Details of Research Outputs

We extend the extant research on neutral recommendation agents (RAs) to those that lack recommendation neutrality and are biased toward sponsors. We first investigate the e ects of recommendation neutrality on users’ trust and distrust in RAs by comparing a biased RA with sponsorship disclosure with a neutral RA. We then apply a contingency approach to examine the e ects of sponsorship disclosure on users’ trust and distrust in biased RAs, with explanations for organic recommendations as a contingent factor. A laboratory experiment was conducted in the United States. We determine that users’ trust in the biased RA with sponsorship disclosure is lower and that their distrust is higher than that in the neutral RA. Results also show that user trust in a biased RA increases only when explanations for organic recommendations and sponsorship disclosure are both provided. Users’ perceived psychological contract violations of an RA have been verified as a key mediator of the examined e ects. However, explanations for organic recommendations, sponsorship disclosure, or their combination fail to significantly lower users’ distrust in a biased RA. A second experiment conducted in Hong Kong confirms the major findings of the experiment conducted in the United States. Theoretical contributions and practical implications for e-commerce RAs are discussed.