3.1. Specific Topics of COVID-19 Stigma
Our first study goal is to understand the specific topics of mark, group labeling, responsibility, and peril in COVID-19 related tweets. Approximately 25% of the 7000 tweets (1759) included at least one type of message content that is instrumental in developing and disseminating COVID-19 stigma. Specifically, 22.56%, 2.51%, and 0.06% of the 7000 tweets included one, two, and three types of stigma message content, respectively. No tweets included all the four components.
Marks are cues to identify members of a stigmatized group. Among the 243 tweets (3.47%) that included marks for COVID-19, four specific types of marks emerged from the data, including flu-like symptoms, personal protective equipment (PPE), Asian origin, and healthcare providers and essential workers. About 1.44% of tweets mentioned that because a person has flu-like symptoms, they may have or transmit COVID-19. For instance, a user posted, “I have a terrible coughing spell at work and these people probably think I have coronavirus. Whole time I’m choking on a piece of lint.” Another user shared a story where a family was denied service due to a daughter coughing, “Panicked passengers get family kicked off flight over coughing daughter…” Here, people use flu-like symptoms to identify people who (may) have COVID-19. In addition, 1.67% of tweets mentioned that people may have COVID-19 if they are using PPE, such as face masks, goggles, and shoe covers. A user wrote, “If they are healthy, why are they wearing face masks and goggles?” Another posted, “Face mask is for those with suspected signs. Leave protective gears for health workers and affected persons.” By linking PPE with “suspected signs,” people turned PPE into a mark to single out people with COVID-19.
Because COVID-19 was first reported in China, another mark for COVID-19 emerged in the data was Asian origin. About 2.11% of tweets mentioned that because a person is Asian, a product is made in Asia, or a place is linked to Asia (e.g., Chinatown, Chinese restaurant), that person, product, or place may have or transmit COVID-19. A user wrote, “I ordered a package from China before this coronavirus stuff became serious and once it gets here, I’m spraying it with alcohol, tying it up in a bag and throwing it in the garage.” Similarly, another user suggested that face masks from China may also have COVID-19, “When you put on a mask to avoid coronavirus but realized mask has also been made in china.” About 0.83% of tweets also mentioned that Asian people have “coronavirus privilege,” meaning that they are more likely to contract and spread COVID-19. Those statements linked Asian descent to COVID-19. Moreover, a few (0.17%) tweets also considered certain careers indicators for having or transmitting COVID-19, including healthcare providers (e.g., “I need to stay away from my nurse neighbor. She may make me sick.”) or essential workers (e.g., “My sister has to self-isolated herself since she works at a grocery store and my parents think she will get them sick.”). In short, flu-like symptoms, PPE, Asian origin, and certain careers are used to mark people who may have and spread COVID-19.
3.1.2. Group Labeling
In stigma communication, labels are created to categorize the stigmatized group as a separate group entity. A total of 83 tweets (1.19%) included group labeling and two types of labels emerged from the data. About 0.86% of tweets referred to COVID-19 as “Wuhan virus,” “Chinese/China virus,” or “Asian virus,” separating Wuhan residents, Chinese people, or Asian individuals from the general population and suggesting that those groups are more susceptible to COVID-19 and are threats to other communities. Some tweets explicitly defended such stigmatized names as “accurate” and “factual,” because they specify the origin of the disease. For instance, a user wrote, “#COVID-19 originated from #wuhan to call it #wuhanvirus is a factual statement.” Another said, “Actually calling it coronavirus is super generalized because SARS and MERS were types of coronaviruses and calling it Wuhan coronavirus or China virus actually gives it a specification based on where it originated like MERS name does.”
A few tweets (0.33%) also referred to COVID-19 as “trumpvirus” or “trumpdemic.” A user wrote, “the #coronavirus does not care if you’re a republican. This pandemic of #covid19 is not a false story. So remain calm and do not blame the dem’s for the #trumpdemic.” Another user posted, “@xxx so you think #coronavirus is a #hoax? #trumpvirus.” While users often utilized the terms to express their political views, labeling a disease with the name of a controversial political leader may unintentionally impose negativity on people who have COVID-19, facilitating the stigmatization of COVID-19.
