Weibo and gender during COVID-19
The first outbreak of a ‘novel pneumonia’ in Wuhan, China was circulated on Weibo in late December 2019.23 The (approximately) 11 million population in Wuhan were placed under an abrupt and strict cordon sanitaire from 23 January 2020 to 8 April 2020. This scale of cordon sanitaire was unprecedented. One means of communication for people in Wuhan during this time was to socially engage with each other and the rest of the world via Weibo. Weibo is the biggest social media platform in China. There are over 500 million registered Weibo users, and of these 50.10% are registered as male and 49.90% as female. Most users post in the Chinese language. There was a significant upsurge of posts on Weibo during the first lockdown, the passing of Dr. Li Wenliang, and the publicity of the Wuhan Diary. These events became popular hashtags such as #Stand by her, #Healthy China, #female HCWs, #WuhanDiary and #WuhanLockdown.23–27
Existing research on COVID-19-related posts on Weibo have been concerned with tracking community-level sentiment,6 28 29 knowledge and understanding of public health information posts,27 public opinions on trust and government,15 23 25 30 31 and response to ‘fake news’ and rumours.32 33 During the pandemic, Weibo was examined as a potential tool for real-time surveillance, including case characteristic prediction34–36 and tracking infection transmission.37 With regards to Weibo, gender and COVID-19, the only two papers to consider gender difference focused on mental health and feminist activism. The first is a sex-disaggregated survey of people who posted feelings of depression and anxiety on Weibo during the first wave of COVID-19 in China.38 The second publication examined Weibo as a ‘site’ for feminist resistance against the lockdown during COVID-19.6 We sought to examine the gender differences in posts among Weibo users who were interacting with two hashtags most associated with COVID-19: #WuhanLockdown and #WuhanDiary, during the first emergency phase of the COVID-19 outbreak in Wuhan, China (23 January 2020–8 April 2020). Wuhan to date remains the only large scale outbreak in China, and the introduction of the measures used to control the outbreak at this time was relatively novel. The significance of this study is that it provides an opportunity to understand how individuals were reflecting on the first lockdown experienced globally during the pandemic. The data provides insight into what individuals were thinking and experiencing as they lived the social and economic impacts of the pandemic.
We collected social media posts through Weibo’s Open Application Programming Interfact, by Weiboscope, a research project led by coauthor since 2010.23 This was a data pull of 66 235 posts on Weibo posted during the first wave of the pandemic in Wuhan, China, between 1 January 2020 and 1 July 2020. Two data pulls from Weiboscope were conducted according to posts using one of two hashtags: #WuhanDiary and #WuhanLockdown. These two hashtags were selected as they were the most used to journal and record daily lives during lockdown. All posts pulled between 1 January 2020 and 1 July 2020 included one of the following hashtags #Wuhandiary and/or #WuhanLockdown. Registered Weibo users are provided with male or female registration. The data pull for each post included gender identifier (if provided); location of user (if provided); and date of post. Over 90% of posts were gender identified and we only included the posts for which user’s gender was provided.
Hashtags allow Weibo users to find relevant posts on a certain topic, to foster public attention and engagement.26 39–41 Our analysis involved three methods designed to examine whether Weibo exhibited similar gendered user patterns and behaviours identified in other social media platforms during the COVID-19 pandemic. First, we examined gendered patterns in posting behaviour. We plotted the number of posts, disaggregated by gender, against the 6-month timeline to track increase/decline of posts per week by gender. We also noted post behaviour (weekdays vs weekends). Second, we graphed the percentage of posts that adopted a negative tone to align with our definition of negative sentiment posts over time among female and male registered users. Previous sentiment studies of Weibo posts at the early stage of outbreak (prelockdown) found that use of words including fear, disappointment, guilt and anger increased significantly during the outbreak period (between 20 January to 23 February).42 To plot sentiment during the lockdown period we also adopted a timeline of ‘significant’ public events to monitor the content of posts during and after these events within the 6-month period; we also traced gender differences in post sentiment and content over this timeline. Finally, we examined the keywords used in posts that included hashtags #WuhanDiary or #WuhanLockdown and disaggregated by user’s gender. We sought to understand the association between the most frequently used keywords and gender during the 6-month period. In total, the final analysis included 62 297 posts (#WuhanDiary: 34,237; #WuhanLockdown: 28,060). Below, we present our results on the number of posts between men and women users referencing #Wuhandiary or #WuhanLockdown; we compare the negative sentiment expressed in posts by gender over the 6-month timeline; and we examine the keywords associated with #Wuhandiary and #WuhanLockdown and gender.
We acknowledge the limitations of this dataset. The sample is representative and does not reflect the diversity of all Weibo users. However, the two hashtags were chosen to collect specifically the social media posts that represented public response to the unprecedented large-scale public health intervention–entire city lockdown in Wuhan in mid-January 2020–during the time when the unprepared public had very limited knowledge about COVID-19, that is, unknown infectiveness, mortality and no vaccine/treatment. This scenario highlights the gender difference of Chinese people under isolation and anxiety in the crisis. The two terms were selected to represent the hashtag campaign launched and popularised by the public and the social media platform with dedicated purposes to encourage individual’s help-seeking, coping, and emotional expression via social media at that very moment.23–27 It is common practice to draw representative samples of Weibo posts rather than filter and analyse an entire Weibo dataset.23 User registration is another limitation in the sample. People who self-identify as male or female for Weibo registration (only two options provided) is the basis of the analysis. We recognise that this binary registration requirement has limitations on our findings as it excludes other gender identities and their experiences.