Since more and more consumers rely on online consumer reviews in their purchase decision making, we have observed the emergence of empirical studies on online reviews in different contexts. Such studies look at various dimensions surrounding online reviews; nevertheless, none of them actually gives a more comprehensive picture as to the rating differences across online rating platforms and/or across time, let alone explains such differences. This study aims to address these understudied but managerially relevant issues. In our empirical study, we focus on hotel ratings across platforms and across time (i.e., across identified change point in time), and explain such differences by review text associating with the ratings. We investigated San Francisco hotels that were rated on both Tripadvisor and Expedia, and utilized the General Inquirer dictionary to identify the categories of words in a review. Regressions analyses were then conducted for the explanation of rating differences. We found that rating differences, in both the cross-platform and the longitudinal cases, are substantially associated with the number of words (in a negative relationship). Furthermore, word tags identified from the review text associated with ratings also explain the rating differences. To conclude, the research reveals that for a specific service provider, online ratings may vary significantly in both the cross-platform and the longitudinal senses. It demonstrates that adopting a systematic approach and utilizing an established semantic dictionary, information extracted from review text can substantially explain such differences.

• Chun-Yao Huang(National Taiwan University, Taiwan)
• Ping-Yu Liu(National Taiwan University, Taiwan)