Opinion leaders on Twitter immigration issue networks : combining agenda-setting effects and the two-step flow of information
MetadataShow full item record
This dissertation focuses on opinion leaders on the Twitter issue networks to examine the two-step flow of information and agenda-setting effects. It analyzes the US immigration issue network as a case study because it is controversial on Twitter, and many elites such as lawyers and politicians who can influence others’ opinions are engaged in Twitter debates. Twitter is an important platform for active public debates, serving as a networked public sphere that can be used as a source and disseminator of information. Research has shown that communication patterns on the networked public spheres vary, including top-down, bottom-up, and side-by-side communications. This dissertation asked the following questions: (1) what is the shape of the Twitter immigration issue network? (2) who are Twitter opinion leaders, and what are their characteristics? and (3) who sets the agenda on Twitter: news media, opinion leaders, or the public? To answer those questions, this dissertation employs (1) social network analysis to identify immigration issue networks and opinion leaders, (2) hierarchical linear regressions to examine factors that can predict opinion leadership, and (3) Granger causality tests to measure the longitudinal agenda-setting effects of each group (news media, opinion leaders, and the public). The author differentiated between the retweet and mention networks because while the retweet network is intended to disseminate information, the mention network is intended to elicit responses, motivating users to participate in Twitter conversations. Through social network analysis, the author found divisions among clusters. Especially, the retweet network was classified as a polarized network and the mention network was described as a community cluster. The results of hierarchical linear regression analyses indicated that elite status, verified status, the number of followers, and individual issue involvement were common predictors of opinion leadership. The results of time-series Granger causality tests showed a mixture of top-down and bottom-up agenda-setting effects. This dissertation extends our theoretical understanding of opinion leaders based on traditional theories including two-step flow of information and agenda-setting effects. A key practical implication is that active Twitter users can be opinion leaders and can contribute to setting an issue agenda.