Attributions of Trust in Decision Support Technologies: A Study of Recommendation Agents for E-Commerce

Abstract

As organizations increasingly utilize Web-based technologies to support customers better, trust in decision support technologies has emerged as an important issue in online environments. In this study, we identify six reasons users trust (or do not trust) a technology in the early stages of its use by extending the theories of trust formation in interpersonal and organizational contexts to that of decision support technologies. We study the particular context of decision support technologies for e-commerce: online recommendation agents (RAs), which facilitate users' decision making by providing advice on what to buy based on user-specified needs and preferences. A laboratory experiment is conducted using a multimethod approach to collect data. Both quantitative data about participants' trust in RAs and written protocols that explain the reasons for their levels of trust are collected. A content analysis of the written protocols identifies both positive and negative trust attributions that are then mapped to six trust reasons. A structural equation modeling analysis is employed to test the causal strengths of the trust reasons in explaining participants' trust in RAs. The results reveal that in the early stages of trust formation, four positive reasons (i.e., knowledge-based, interactive, calculative, and dispositional) are associated with higher trust in RAs and two negative reasons (i.e., calculative and interactive) are associated with lower trust in RAs. The results also demonstrate some distinctive features of trust formation with respect to decision support technologies. We discuss the research and practical implications of the findings and describe opportunities for future research.