Published on 22.05.20 in Vol 6, No 2 (2020): Apr-Jun
Preprints (earlier versions) of this paper are available at http://preprints.jmir.org/preprint/19447, first published Apr 17, 2020.
Global Sentiments Surrounding the COVID-19 Pandemic on Twitter: Analysis of Twitter Trends
Background: With the World Health Organization’s pandemic declaration and government-initiated actions against coronavirus disease (COVID-19), sentiments surrounding COVID-19 have evolved rapidly.
Objective: This study aimed to examine worldwide trends of four emotions—fear, anger, sadness, and joy—and the narratives underlying those emotions during the COVID-19 pandemic.
Methods: Over 20 million social media twitter posts made during the early phases of the COVID-19 outbreak from January 28 to April 9, 2020, were collected using “wuhan,” “corona,” “nCov,” and “covid” as search keywords.
Results: Public emotions shifted strongly from fear to anger over the course of the pandemic, while sadness and joy also surfaced. Findings from word clouds suggest that fears around shortages of COVID-19 tests and medical supplies became increasingly widespread discussion points. Anger shifted from xenophobia at the beginning of the pandemic to discourse around the stay-at-home notices. Sadness was highlighted by the topics of losing friends and family members, while topics related to joy included words of gratitude and good health.
Conclusions: Overall, global COVID-19 sentiments have shown rapid evolutions within just the span of a few weeks. Findings suggest that emotion-driven collective issues around shared public distress experiences of the COVID-19 pandemic are developing and include large-scale social isolation and the loss of human lives. The steady rise of societal concerns indicated by negative emotions needs to be monitored and controlled by complementing regular crisis communication with strategic public health communication that aims to balance public psychological wellbeing.
JMIR Public Health Surveill 2020;6(2):e19447
The coronavirus disease (COVID-19) pandemic has infected individuals in more than 200 countries and resulted in many deaths . With the World Health Organization’s (WHO’s) pandemic declaration and government-initiated actions against the disease, sentiments about COVID-19 are rapidly evolving. In the past decade, social media analytic tools have been utilized to monitor public sentiments and communication patterns of public health emergencies like the Ebola and Zika epidemics. Although many studies have investigated general sentiment valences and discourse topics [ , ], specific emotions have been found to be more closely linked to psychological processes and behaviors than the overall positive and negative valences [ ]. Therefore, we postulate that distinct emotions emerging from social media and their underlying narratives are highly relevant to the current COVID-19 crisis and can provide actionable insights into the efficacy of public health messaging.
Particularly, we focused on four emotions: fear, anger, sadness, and joy. According to Plutchik’s Wheel of Emotions , fear-anger and sadness-joy are the basic emotion pairs of opposite experiences. Fear is an unpleasant emotion typically arising from danger or uncertainties caused by circumstances, while anger results from uncertainties caused by others [ ]. Sadness is a negative emotion experienced typically after unpleasant circumstances that are out of one’s control, and joy is a positive feeling after pleasant events that are appraised as certain and under control [ ]. Investigating the evolution of these four basic emotions can demonstrate the changing dynamics of the public’s experience to the crisis.
In this report, we present the results of Twitter users’ public emotional responses to the pandemic. Trends of the four basic emotions and the narratives underlying those emotions were examined.
English tweets related to COVID-19 worldwide posted from January 28 to April 9, 2020, were collected from Twitter’s standard search application programming interface using “wuhan,” “corona,” “nCov,” and “covid” as search keywords. These keywords were selected because they were widely used during the early assessment of the COVID-19 situation. Publicly accessible tweets from any type of account that contained any of the keywords were collected. The underlying emotions of tweets were analyzed using the algorithm CrystalFeel, a sentiment analytic technology whose accuracy had been demonstrated (see details and examples in) [ ]. Pearson r correlations were conducted between emotions and date to demonstrate the trends of emotions across time statistically. Word clouds were generated for each of the four emotions based on the top frequent unigrams and bigrams.
