Opinion leadership in sustainable fashion communities : an online social network inquiry

• Main Author: Orminski, Jeanette
• Other Authors: Benjamin Hill Detenber
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Social sciences::Communication::Communication theories and models; Social sciences::Communication::Public opinion
• Online Access: https://hdl.handle.net/10356/156157
• Institution: Nanyang Technological University

Fast-changing fashion trends and increased consumption has significant ramifications for people and the environment. Sustainability offers a roadmap for development promising long-term economic success without harming people and the environment. The dissertation systematically explains the linkages between fashion, people’s values and personality, sustainable consumer behavior, and online social networks. Opinion leaders have the opportunity to share information and influence the public to adopt new ideas and behaviors. The original approaches of opinion leadership and behavior change do not consider new technologies and changing communication patterns, which have obscured borders between interpersonal communication and mass media coverage on social network platforms. The dissertation addresses this issue in a multi-method approach. A qualitative discourse analysis serves as a pilot study and presents a sustainable fashion framework to define the setting of the dissertation. To overcome current limitations to identify opinion leaders in online social networks, the main focus of this dissertation builds the application of the novel method of Massive Unsupervised Outlier Detection to identify outliers as influential communicators on Twitter that differ from typical users in significant ways. To validate outliers as opinion leaders, I measure opinion leadership through two other approaches. Using traditional measures of opinion leadership, I survey self-reported opinion leadership among a sample of Twitter users from the original data set. An explorative network approach using Twitter lists provides additional insights and explains different types of opinion leaders. The results partially validate the computational method and provide a basis for future studies on opinion leadership in online social networks. In conclusion, the dissertation underscores how computational methods support communication research and sustainable fashion practice, adding knowledge to areas of political participation, opinion leadership, and online social network analyses.