Understanding the Influence of eWOM Behavior as Reflected in Online Opinion Leaders & Online Forums via A Mixed Methods Approach

• Main Authors: Yang, Chiao-Wen, 楊喬文
• Other Authors: Lai, Chih-Hui
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/vcv9b5

碩士 === 國立交通大學 === 客家學院傳播與科技學系碩士班 === 107 === The objective of the current study is based on Diffusino of Innovation Theory, combined with Two-Step Flow Theroy and Social Influence Theory to understand how different sources of opinions, online opinion leaders and online forums, predict electronic word of mouth (eWOM) and purchase intention. Additionally, this study explores the structural patterns of eWOM. Situated in the context of cosmetic product consumption, the study employed a mixed-methods approach, consisting of focus groups (N=15), online survey (N=682), and automatic semantic network analysis of the posts collected from PTT and Dcard (2017/05 -2018/04, N=29,359). The results of the survey data showed that, in both models with online opinion leaders and online forums, the source of eWOM credibility and eWOM seeking behavior significantly predicted customers’ purchase intention. In the meantime, the results from the semantic network analysis revealed different patterns in terms of the topics in different online forums. For example, luxury cosmetic brands were discussed more on PPT whereas open-shelf cosmetic brands and generic information exchanges were more popular on Dcard. Together, this study combines the examination of two different sources of opinions through both self-reported and behavioral data, which enriches the existing Diffusion of Innovation Theroy and Two-Step Flow Theory. It also provides an alternative structural angle to the existing literature of eWOM. Based on these findings, this study offers practical suggestions as to how to maximize the effects of eWOM by taking into account the needs and communication patterns among consumers.