Location privacy exchange behavior: development and testing of two proposed models
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Researchers studying the role of privacy perceptions in Location-Based Services (LBS) have primarily focused on information disclosure decisions at the individual transaction level, weighing the benefits of LBS against the risks of a potential privacy breach. In contrast, two of the three essays in this dissertation use a relationship commitment' conceptualization, with each transaction decision viewed as a component in a long-term process. The
privacy paradox' has been frequently observed, users disclose high levels of personal information in contrast to their low levels of intent to disclose such information. This dissertation includes measures of intent and behavior to determine relevant antecedents of actual location privacy disclosure in the context of LBS on mobile computing devices. In Essay One, qualitative research is used to discover the costs and benefits considered to be important by researchers and LBS users. While there is overlap between the decision considerations of LBS privacy researchers and users, there are also significant differences in the relative priorities each group assigns to these factors. In Essay Two, a behavior prediction model based on privacy calculus, prospect theory and Intertemporal choice theory is proposed and tested. In Essay Three, a behavior prediction model based on social exchange theory and the commitment-trust theory of relationship marketing is proposed and tested. Both of the proposed models were tested in a field experiment using a trial version of a working LBS app with 1047 participants. Both of the research models exhibit significant predictive power, particularly for users' intentions to disclose personal information in exchange for benefits provided by a LBS. The research model in Essay Two explains 39.5 percent of the total variance in mobile users' intention to disclose personal and location information, while the research model in Essay Three explains 54.1 percent. However, the privacy paradox persists as the antecedent variables account for only 9.9 percent of the total variance in actual user information disclosure. Recommendations for researchers, providers of LBS on mobile computing devices, and for LBS users are provided in Chapter Five.