Order Bias in Online Review Evaluation: Experimental Dataset
Description
This dataset was generated as part of an academic study examining how the display order of online reviews influences users’ perceptions of review helpfulness. The data come from a controlled laboratory experiment involving 226 MBA students, who were randomly assigned to one of three experimental conditions differing in review sequence and length. Participants evaluated the helpfulness of 20 mobile phone reviews using a 5-point Likert scale (1 = not helpful at all, 5 = very helpful). The dataset includes participant-level responses, experimental conditions, review identifiers, presentation orders, helpfulness ratings, and sentiment annotations for each review. All personal data have been anonymized. The dataset is intended to support replication, secondary analysis, and future research on cognitive biases in online review evaluation.
Files
Institutions
- Xi'an Jiaotong University School of Management