Does KFC sell rat? Analysis of tweets in the wake of a rumor outbreak
| Pages | 660-673 |
| Date | 20 November 2017 |
| Published date | 20 November 2017 |
| DOI | https://doi.org/10.1108/AJIM-01-2017-0026 |
| Author | Anjan Pal,Alton Y.K. Chua,Dion Hoe-Lian Goh |
| Subject Matter | Library & information science,Information behaviour & retrieval,Information & knowledge management,Information management & governance,Information management |
Does KFC sell rat? Analysis
of tweets in the wake of a
rumor outbreak
Anjan Pal, Alton Y.K. Chua and Dion Hoe-Lian Goh
Wee Kim Wee School of Communication and Information,
Nanyang Technological University, Singapore, Singapore
Abstract
Purpose –In the wake of a rumor outbreak, individuals exchange three types of messages: rumor
messages, counter-rumor messages, and uncertainty-expressing messages. However, the properties of the
three types of messages are relatively unknown particularly in the social media context. Hence, the
purpose of this paper is to examine these three types of messages posted on social media in the wake of a
rumor outbreak.
Design/methodology/approach –Data included tweets posted after the outbreak of a rumor that wrongly
accused the fast food chain Kentucky Fried Chicken (KFC) for selling rats instead of chicken. Using a
deductive approach, codes were derived via content analysis on the tweets. Volume and exposure of tweets
were also examined.
Findings –Counter-rumor tweets (52 percent) outnumbered rumors tweets (32 percent) and
uncertainty-expressing tweets (16 percent). Emotions and personal involvement were abundant in rumor
tweets. Expressions of credence and references to URLs were high in counter-rumor tweets. Social ties were
found widely in uncertainty-expressing tweets. The high volume and exposure of counter-rumor tweets
compared with those of either rumor tweets or uncertainty-expressing tweets highlight the potential of
counter-rumors to mitigate rumors.
Originality/value –This research ventures into a relatively unexplored territory by concurrently
examining rumor messages, counter-rumor messages and uncertainty-expressing messages in the wake
of a rumor outbreak. It reveals that counter-rumor messages have the potential to mitigate rumors on
social media.
Keywords Crisis informatics, User-generated content, Social media, Counter-rumor, Online rumors,
Rumoring phenomena
Paper type Research paper
1. Introduction
Online rumors refer to unverified messages circulating on the internet (DiFonzo, 2008).
They often become viral given their ease of dissemination through one-click social plugins
such as the retweet button of Twitter, or the share button of Facebook (Lee and Hong, 2016).
When rumors become viral, they might induce panic and anxiety. Hence, they represent a
growing concern for online users (Kostka et al., 2008).
Scholars have also been intrigued by rumoring phenomena. The context of the studies
includes cases ranging from natural disasters such as the Haiti Earthquake (Oh et al., 2010)
to terrorist attacks such as the Boston Marathon Bombing (Starbird et al., 2014).
These studies have been largely trained on examining content characteristics and
prevalence of rumor messages. Content characteristics include message features such as
credence, emotions, personal involvement, social ties, and the presence of images as
well as URLs (Oh et al., 2010, 2013; Sun et al., 2013; Tanaka et al., 2014). Message
prevalence includes estimating their volume and exposure in rumoring phenomena
(Arif et al., 2016; Starbird et al., 2014). Volume captures the quantity of messages whereas
exposure estimates the number of individuals who encounter the messages.
However, in the wake of a rumor outbreak, rumor messages do not swirl in isolation.
They are often accompanied by at least two other types of messages, namely,
counter-rumor messages, and uncertainty-expressing messages. Counter-rumor
Aslib Journal of Information
Management
Vol. 69 No. 6, 2017
pp. 660-673
© Emerald PublishingLimited
2050-3806
DOI 10.1108/AJIM-01-2017-0026
Received 18 January 2017
Revised 27 April 2017
Accepted 31 July 2017
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/2050-3806.htm
660
AJIM
69,6
To continue reading
Request your trial