Responsibility is message content that implies blame by making attributions about individual’s choices and control. These messages blame people for purposefully engaging in certain behaviors that may put them in the stigmatized situations, in this case as a vector for COVID-19. In total, 124 tweets (1.77%) included responsibility information and we identified three specific types of behaviors using blaming language. First, about 0.30% of tweets blamed COVID-19 on people who have different culturally linked food preferences. One user wrote, “um guys we shouldn’t support BTS [author note: A South Korean boy band] because those sick ass Chinese people trying to have a concert in the U.S. and give us the coronavirus. It’s just rude and disgusting us whites don’t want to get sick from them bat eating fuckers.” Another tweet read, “All this coronavirus shit all because someone ate something weird. Well I hope their dead.” Here, the tweets implied that people who have different eating habits should be held responsible for COVID-19. This aspect of stigma creates a cultural responsibility for the COVID-19 outbreak due to food preferences and stigmatizes all Asian populations.
Travelers were also being blamed for contracting and spreading COVID-19 in approximately 0.91% tweets. A user wrote, “People from abroad who have the coronavirus and have paid for a trip to UK are still going to travel here because they can’t get their money back.” Another posted, “It blows my mind the number of people still willing to travel with this COVID-19. it’s going to be interesting when the snowbirds get home.” Those tweets suggested that because people purposely chose to travel, they should be held accountable for the surging cases of COVID-19. Similarly, about 0.59% of tweets also blamed people who choose not to adopt COVID-19 precautions. A tweet read, “You are stupid and ignorant if you still don’t take the coronavirus situation seriously. wear a mask. wash your hands frequently. take responsibility of your life.” Another user wrote, “What I don’t understand is ppl [people] who put their mouth on the dome lids for slushies; then touch it back to another flavor nozzle... triflin! I just watched a video of a girl on sc [Snapchat]; no one thought twice about it; u [you] guys wonder why ppl have the coronavirus bc yall [because you’re all] disgusting humans.” Here, the tweets implied that people who do not follow precautions “deserved COVID-19” because they choose not to comply with public health guidelines, which put themselves and other community members in danger.
Peril is message content underlining the danger that a stigmatized group poses to the rest of the society. In the context of infectious diseases, threats of the diseases are common peril information that evoke and amplify stigma related to the diseases [16
]. About one in five tweets (1396, 19.94%) mentioned threats of COVID-19 on people’s health, their normal life, the economy, and healthcare systems.
About 9.34% of tweets mentioned the negative mental and physical health consequences of COVID-19. Some users highlighted high severity and susceptibility of COVID-19: “It is way worse than flu,” “Has higher death rates than flu,” “It’s about the rate of mortality of the coronavirus which is what makes it more dangerous and also how infectious it is,” and “About 70% of the world population will get it.” Other tweets mentioned the adverse mental health impact of COVID-19. One tweet read, “Work is so stressful. Come home news is so stressful. And I’m living in fear of bringing home coronavirus!!!” Another user wrote, “I’m so stressed about COVID-19 and today is our travel day out of TX through Atlanta to Newark. They say the body keeps the score and as much as I try to remain calm, I’ve triggered a period even though I have an IUD [intrauterine contraceptive device]. This is great.” These examples demonstrate Twitter users’ strong concerns about the health threats of COVID-19.
Nearly 6.84% of tweets focused on how COVID-19 threatens people’s ability to live a normal life, such as lack of daily resources (e.g., “Went to CVS to get some Dayquil and they’re cleaned out. I mean the shelves were empty. Is that what people are stocking up on for coronavirus??”), cancellations of vacations or activities (e.g., “Stupid coronavirus! Ruined my entire vacation plan! Today, at this time, I would’ve been chilling with @xxx in Singapore”), and home schooling (e.g., “They will close campus after Spring break. Welcome to Zoom University at my bedroom!”). These tweets reflect the concern about disruptions in daily life due to COVID-19.
