ALEXA, WHO TOLD YOU THIS? EXAMINING HOW MEDIA PLATFORM AND SOURCE TAILORING AFFECT USERS’ PERCEPTIONS OF INFORMATION DELIVERED BY A VIRTUAL ASSISTANT

Restricted (Penn State Only)
Author:
Kim, Jinyoung
Graduate Program:
Mass Communications
Degree:
Doctor of Philosophy
Document Type:
Dissertation
Date of Defense:
June 11, 2018
Committee Members:
  • S. Shyam Sundar, Dissertation Advisor
  • S. Shyam Sundar, Committee Chair
  • Fuyuan Shen, Committee Member
  • Michael Grant Schmierbach, Committee Member
  • Mary Beth Rosson, Outside Member
Keywords:
  • Alexa
  • information credibility
  • source tailoring
  • customization
  • virtual assistant
  • media platform
Abstract:
When we evaluate the credibility of various online information, one of the most basic yet important questions that we ask is “What is the source of information?” This question has never been successfully answered by virtual assistants such as Amazon’s Alexa that we commonly use to obtain information. This dissertation examines how users’ credibility of Alexa’s information and their psychological reactions to the virtual assistant (i.e., attitudes and behavioral intentions toward Alexa) change as a function of source attribution made by Alexa. Specifically, the current dissertation empirically tests whether different types of media platform (online news vs. forum sites) and source tailoring (customization vs. personalization) influence how users perceive Alexa’s information and the virtual assistant by using a lab experiment (N=178). Results show that users perceive Alexa’s information as more credible when she reveals the source. When users receive Alexa’s information from news websites, they are likely to perceive gatekeepers of the news websites as experts in the fields (i.e., high source expertise) than those from forum websites. Such different psychological reactions to Alexa’s sources in turn affect their perceived credibility of Alexa’s information. Moreover, users who customize Alexa’s sources perceive greater user control over their interaction with Alexa, self-identity projected on to the Alexa’s online interface, and involvement in Alexa’s information, compared to those who receive a personalized list of Alexa’s sources or those who do not experience any kind of source tailoring. These are in turn positively associated with their attitudes, behavioral intentions toward, and trust in Alexa. The key findings of this dissertation hold theoretical implications for research on source orientation in human-computer interaction, information credibility, and user psychology, as well as practical implications for product designers and researchers.