Judgments in the Sharing Economy: The Effect of User-Generated Trust and Reputation Information on Decision-Making Accuracy and Bias

Abstract

The growing ecosystem of peer-to-peer enterprise - the Sharing Economy (SE) - has brought with it a substantial change in how we access and provide goods and services. Within the SE, individuals make decisions based mainly on user-generated trust and reputation information (TRI). Recent research indicates that the use of such information tends to produce a positivity bias in the perceived trustworthiness of fellow users. Across two experimental studies performed on an artificial SE accommodation platform, we test whether users' judgments can be accurate when presented with diagnostic information relating to the quality of the profiles they see or if these overly positive perceptions persist. In study 1, we find that users are quite accurate overall (70%) at determining the quality of a profile, both when presented with full profiles or with profiles where they selected three TRI elements they considered useful for their decision-making. However, users tended to exhibit an "upward quality bias" when making errors. In study 2, we leveraged patterns of frequently vs. infrequently selected TRI elements to understand whether users have insights into which are more diagnostic and find that presenting frequently selected TRI elements improved users' accuracy. Overall, our studies demonstrate that - positivity bias notwithstanding - users can be remarkably accurate in their online SE judgments.

Keywords: accuracy; bias; digital identity; reputation; sharing economy; trustworthiness; user judgment; user-generated content.