Statistical analyses
Different from other topic extraction in NLP which automatically generate a given number of topics by clustering a group of key words based on machine-learning techniques, the current topic extraction process relied on the calculating the frequency and range of noun chunks (i.e., topics). Since we are interested in how topics evolved in the time frame examined, a statistical technique which can relate a variable to its past values in the time series was needed. Similarly, a technique that is able to capture to the trends of sentiments and emotions over time is also needed to serve our research goal.
For RQ1, first-order autoregression models [21,28] were fit to identify “hot” and “cold” topics based on the normalized frequency of noun chunks in each month. A noun chunk which was detected with a significant increase of normalized frequencies was considered a “hot” topic, whereas a significant decrease was indicative of a “cold” topic [24].
For RQ2, Mann-Kendall Trend tests were performed to detect possible trends in the overall sentiment and eight categories of emotions over time. Sen’s slope and the corresponding confidence intervals were computed for the rate of change.
Results
Topics
A total of 91 topics were identified in the dataset (see S1 Table). The top 10 topic in terms of raw frequencies were health system, severe disease, act accelerator, covax facility, contact tracing, healthcare worker, public health, young people, clinical trial, and physical distancing.
Results of the first-order autoregressive models showed that, of the 91 topics, the normed frequencies of 11 topics increased significantly over the past 26 months (see Table 2 and Fig 1). These topics were considered ‘hot’ topics. Interestingly, none of the topics showed significant decreasing trend in the period. That is, no significant cold topics were identified.
Sentiments and emotions
Results of the Mann-Kendall trend test on sentiment scores (Z = -1.2343, τ = -0.1753846, p. = 0.2171) showed that no significant trend in sentiment was found between January, 2020 and February, 2022. Although the overall sentiment was positive, clear fluctuations were observed during the period (see Fig 2A). That is, the trajectory of sentiment underwent constant changes as the pandemic developed.
Results of the emotion analysis showed that trust and anticipation were the most predominant categories of emotion, followed by fear, sadness, joy, surprise, anger, and disgust (see Fig 2B). In addition, results of the Mann-Kendall trend tests indicated that significant downward trends were found in anticipation, surprise, anger, disgust, and fear. However, no significant trends were found in joy, trust, and sadness (see Table 3).
Discussion
The present study investigated the topics, sentiments, and emotions in the WHO COVID-19 press conferences. Results of our study showed that 1) the normalized frequencies of 11 topics (noun chunks) increased in the period, 2) no significant monotonic trend in sentiment was found, and 3) trust, anticipation, and fear were the most predominate categories of emotions. Below, we discuss the findings in details with possible explanations.
A close reading revealed that these “hot topics” fall into three broad categories: 1) anti-pandemic measures (i.e., contact tracing, lock-down, global emergency programme, disease control, and global solidarity); 2) disease surveillance and development (i.e., mild case, report case, and severe disease); 3) vaccine-related issues (i.e., vaccination coverage, and vaccination manufacturer). Topics pertaining to anti-pandemic measures (e.g., contact tracing) attracted public attention mainly in the early stage of the pandemic [10,29,30]. On January 30 of 2020, the WHO officially declared the outbreak of COVID-19 a public health emergency concern. Since then, it has been suggested that the transmission of COVID-19 can be blocked by measures such as contact tracing systems, isolation and lock-down [31,32]. Topics concerning disease surveillance and development were stable in terms of frequency except the “severe disease”. The term “severe disease” has attracted increasing attention probably because it is associated with the rate of mortality [10,29,30,33]. In addition, severe disease is also important for studying the treatment, prognosis, control measures of the disease [31,34], especially when new variants were detected (e.g., Delta, Omicron) [35–38]. As vaccine has been considered one of the key solutions to the pandemic and multiple pharmaceutical companies and institutions (i.e., vaccination manufacturer) around the globe have actively committed to developing COVID-19 vaccines [5,39].