Moreover, 3.93% of tweets highlighted the adverse impact of COVID-19 on the economy, such as unemployment, pay cuts, decreased GDP, and crashed stock market. For example, “If revenue continues to fall as a consequence of the coronavirus, firms will make employees redundant to cut costs,” “The first U.S. layoffs from the coronavirus are here — with more feared to come,” and “I’m feeling sick and it is not because of COVID-19 it’s the stock market ride.” Those examples indicate that in the beginning of the crisis, the economic peril of COVID-19 was on people’s minds.
A few users (1.13%) also worried about the burden of COVID-19 on the healthcare system. One user wrote, “This new normal is going to be hard and scary. but we have to start now. before we think we need to. Hospitals won’t be able to handle the massive influx.” Another posted, “If this coronavirus is so contagious, then our healthcare staff are also highly susceptible to also being struck down with the virus. who is going to man the hospitals in this case? Very concerning.” These examples show that COVID-19 poses danger to society by threatening healthcare systems.
About one in five tweets highlighted the peril of COVID-19 in various aspects of people’s life. While these are legitimate concerns related to COVID-19 and most of the information may not intend to stigmatize people with COVID-19, the fear and anxiety associated with the threats of COVID-19 can encourage people to mark, label, and blame “others” for the situation, which facilitates the creation and spread of COVID-19 stigma.
3.2. Misinformation, Conspiracy Theories, and COVID-19 Stigma
Our second study objective is to explore how the presence of COVID-19 misinformation and conspiracy theories is related to the presence of mark, group labeling, responsibility, and peril content in tweets. We coded for the presence of common misinformation about COVID-19 identified in existing studies [26
], such as COVID-19 “is fake,” “is just a flu,” “heat kills the virus,” “drinking tea will stop the coronavirus,” as well as factually false statements, such as “only Asians will get corona” and “young people will not die from the virus.” When necessary, government sources were used for fact checking. We also coded for the presence of major conspiracy theories of COVID-19 [23
], such as, “It originated in a lab in Wuhan and some idiots let it loose,” “God sent coronavirus to destroy LGBTQ people,” and “Bill Gates created the virus to test 5G.” In total, 4.21% of the tweets included misinformation about COVID-19 and 2.00% of tweets mentioned at least one COVID-19 conspiracy theories.
Chi-square tests (see Table 2
) showed that tweets with misinformation did not include more mark content than those without misinformation,
= 7000) = 2.91, p
= 0.089. There was no difference in the presence of group labeling between tweets with misinformation and without misinformation,
= 7000) = 3.71, p
= 0.054. Tweets with misinformation did not include more responsibility information than those without misinformation,
= 7000) = 0.12, p
= 0.727. Peril of COVID-19 was more likely to be present in tweets without misinformation (20.34%) than those with misinformation (10.85%),
= 7000) = 15.96, p
Tweets with conspiracy theories did not mention more mark content than those without conspiracy theories, (1, N = 7000) = 3.72, p = 0.054. Group labeling was more likely to be present in tweets with conspiracy theories (6.43%) than those without conspiracy theories (1.08%), (1, N = 7000) = 33.52, p < 0.001. Tweets with conspiracy theories (7.14%) were more likely to have responsibility information than those without conspiracy theories (1.66%), (1, N = 7000) = 23.69, p < 0.001. Peril of COVID-19 was more likely to be present in tweets without conspiracy theories (20.22%) than those with conspiracy theories (6.43%), (1, N = 7000) = 16.34, p < 0.001.
In summary, compared to tweets without misinformation, those with misinformation were less likely to mention the threats of COVID-19. Compared to tweets without conspiracy theories, tweets with conspiracy theories were more likely include group labeling and responsibility information, but less likely to mention the peril of COVID-19.