Theorizing the textual differences between authentic and fictitious reviews: Validation across positive, negative and moderate polarities

Purpose: The purpose of this paper is twofold: to build a theoretical model that identifies textual cues to distinguish between authentic and fictitious reviews, and to empirically validate the theoretical model by examining reviews of positive, negative as well as moderate polarities. Design/methodology/approach: Synthesizing major theories on deceptive communication, the theoretical model identifies four constructs – comprehensibility, specificity, exaggeration and negligence – to predict review authenticity. The predictor constructs were operationalized as holistically as possible. To validate the theoretical model, 1,800 reviews (900 authentic + 900 fictitious) evenly spread across positive, negative and moderate polarities were analyzed using logistic regression. Findings: The performance of the proposed theoretical model was generally promising. However, it could better discern authenticity for positive and negative reviews compared with moderate entries. Originality/value: The paper advances the extant literature by theorizing the textual differences between authentic and fictitious reviews. It also represents one of the earliest attempts to examine nuances in the textual differences between authentic and fictitious reviews across positive, negative as well as moderate polarities.