Contrary to the sentiment of COVID-19 related messages detected in social media [9,10,40,41], no significant trend of sentiment was observed in the WHO COVID-19 press conferences. Although the official statements are usually presented in an objective and calm manner, Frewer et al. [42] argued that if government or institutions failed to respond to a crisis timely, public outrage and instability may arise. In this connection, fluctuations in the sentiment may be observed as significant events occurred. For example, the rises of sentiment in September and November in 2020 were seemingly corroborated with the major progress in vaccine development at that time. Since vaccines are crucial in fighting the pandemic, these breakthroughs would contribute to the positive sentiments [5]. In contrast, the decrease of sentiment may be associated with negative events such as virus mutation or the severe situations during the pandemic. For example, the values of sentiment scored relatively low when the Delta variant was detected in October 2020 [35,43] and the Omicron variant was detected in November 2021[37,38]. In February 2020, the number of confirmed cases and death cases rose sharply [31], which in turn, triggered the negative sentiment.
In line with previous studies of emotions on COVID-19 [9,10,44,45], our findings suggested that anticipation and trust were the most predominate emotions. These two emotions are more inclined to reflect group cohesiveness instead of individual emotions [9]. Tziner [46] argued that group threats such as natural disasters and pandemic can make groups a community of interests and generate more beneficial and positive behaviors and social solidarity. Under such conditions, the prevalence of positive psychological reactions such as trust and anticipation can enhance public’s coping ability [47,48] and increase trust in government during a crisis [49,50]. The emotions “trust” and “anticipation” can be generated by some situation improvement such as the decline in the number of people involved in crisis (i.e., confirmed and death cases) and government’s effective effort [51,52]. As the most authoritative health organization, WHO needs to spread trust and claim to help build confidence in fighting COVID-19 [53]. On the other hand, previous studies also suggest that public health emergencies can trigger a series of negative emotional responses, including a higher level of anxiety, anger, depression and fear [9,54–56]. For example, “fear” is usually related to emergencies and crisis, since it is one of the most pivotal emotions linked with life safety and survival [57,58]. Although these emotions also can be captured in our result, the levels were low indicating WHO’s effort to avoid spreading negative emotions. In other words, WHO has consistently been positive in tone regardless of the status quo of the pandemic, which in turn, may have helped the public build confidence and look the bright side of life. In terms of the trends of emotions, positive psychological reactions (i.e., joy, anticipation, trust, surprise) exhibited a downward trend while negative emotions such as “fear” exhibited an increasing trend during the early stage of COVID-19 pandemic. Then with the growing number of recovered patients and the development of vaccines, the trends of trust and joy turned upward [5,51].
Conclusions
The retrospective study provided new empirical evidence on how WHO communicated issues pertaining to COVID-19 to the general public through its press conferences. With the help of this research, members of general public, health organizations, and other stake-holders will be able to better understand the way in which WHO has responded to various critical events during the first two years of the pandemic.
Since we only examined the monotonic trend, which may not be sufficient to account for the dynamic development of sentiments and emotions over the period. Future studies may consider examine these trends with different measures.
Supporting information
S1 Table. Topics in WHO COVID-19 press conferences from January, 2020 to February, 2022.
https://doi.org/10.1371/journal.pone.0282855.s001
(DOCX)
Acknowledgments
We would like to express our appreciation to the editor and anonymous reviewers for their valuable comments and suggestions.
References
- 1. Zahry NR, McCluskey M, Ling J. Risk governance during the COVID-19 pandemic: A quantitative content analysis of governors’ narratives on twitter. Contingencies & Crisis Mgmt. 2022; 1468–5973.12412.
- 2. Barchielli B, Cricenti C, Gallè F, Sabella EA, Liguori F, Da Molin G, et al. Climate changes, natural resources depletion, COVID-19 pandemic, and Russian-Ukrainian war: What is the impact on habits change and mental health? Int J Environ Res Public Health. 2022;19: 11929. pmid:36231229
- 3. Rao HR, Vemprala N, Akello P, Valecha R. Retweets of officials’ alarming vs reassuring messages during the COVID-19 pandemic: Implications for crisis management. Int J Inf Manage. 2020;55: 102187. pmid:32836644
- 4. Ye W. COVID-19 Twitter communication of major societal stakeholders: Health Institutions, the government, and the news media. Int J Commun. 2021; 37.
