Disaster Misinformation and Its Corrections on Social Media: Spatiotemporal Proximity, Social Network, and Sentiment Contagion

Wei Zhai*, Hang Yu, Celine Yunya Song

*Corresponding author for this work

Research output: Contribution to journalJournal articlepeer-review

2 Citations (Scopus)

Abstract

Misinformation disseminated via online social networks can cause social confusion and result in inadequate responses during disasters and emergencies. To contribute to social media-based disaster resilience, we aim to decipher the spread of disaster misinformation and its correction through the case study of the disaster rumor during Hurricane Sandy (2012) on Twitter. We first leveraged social network analysis to identify different types of accounts that are influential in spreading and debunking disaster misinformation. Second, we examined how the spatiotemporal proximity to the rumor event influences the sharing of misinformation and the sharing of corrections on Twitter. Third, through sentiment analysis, we went further by examining how spatiotemporal and demographic similarity between social media users affect behavioral and emotional responses to misinformation. Finally, sentiment contagion across rumor and correction networks was also examined. Our findings generate novel insights into detecting and counteracting misinformation using social media with implications for disaster management.
Original languageEnglish
Pages (from-to)408-435
Number of pages28
JournalAnnals of the American Association of Geographers
Volume114
Issue number2
Early online date21 Nov 2023
DOIs
Publication statusPublished - 7 Feb 2024

Scopus Subject Areas

  • Geography, Planning and Development
  • Earth-Surface Processes

User-Defined Keywords

  • disaster management
  • misinformation
  • social networks
  • social media
  • sentiment contagion

Fingerprint

Dive into the research topics of 'Disaster Misinformation and Its Corrections on Social Media: Spatiotemporal Proximity, Social Network, and Sentiment Contagion'. Together they form a unique fingerprint.

Cite this