PUB - Publikationen an der Universität Bielefeld

In recent years online consumer reviews (OCRs) have developed into a popular subject of marketing research (Hu & Liu, 2004; T. Y. Lee & Bradlow, 2011). Most often OCRs are composed of free texts, pros and cons, overall (star) ratings and recommendation indicators. Therefore, the aim of this paper is to examine whether these components are correlated in a reasonable way or whether there are inconsistencies that might influence the interpretation of corresponding research outputs. Using a regression analysis framework we investigate OCRs from a popular opinion portal regarding the relationship between rating and recommendation. Furthermore, the basic effects of posted pros and cons on these two types of product evaluations (rating and recommendation) are studied. Although our research clearly supports the general appreciation of OCRs as a powerful database for modern marketing research it also documents the necessity of being aware of possible inconsistencies.