Factors influencing willingness to provide personal information for personalized recommendations

Abstract

By reducing the trouble of searching and selecting the right information, recommendation systems help users obtain information efficiently. However, recommendation systems also raise concerns over privacy issues. This paper thus examined consumers' intentions to provide personal information based on the dual calculus model which incorporates two interrelated trade-offs. An online survey targeting those who have experience using YouTube was conducted. Five distinct types of personal information (usage patterns, personal identifiers, biographical information, usage context, and feedback information) were selected based on an exploratory factor analysis. The results imply that perceived benefits, convenience, and vulnerability do not have significant effects on intention to disclose information. Perceived severity, however, had a significant negative effect on intention to disclose all types of information. Coping efficacy showed positive effects on willingness to disclose personal data except for feedback information. Self-efficacy, on the other hand, had a positive effect only on disclosing feedback information. This study expands the scope of discussion on personal information by reflecting the recent expansion of the collection and use of information for recommendation systems. It also sheds light on how to identify and handle each type of information differently.