A total of 20,325,929 tweets were collected, including 7,033,158 unique users from more than 170 countries. The daily proportion of tweets stratified by emotion were plotted across time ().
Expectedly, fear was the dominant emotion at the end of January when the disease first surfaced. The prominence of fear gradually dropped to less than 30% of daily tweets in early April as the crisis developed (r71=–0.92; P<.001). In contrast, tweets on anger progressively increased from late January to early March, peaking at 29% on March 12, a day after the pandemic declaration by the WHO. Tweets on anger slightly decreased since then, but remains at a relatively high level (r71=0.75; P<.001). Coinciding with the decrease of tweets on both fear and anger after the pandemic announcement, tweets on sadness, although proportionally lower than those of the other emotions, doubled since the WHO declaration (r71=0.88; P<.001). Similarly, tweets on joy, suggesting a sense of pride, gratitude, hope, and happiness , also increased (r71=0.86; P<.001).
Further analyses using word clouds suggest that narratives underlying those emotions evolved as the pandemic developed (). In late January, fear was possibly related to the emergence of COVID-19 and its unknown nature, causing uncertainty about containment and spread, indicated by words such as “first case” and “outbreak.” However, as the pandemic escalated, the narratives suggested fear about shortages of COVID-19 tests and medical supplies indicated by words such as “test shortages” and “uncounted.” The anger word clouds suggest xenophobia at the beginning of the pandemic when the disease was predominantly localized to China and Asia, indicated by words such as “racist” and “Chinese people.” Anger then shifted to discourse around isolation fatigue that can occur from social seclusion, indicated by words such as “stay home” and several swear words. Narratives of recent sadness surrounding the topics of losing friends and family members are surfacing, with words relating to “loved one” and “passed away,” highlighting potential social concerns arising from personal traumatic experiences of the pandemic. The world has also seen a concurrent increase in the sense of joy encompassing hope, gratitude, and human resilience with words such as “Thank,” “good news,” and “feel good.”
Our initial findings suggest that global online discourse is swiftly evolving. The discourse is driven by shared public experiences of the COVID-19 pandemic, including large-scale social isolation and the loss of human lives. Although existing studies have demonstrated the immediate psychological reactions to COVID-19 [, ], our study is the first to demonstrate the evolution of responses across time.
Our findings reveal that negative emotions are dominant during the COVID-19 pandemic, supporting the recent call for action to maintain the public’s mental wellbeing for this unprecedented crisis . Negative emotions such as anger and sadness, which are increasing, need to be heeded and counterbalanced by complementing regular crisis communication with strategic public health communication that aims to balance public psychological wellbeing [ ]. If such overbearing public emotions are not addressed, there is potential for the emergence of unintended outcomes such as breeding mistrust in the handling of the disease and a belief in online falsehoods that could hinder the ongoing control of the disease [ , ].
Although the data, collected from Twitter's standard application programming interface, looks at only public tweets surrounding the four selected keywords, the estimation is appropriate for the public discourse surrounding the pandemic at present that abides to ethnical guidelines. Future studies should further investigate sentiments by examining specific countries and expanding the scope to include other media platforms such as Facebook and Weibo.
MOL conceptualized, initiated, and led the manuscript; JL and AS drafted the manuscript; PJS and WS provided discussions surrounding global results; RG and YY provided data analysis. All authors contributed to the manuscript writing, reviewed the content, and agreed with the submission.
This work is supported by the Agency for Science, Technology and Research (A*STAR) under its A*ccelerate Gap Fund (ETPL/18-GAP050-R20A) and the Singapore Ministry of Health’s National Medical Research Council under its COVID-19 Research Fund (COVID19RF-005).
Conflicts of Interest
RG and YY are co-inventors of the CrystalFeel algorithm. No other conditions or circumstances present a potential conflict of interest for the other authors.
