Doctor of Philosophy
Transactions in the peer-to-peer sharing economy carry high risk and uncertainty. Consumers exchange with non-professional providers with whom they have no past history, and must rely on ratings and reviews for choice selection. However, there is a large positive bias in the ratings, making differentiation difficult, and causing some consumers to lose trust. Despite these concerns, little progress has been made to demonstrate the cause of the bias or how it can be fixed. I address this gap by demonstrating that consumers evaluate peer-peer experiences based on trust. This trust evaluation, in concert with network and social factors, contributes to the bias.
Research on service evaluation is often informed by the expectancy disconfirmation process (Oliver, 1980, 2010). Consumers compare a provider’s performance against prior expectations; the resultant satisfaction or dissatisfaction leads to online ratings. I demonstrate that the process works differently for peer-to-peer services; a consumer’s determination of whether a provider met expectations has an effect on ratings beyond the effect of satisfaction (Study 1). When uncertainty and risk are high, a provider demonstrates that they can be trusted by meeting a consumer’s prior expectations (Study 2). Contextual factors in peer-to-peer networks cause consumers to feel that their ratings are more important to peer providers, and that they may need to justify ratings. This elevates trust as an important driver of ratings at the expense of satisfaction, because satisfaction is more subjective and more difficult to justify (Study 3).
Consumers may give peer providers positive ratings even if performance is worse than expected. Standards of evaluation are relatively unclear for peer-to-peer services (making it more difficult to identify performance failure), and social norms of gratitude and empathy motivate consumers to forgive peer providers for unreliable service (Studies 4 and 5). Negative ratings for peer providers may result only if consumers believe that a provider caused and controlled a negative outcome, which suggests a lack of integrity (Study 6). I demonstrate that platforms can attenuate the positive bias by making ratings anonymous, by clearly defining service standards, and by increasing perceived controllability by providers for expectations and performance failure.
Summary for Lay Audience
The peer-to-peer sharing economy is growing quickly behind platforms such as Airbnb and Uber that help people rent or share their skills and belongings with other consumers. Online ratings and reviews are extremely important for consumers of peer-to-peer services because they establish trust with unknown (and mostly non-professional) providers. However, nearly all peer-to-peer ratings are five-stars, which makes it difficult for consumers to distinguish between providers. It suggests that peer-to-peer ratings may be biased, and may not reflect a provider’s true quality. I attempt to determine the cause of this positive ratings bias, and provide solutions to fix the bias.
The dissertation is comprised of six studies. I first explore how consumers of peer-to-peer services evaluate their experiences differently than consumers of traditional services. Research shows that for consumers who rent from a traditional business, their satisfaction is the main driver of the ratings decision. I show that this is not true for peer-to-peer services (Study 1). In peer-to-peer services, consumers experience higher feelings of risk and uncertainty because they are dealing with strangers. I show that when risk and uncertainty are high, a provider who meets a consumer’s expectations demonstrates their trustworthiness (Study 2).
Next, I demonstrate that the feeling of trust in the provider is directly reflected in peer-to-peer ratings and may lead to positive ratings even when performance is worse than expected. This is because peer-to-peer services have important differences compared to traditional services that cause peer-to-peer consumers to feel that they need to justify their ratings decisions, and to feel gratitude and empathy toward peer providers (Studies 3-5). This leads to high ratings even if a consumer is relatively unsatisfied, as long as the provider was relatively trustworthy. I show that peer-to-peer consumers give low ratings only if they feel that an untrustworthy provider caused and could have prevented a service failure (Study 6), but that it is difficult for peer-to-peer consumers to make these assessments. Recognizing this, I test three ways that platforms can reduce the ratings bias by reducing the perceived need to justify ratings and by making it easier to recognize service failures.
Moorhouse, Michael, "Trust and Bias in Peer-to-Peer Ratings: Why Peer-to-Peer Service Ratings are Nearly Always Positive, and How They Can be Fixed" (2021). Electronic Thesis and Dissertation Repository. 7863.