Predicting Opinion Leadership in
The Case of the Wisconsin Recall Election
Weiai Xu, Department of Communication, SUNY-Buffalo
Yoonmo Sang, Department of Radio-Television-Film, University of Texas –
Stacy Blasiola, Department of Communication, University of Illinois at
Corresponding author: Dr.Han Woo Park, Associate Professor, the
Department of Media & Communication, YeungNam University, South Korea
Opinion leadership in digital age is the
ability to influence online information flow.
grab others’ attention and persuade
others to maximize the attention.
On Twitter, opinion leadership means
getting your message retweeted.
Characteristics associated with opinion leaders (see Rogers, 2003):
opinion leaders are:
3. involved in specific issues or topics
4. highly regarded by others,
Are these characteristics still relevant in digital
Rogers, E. M. (2003). Diffusion of innovations (5th ed. ed.). New York: Free Press. 3
In this project, we test how opinion leadership
in Twitter relates to
1. social connectivity
3. user identity
What does connectivity means?
• Twitter forms information flow networks
• Graph: Twitter hashtag #wirecall
means a higher
What does involvement means?
• General involvement: Twitter users’ political
• Situational involvement: issue involvement,
whether a user is interested in or personally
affected by the WI recall election
Three ways to infer whether a user is involved:
• user’s geographic location (in WI or not)
• Self-disclosure of political identity on Twitter
• The content of tweets
Location info on Twitter profile
Self-disclosure of political identity on Twitter profile
Tweets indicating a hierarchy of involvement
Explicitly ask other users to engage in certain acts
Providing original feedback
Simply passing along others’ messages
Higher proportion of engaging tweets = more involved
What does identity means?
• Individual user or organizational users
(e.g. news media, advocacy groups,
government agencies, etc.)
• News media or non-news-media
Data-mining: most recent 1500 tweets every two
hours, from 5-29-2012 to 6-5-2012
1000 users randomly sampled from 8957 Twitter
users that tweeted #wirecall during the timeframe
The sampled users sent 3546 tweets containing the
• connectivity positively predicts the ability in getting
retweeted by other users.
• Involvement indicated by geographic location
positively predicts the ability in getting retweeted.
• Involvement indicated by self-disclosure of political
identity: no significant results.
• Organizational users more likely to get retweeted.
• But no significant differences in the ability to get retweets
between Twitter users of traditional news media and nonmedia.
• Characteristics associated with traditional opinion leaderships
are still relevant in Twitter communication
• the flexibility of inferring online users’ characteristics (e.g.
involvement and social connectivity) using actual behavioral
data instead of self-reported data
• Methodologically, combing network analysis with content
analysis in studying online behavior
• Combining behavior data and perception data (content
analysis + network analysis + survey)
• Connectivity in various types of networks (issue network
vs. general Twitter network)
• Non-issue specific
• longitudinal analysis
• Questions to