- 5. Lyu JC, Han EL, Luli GK. COVID-19 vaccine–related Discussion on Twitter: Topic modeling and sentiment analysis. J Med Internet Res. 2021;23: e24435. pmid:34115608
- 6. Su Z, McDonnell D, Wen J, Kozak M, Abbas J, Šegalo S, et al. Mental health consequences of COVID-19 media coverage: the need for effective crisis communication practices. Global Health. 2021;17: 4. pmid:33402169
- 7. Wang Y, Hao H, Platt LS. Examining risk and crisis communications of government agencies and stakeholders during early-stages of COVID-19 on Twitter. Comput Hum Behav. 2021;114: 106568. pmid:32982038
- 8. Ibrahim J-N, Fares C, Bader R. Relationship between emotional expressions and lifestyle changes among university students during COVID-19 lockdown in Lebanon. J Infect Dev Ctries. 2022;16: 1148–1158. pmid:35905019
- 9. Li S, Wang Y, Xue J, Zhao N, Zhu T. The impact of COVID-19 epidemic declaration on psychological consequences: A study on active Weibo users. Int J Environ Res Public Health. 2020;17: E2032. pmid:32204411
- 10. Xue J, Chen J, Hu R, Chen C, Zheng C, Su Y, et al. Twitter discussions and emotions about the COVID-19 pandemic: Machine learning approach. J Med Internet Res. 2020;22: e20550. pmid:33119535
- 11. Enria L, Waterlow N, Rogers NT, Brindle H, Lal S, Eggo RM, et al. Trust and transparency in times of crisis: Results from an online survey during the first wave (April 2020) of the COVID-19 epidemic in the UK. PLoS ONE. 2021;16: e0239247. pmid:33591985
- 12. Antwi Joseph, Nyarkoh Enoch. Government communication during the Covid-19 pandemic; the case of Ghana. Int. J. Public Adm. 2021;44: 1039–1039.
- 13. Kayal G, Balfaqih H, Durani F, Akeel H, Abukwaik A, Azudin N. Evaluating the impact of verified government accounts on the knowledge, attitudes, and intentions of Saudi residents during the COVID-19 pandemic. Int J Electron. 2022;18: 1–23.
- 14. Schueler D, Marx M. Speech acts in the Dutch COVID-19 press conferences. Lang Resour Eval. 2022 Jul 19:1–24. pmid:35874035
- 15.
World Health Organization. Our-work. 2022. https://www.who.int/our-work.
- 16. Liu D, Lei L. The appeal to political sentiment: An analysis of Donald Trump’s and Hillary Clinton’s speech themes and discourse strategies in the 2016 US presidential election. Discourse, Context Media. 2018;25: 143–152.
- 17. Lei L, Deng Y, Liu D. Examining research topics with a dependency-based noun phrase extraction method: a case in accounting. Libr. 2020.
- 18.
Honnibal M, Johnson M. An improved non-monotonic transition system for dependency parsing. Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing. Lisbon, Portugal: Association for Computational Linguistics; 2015. pp. 1373–1378.
- 19.
de Marneffe M-C, Dozat T, Silveira N, Haverinen K, Ginter F, Nivre J, et al. Universal Stanford dependencies: A cross-linguistic typology. In Proceedings of the Ninth International Conference on Language Resources and Evaluation (LREC’14). Reykjavik, Iceland: European Language Resources Association (ELRA); 2014. pp. 4585–4592.
- 20. Zhu H, Lei L. A dependency-based machine learning approach to the identification of research topics: a case in COVID-19 studies. Libr. 2022;40: 495–515.
- 21. Lin Z, Lei L. The research trends of multilingualism in applied linguistics and education (2000–2019): A bibliometric analysis. Sustainability. 2020;12: 6058.
- 22.
Jockers, M. J. and H. P. Bang. Syuzhet: Extracts sentiment and sentiment-derived plot arcs from text. 2017. https://CRAN.R-project.org/package=syuzhet.
