Whose Story Wins on Twitter?

Abstract

The South China Sea dispute is one of the most complicated geopolitical issues of the twenty-first century. While this international conflict revolves around military and economic disputes, in today’s information age international politics also hinges on how each country presents the dispute in the news and whose “story” wins. Based on the Network Agenda Setting model, this study analyzed the news coverage in three involved countries—China, the Philippines, and the United States—and examined whose “story” gains the most prominence on Twitter, an emerging transnational public sphere. A combination of network analysis and the Granger causality test were used to explicate the media effects. Network visualization techniques were adopted to graphically represent the media network agendas. Overall, the results showed that the Twittersphere to a great extent followed the US news media in reasoning the association and dynamics between different countries involved in the South China Sea dispute, whereas the media in China and the Philippines showed minimal to moderate impact on this global social media platform. The findings demonstrate that the US media’s agenda-setting effect can transcend national borders, and that the imbalanced power structure has been reinforced in this new media environment.

KEYWORDS:

• agenda setting
• data visualization
• intermedia agenda setting
• international communication
• network analysis
• Network Agenda Setting model

ACKNOWLEDGEMENTS

The authors would like to thank Eric Kolaczyk for his consultation on network analysis, and Yue Wang for assistance with manual content analysis.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

1. While some alternative terms other than the “South China Sea” were used to refer to the sea or its regions, the newspapers selected for the analysis used the term “South China Sea.” Therefore, we also used the term to search for relevant tweets. We acknowledge that this choice would inevitably exclude relevant conversation on Twitter that used alternative terms, but it could more accurately measure the media influence, which is the goal of this study.

2. We acknowledge that, however, including non-English tweets may provide a more comprehensive analysis of the conversation on Twitter, and will allow for measuring the media effects in specific countries and regions.

3. Our final sample represents about 1.5 percent of the original data. It is important to reiterate that our original data were collected via the full Twitter firehose and the sample was generated systematically. In other words, our sample is more representative than many recent studies (e.g., Vargo et al. 2014) that used Twitter's public API, which only generates an approximately 1 percent, non-random sample of all of tweets at the time (Morstatter et al. 2013).

4. A manual content analysis of Twitter user profiles (α = 0.83) showed that 89.8 percent of tweets in our sample were sent by ordinary citizens or groups, while the rest were sent by news organizations or news aggregators from different countries around the world. Like previous research, we treated all Twitter users—media and non-media—discussing the South China Sea dispute as one international community. Future research could measure media effects on different Twitter users separately.

5. Offline journal readers should check the manuscript's online version for color versions of the figures.