Extended details on methods and data analysis.DOCX File , 209 KB
Narratives of emotions during the COVID-19 pandemic.DOCX File , 1913 KB
- World Health Organization. Coronavirus disease 2019 (COVID-19) Situation Report - 83 URL: https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200412-sitrep-83-covid-19.pdf?sfvrsn=697ce98d_4 [accessed 2020-05-18]
- Lwin MO, Lu J, Sheldenkar A, Schulz PJ. Strategic Uses of Facebook in Zika Outbreak Communication: Implications for the Crisis and Emergency Risk Communication Model. Int J Environ Res Public Health 2018 Sep 10;15(9) [FREE Full text] [CrossRef] [Medline]
- Zhang EX, Yang Y, Di Shang R, Simons JJP, Quek BK, Yin XF, et al. Leveraging social networking sites for disease surveillance and public sensing: the case of the 2013 avian influenza A(H7N9) outbreak in China. Western Pac Surveill Response J 2015;6(2):66-72 [FREE Full text] [CrossRef] [Medline]
- Moors A. The Integrated Theory of Emotional Behavior Follows a Radically Goal-Directed Approach. Psychological Inquiry 2017 Feb 26;28(1):68-75. [CrossRef]
- Plutchik R. A general psychoevolutionary theory of emotion. In: Plutchik R, Kellerman H, editors. Emotion: Theory, Research, and Experience. Cambridge, Massachusetts: Academic press; 1980:3-33.
- Roseman IJ. Appraisal Determinants of Emotions: Constructing a More Accurate and Comprehensive Theory. Cognition & Emotion 1996 May;10(3):241-278. [CrossRef]
- Gupta R, Yang Y. CrystalFeel at SemEval-2018 Task 1: Understanding and Detecting Emotion Intensity using Affective Lexicons. In: ACL Anthology. 2018 Presented at: Proceedings of the 12th International Workshop on Semantic Evaluation (SemEval-2018); June 5–6, 2018; New Orleans, Louisiana p. 256-263. [CrossRef]
- Wang C, Pan R, Wan X, Tan Y, Xu L, Ho CS, et al. Immediate Psychological Responses and Associated Factors during the Initial Stage of the 2019 Coronavirus Disease (COVID-19) Epidemic among the General Population in China. Int J Environ Res Public Health 2020 Mar 06;17(5) [FREE Full text] [CrossRef] [Medline]
- 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 Mar 19;17(6) [FREE Full text] [CrossRef] [Medline]
- Holmes EA, O'Connor RC, Perry VH, Tracey I, Wessely S, Arseneault L, et al. Multidisciplinary research priorities for the COVID-19 pandemic: a call for action for mental health science. Lancet Psychiatry 2020 Apr 15 [FREE Full text] [CrossRef] [Medline]
- Bavel JJV, Baicker K, Boggio PS, Capraro V, Cichocka A, Cikara M, et al. Using social and behavioural science to support COVID-19 pandemic response. Nat Hum Behav 2020 Apr 30. [CrossRef] [Medline]
- Duan L, Zhu G. Psychological interventions for people affected by the COVID-19 epidemic. Lancet Psychiatry 2020 Apr;7(4):300-302 [FREE Full text] [CrossRef] [Medline]
|COVID-19: coronavirus disease|
|WHO: World Health Organization|
Edited by T Rashid Soron, G Eysenbach; submitted 17.04.20; peer-reviewed by H Hochheiser, N Cesare, E Da Silva, M Toledo, J Villegas, C Gandhi; comments to author 06.05.20; revised version received 11.05.20; accepted 13.05.20; published 22.05.20
©May Oo Lwin, Jiahui Lu, Anita Sheldenkar, Peter Johannes Schulz, Wonsun Shin, Raj Gupta, Yinping Yang. Originally published in JMIR Public Health and Surveillance (http://publichealth.jmir.org), 22.05.2020.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Public Health and Surveillance, is properly cited. The complete bibliographic information, a link to the original publication on http://publichealth.jmir.org, as well as this copyright and license information must be included.