- 23.
Rinker, T. Trinker/sentimentr: Dictionary based sentiment analysis that considers valence shifters (version 2.6.1). 2018. http://github.com/trinker/sentimentr.
- 24.
Rinker, T. Calculate text polarity sentiment. 2019. https://cran.r-project.org/web/packages/sentimentr/sentimentr.pdf.
- 25.
Mohammad, S. M. Sentiment and emotion lexicons. 2010. http://saifmohammad.com/WebPages/lexicons.html.
- 26. Plutchik R. The Nature of emotions: Human emotions have deep evolutionary roots, a fact that may explain their complexity and provide tools for clinical practice. Am. Sci. 2001; 89(4):344–350.
- 27.
Lei L, Liu D. Conducting sentiment analysis. 1st ed. Cambridge University Press; 2021.
- 28. Gardner G., Harvey A. C. and Phillips G. D. A. Algorithm AS 154: An algorithm for exact maximum likelihood estimation of autoregressive-moving average models by means of kalman filtering. J R Stat Soc. 1980;29(3): 311–322.
- 29. Xue J, Chen J, Chen C, Zheng C, Li S, Zhu T. Public discourse and sentiment during the COVID 19 pandemic: Using latent Dirichlet allocation for topic modeling on Twitter. PLoS ONE. 2020;15: e0239441. pmid:32976519
- 30. Abd-Alrazaq A, Alhuwail D, Househ M, Hamdi M, Shah Z. Top concerns of tweeters during the COVID-19 pandemic: Infoveillance study. J Med Internet Res. 2020;22: e19016. pmid:32287039
- 31. Sohrabi C, Alsafi Z, O’Neill N, Khan M, Kerwan A, Al-Jabir A, et al. World Health Organization declares global emergency: A review of the 2019 novel coronavirus (COVID-19). Int J Surg. 2020;76: 71–76. pmid:32112977
- 32.
World Health Organization, Novel Coronavirus(2019-nCoV), Situation Report—12 (2020).2020. https://www.who.int/emergencies/diseases/novel-coronavirus-2019/situation-reports.
- 33. Stokes DC, Andy A, Guntuku SC, Ungar LH, Merchant RM. Public priorities and concerns regarding COVID-19 in an online discussion forum: Longitudinal topic modeling. J Gen Intern Med. 2020;35: 2244–2247. pmid:32399912
- 34. Stepanova M, Lam B, Younossi E, Felix S, Ziayee M, Price J, et al. The impact of variants and vaccination on the mortality and resource utilization of hospitalized patients with COVID-19. BMC Infect Dis. 2022;22: 702. pmid:35996076
- 35. Callaway E. Delta coronavirus variant: scientists brace for impact. Nature. 2021;595: 17–18. pmid:34158664
- 36. Eales O, Page AJ, de Oliveira Martins L, Wang H, Bodinier B, Haw D, et al. SARS-CoV-2 lineage dynamics in England from September to November 2021: high diversity of Delta sub-lineages and increased transmissibility of AY.4.2. BMC Infect Dis. 2022;22: 647. pmid:35896970
- 37. Nealon J, Cowling BJ. Omicron severity: milder but not mild. Lancet. 2022;399: 412–413. pmid:35065007
- 38. Zhang P, Yang S, Dai S, How Jin Aik D, Yang S, Jia P. Global spreading of Omicron variant of COVID-19. Geospat Health. 2022;17. pmid:35829683
- 39. Le TT, Cramer JP, Chen R, Mayhew S. Evolution of the COVID-19 vaccine development landscape. Nat Rev Drug Discov. 2020;19: 667–668. pmid:32887942
- 40. Jabalameli S, Xu Y, Shetty S. Spatial and sentiment analysis of public opinion toward COVID-19 pandemic using twitter data: At the early stage of vaccination. Int J Disaster Risk Reduct. 2022;80: 103204. pmid:35935613
- 41. Su Y, Xue J, Liu X, Wu P, Chen J, Chen C, et al. Examining the impact of COVID-19 lockdown in Wuhan and Lombardy: A psycholinguistic analysis on Weibo and Twitter. Int J Environ Res Public Health. 2020;17: E4552. pmid:32599811
- 42. Frewer LJ, Miles S, Marsh R. The media and genetically modified foods: evidence in support of social amplification of risk. Risk Anal. 2002;22: 701–711. pmid:12224744
- 43. Vaidyanathan G. Coronavirus variants are spreading in India—what scientists know so far. Nature. 2021;593: 321–322. pmid:33976409
- 44. Zhang W, Mukerjee S, Qin H. Topics and sentiments influence likes: A study of Facebook public pages’ posts about COVID-19 vaccination. Cyberpsychol Behav Soc Netw. 2022;25: 552–560. pmid:35969378
- 45. Li Q, Wei C, Dang J, Cao L, Liu L. Tracking and analyzing public emotion evolutions during COVID-19: A case study from the event-driven perspective on Microblogs. Int J Environ Res Public Health. 2020;17: E6888. pmid:32967163
- 46. Tziner A. Group cohesiveness: A dynamic perspective. Soc Behav Pers. 1982;10: 205–211.
