Theme 1 | Social Media: Human Behavior and Response
| 23 AUGUST 2021 (Monday) - Singapore Time | |
| 14:00 - 15:30 | THEME 1 | SOCIAL MEDIA: HUMAN BEHAVIOUR AND RESPONSE |
| Moderator: Feng Qiushi | Methodological Collaborations to Achieve Meaningful and Practical Insights on User-Group Discussion and Behaviours in Social Media
Elmie Nekmat | Department of Communications and New Media, NUS |
| Trivium: A Custom-built Experimental Platform to Inform Social Media Design for Better Conversation Health
Kokil Jaidka | Department of Communications and New Media, NUS |
|
| Understanding Data-driven Campaigns in an Underdeveloped Region: Evidence from the 2020 Bihar Election campaigns, India
Taberez Neyazi | Department of Communications and New Media, NUS |
|
| Social Media Communication about HPV Vaccine in China: A Study Using Topic Modeling and Survey
Jiang Shaohai | Department of Communications and New Media, NUS |
|
Abstracts
Methodological Collaborations to Achieve Meaningful and Practical Insights on User-Group Discussion and Behaviours in Social Media
Associate Professor Elmie Nekmat
Scientists, industry practitioners, and policymakers are now faced with the unprecedented ability to collect large amounts of information and data on social media uses, users, and their digital communication/interaction patterns. Following which, they are more often than not faced with the problem to produce meaningful knowledge and practical outcomes from such large datasets – which signals the need for greater cross-methodological and cross-disciplinary collaborations. In this talk, I will share the promises and challenges associated with such cross-methodological collaborations. I will do so by drawing from an international collaboration project that had combined big data mining with other more traditional methodological approaches and analyses in the field of communication and media. A qualitative, semantic analysis was applied on 30,796 unique user postings on Weibo to determine how Chinese users create collective memories surrounding three national crises that affected the country.
Trivium: A Custom-built Experimental Platform to Inform Social Media Design for Better Conversation Health
Assistant Professor Kokil Jaidka
Social media platforms are engineered to maximize user engagement on their websites. However, greater user engagement has had unanticipated trade-offs, because social media platforms are increasingly turning toxic. Incivility on social media threatens the viability of online social networks, the inclusivity of online deliberative public spheres, and the rights of social media users – but beyond that, its effects can shake even the stability of modern democracies.This project focuses on informing the design of social media platforms with the objective to improve the health and quality of ensuing online discussion. Experiments on a custom-built platform manipulate the technological affordances that enable or constrain certain kinds of communication. For instance, anonymity is multi-faceted and afforded through different personal and social cues on social media. Personal anonymity in social media encourages minority perspectives but is seen to also increase online toxicity and bullying. On the other hand, social anonymity can reduce the perception of social media as a divisive space, and perhaps reduce affective polarization. I present Trivium, a new custom-built platform designed to test these ideas and hypotheses centered on the relationship of technological design with elicited user behavior, through opt-in online experiments with social media users. I discuss the larger implications of this tool towards building a richer understanding of the role of social media platforms in digital democracies. This work is in collaboration with the University of Pennsylvania and is funded by a Nanyang Presidential Postdoctoral fellowship.
Understanding Data-driven Campaigns in an Underdeveloped Region: Evidence from the 2020 Bihar Election campaigns, India
Assistant Professor Taberez Neyazi
Political campaigns are increasingly deploying user data to micro-target, mobilize and persuade voters, but the data does not come from a single source. In India, as elsewhere, a large amount of data continues to be collected offline, using surveys and through community interactions, in addition to using data from the census and past voting behavior. By deploying computational power, political actors have been able to process data, gather feedback and rework campaigns at a speed previously not possible. This presentation draws upon the study of a subnational election in an underdeveloped state of Bihar 2020 to examine the prevalence and effectiveness of data-driven campaigns in mobilizing and persuading voters. Specifically, we study the Facebook campaign ads of political parties and supplement it with Twitter conversation data around campaigns and a two-wave panel survey data to show how data-driven campaigns are deployed by different political actors to target, mobilize and persuade voters. While there has been a growing belief in the power of data in the campaign, we show the need to study network effects that is generated on account of interactions between offline and online world, and help in mobilizing voters.
Social Media Communication about HPV Vaccine in China: A Study Using Topic Modeling and Survey
Assistant Professor Jiang Shaohai
The human papillomavirus (HPV) vaccine is relatively novel to people in China. Social media is becoming an important channel for learning new health information. However, limited is known about what HPV vaccine information has been disseminated on social media, and how such online information is associated with health-related behaviors in China. Following the mixed-methods approach, we first crawled 67,773 postings about HPV vaccine on Weibo, the largest microblogging website in China, and performed topic modeling to identify HPV vaccine-related topics that are prevalent on Weibo. The results showed six major topics about HPV vaccine, namely policy, guidance information, advertising, scandals, personal experience sharing, and HPV risks. Second, we conducted an online survey (n = 1,982) to investigate how scanning, seeking, and discussing the six HPV vaccine topics identified from big data analytics can affect HPV vaccine knowledge, safety concern, and vaccination intention. We documented significant impacts of social media health communication on users’ health knowledge, attitude and behavioral intention.
CONTACT US
Have a question or feedback?
Check out our FAQs or fill in the following enquiry form to reach us.
Email: fassresearchevents@nus.edu.sg