The Platform Matters: Selection and Measurement Bias in Online Reviews

Abstract

This study investigates biases introduced when soliciting customer satisfaction survey respondents to post online reviews on different platforms. We examine review nonresponse bias and measurement bias arising when customers post reviews on TripAdvisor or Google Travel. Our findings indicate that platform heterogeneity significantly impacts both the likelihood of posting a review and the ratings provided, leading to biases. Review nonresponse bias occurs when not all survey respondents post reviews, creating a selection bias. Measurement bias arises when posted ratings differ from survey responses. Our analysis reveals that Google Travel users with negative experiences are less likely to post reviews, but those who do post tend to give lower ratings compared to their survey responses. Conversely, TripAdvisor users show less review nonresponse bias, but their posted ratings are more positive when survey responses are negative. Understanding platform-specific biases can help managers adjust opinion weights to ensure reliability. Academically, this research advances the literature by distinguishing review nonresponse bias from measurement bias and encourages further comparative studies across platforms.