- 47. Fredrickson BL, Tugade MM, Waugh CE, Larkin GR. What good are positive emotions in crises? A prospective study of resilience and emotions following the terrorist attacks on the United States on September 11th, 2001. J Pers Soc Psychol. 2003;84: 365–376. pmid:12585810
- 48. Fredrickson BL, Joiner T. Positive emotions trigger upward spirals toward emotional well-being. Psychol Sci. 2002;13: 172–175. pmid:11934003
- 49. Guo SJ. The 2013 Boston marathon bombing: Publics’ emotions, coping, and organizational engagement. Public Relat Rev. 2017;43: 755–767.
- 50. Kim HK, Niederdeppe J. The role of emotional response during an H1N1 influenza pandemic on a college campus. J Public Relat Res. 2013;25: 30–50.
- 51. Gupta V, Jain N, Katariya P, Kumar A, Mohan S, Ahmadian A, et al. An emotion care model using multimodal textual analysis on COVID-19. Chaos Soliton Fract. 2021;144: 110708. pmid:33519125
- 52. Lin C-Y, Broström A, Griffiths MD, Pakpour AH. Investigating mediated effects of fear of COVID-19 and COVID-19 misunderstanding in the association between problematic social media use, psychological distress, and insomnia. Internet Interv. 2020;21: 100345. pmid:32868992
- 53. Klabbers J. The second most difficult job in the world: Reflections on covid-19. J Int Humanit Legal Stud. 2020;11: 270–281.
- 54. Tam CWC, Pang EPF, Lam LCW, Chiu HFK. Severe acute respiratory syndrome (SARS) in Hong Kong in 2003: stress and psychological impact among frontline healthcare workers. Psychol Med. 2004;34: 1197–1204. pmid:15697046
- 55. Maunder R, Hunter J, Vincent L, Bennett J, Peladeau N, Leszcz M, et al. The immediate psychological and occupational impact of the 2003 SARS outbreak in a teaching hospital. Can Med Assoc J. 2003;168: 1245–1251. pmid:12743065
- 56.
Nobles, J., Martin, F., Dawson, S., Moran, P. and Savovic, J. The potential impact of COVID-19 on mental health outcomes and the implications for service solutions. 15 April 2020. https://arc-w.nihr.ac.uk/research-and-implementation/covid-19-response/potential-impact-of-covid-19-on-mental-health-outcomes-and-the-implications-for-service-solutions/.
- 57. Pakpour AH, Griffiths MD. The fear of COVID-19 and its role in preventive behaviours. J Concurr Disord. 2020;2: 58–63.
- 58. Adikari A, Nawaratne R, De Silva D, Ranasinghe S, Alahakoon O, Alahakoon D. Emotions of COVID-19: Content analysis of self-reported information using artificial intelligence. J Med Internet Res. 2021;23: e27341. pmid:33819167