An Examination of the Data Quality of Online Reviews

Abstract

This research synthesizes the information systems and marketing research by considering the usefulness of online product reviews in the context of Wang and Strong's Data Quality Framework. It examines the extent to which a review's intrinsic (review anonymity and use of personal pronouns), contextual (review length, verified purchase, rating, and rating extremity), and representational quality (spelling errors, grammar errors, readability) impact the perceived usefulness of a product review. The examination of Amazon reviews for an inexpensive experience product revealed number of words, verified purchase, and grammar errors have a significant positive impact on review usefulness. Rating and number of spelling errors have a negative effect, suggesting consumers use some discernment in assessing the believably of a review. Surprisingly, the opposite effect was found for grammar errors, with more grammar errors being associated with a more useful review.