The Persuasive Impact of High Variance Across Reviewers’ Online Ratings
(A2021-92932)

AUTHORS

Maximilian Gaerth, University of Mannheim; Neeru Paharia, Georgetown University; Florian Kraus, University of Mannheim

ABSTRACT

A growing number of customer review sites (e.g., TripAdvisor, Yelp) display the distributions across reviewers’ past ratings. While online reviewing experts (vs. novices) tend to show lower variance across their ratings (Nguyen, Wang, Li, and Cotte, 2020), the authors find that readers of online WOM incorrectly infer greater reviewer expertise from high (vs. low) rating variance. As a result, consumers are more likely to purchase a product previously purchased or recommended by reviewers with lower expertise. In addition, the authors examine the diagnosticity of rating variance as a signal of reviewer expertise relative to rating volume, and the circumstances under which rating variance becomes more diagnostic as a signal of reviewer expertise than rating volume. Lastly, the authors demonstrate that reviewers are not aware that high (vs. low) rating variance signals greater expertise, since people who are motivated to convey expertise neglect to exhibit higher variance across their ratings.