A Study on Determinants of Intention to Reuse Recommendation Agents from Information Adoption Model--- The Moderating Effects of Shopping Goals and Domain Knowledge

Online recommendation agents can help users reduce information overload and improve users’ decision quality by providing shopping assistance. Yet, many online shoppers have negative reaction to recommendation agents and stop use them or have no motivation to use them. The main reasons may be due to users have no idea of how to evaluate recommendation agents or of whether users can achieve their shopping goal with less effort. We think that “information” is fundamental elements of recommendation agents. However, it is worth paying attention that what information should be provided and could persuade users, and could enhance users’ acceptance and confidence in recommendation agents. We think this issue is critical to understanding how to improve users’ intention to reuse. This study develops a research framework derived from the perspectives of information adoption model and accessibility-diagnosticity model to investigate the research questions. We propose that explanation facilities and perceived similarity are important determinants of users’ intention to reuse RAs. Recommendation agents could build a cognitive and emotional connection with users through proper design explanation facilities and then persuade users into adoption information to improve their capability of product evaluation and intention to reuse. This study also considers the influence of personal characteristics in shopping goods on how users deal with and accept proposed information by recommendation agents, such as product domain knowledge and shopping goals. For example, novice shoppers, who are unfamiliar with attributes and functions of product, are more moved by explanation of importance and purpose of RAs’ questions and perceived decision outcome similarity than explanation of the whole logic process and perceived decision process similarity. Shoppers with hedonic motivations are more likely accept RAs’ suggestion when they have perceived RAs’ care about what users want. This study will conduct a laboratory experiment with 2*2*2 factorial design to collect data. This study will investigate the relationship of explanation facilities, perceived similarity, information diagnosticity, and intention to reuse, as well as the moderating effects of product domain knowledge and shopping goals. Based on our results, several theoretical and managerial implications will be proposed.

Project ID:PF9907-7887
External Project ID:NSC99-2410-H182-020