Word-of-Mouth, Social Learning and Network Effects as Mechanisms of Social Contagion

• Main Author: Hu, Mantian Mandy
• Format: Dissertation
• Language: English
• Published: ProQuest Dissertations & Theses 01.01.2012
• Subjects: Consumer behavior; Datasets; Marketing; Social network analysis; Social networks; Technology adoption; Word of mouth advertising
• ISBN: 1303477882; 9781303477881

Behaviors are contagious in the way that the behavior of one's peers affects the probability that s/he engages in that same behavior. Social contagion is used to denote this phenomenon. While previous literature has focused on demonstrating the existence of social contagion, the frontier of the area is moving toward understanding why it occurs and the mechanisms behind it. With this knowledge, different managerial recommendations are generated. My dissertation comprises two papers that focus on the different mechanisms of social contagion. The first one is on word-of-mouth. Utilizing a unique dataset which collects information from the automobile category on whether a consumer generates WOM to others and uses WOM for making purchase decisions, I build a discrete-choice model to simultaneously study consumer WOM generation and WOM consumption decisions and empirically answer questions which have not been explored previously. I am particularly interested in studying the key drivers of WOM generation/consumption and the synergy effect between the two WOM-related activities. The second one is on social learning and network effects. Previous literature on social contagion mainly focuses on new technology adoption, while in this study I focus on the opposite side, customers' wireless carrier switching decisions. The nature of wireless service provides me with the opportunity to use functional forms to separately identify the two mechanisms, social learning and network effects. It is the first research to incorporate a learning framework within a social network context where customers are forward-looking and decisions are endogenously made. I present an empirical application of the framework to 2008 data from a mobile network operator in a European country.