Feedback reviews and bidding in online auctions: An integrated hedonic regression and fuzzy logic expert system approach

In online auctions, user-generated feedback reviews provide first-hand information on the trustworthiness of transaction partners to the community. To examine how the feedback reviews are taken into account of the buyers' bidding decisions and thus affect the final winning price of an auction, we thoroughly examine how buyers mentally interpret the seller's reviews and adjust the bids accordingly. With ample bidding results data from a popular auction website eBay.com, this paper adopts an integrated approach of Fuzzy Logic Expert System (FLES) model and a statistical hedonic regression model to examine the research question. In particular, we use the hedonic regression approach to select key variables, which are then entered into a FLES analysis to generate knowledge base regarding the relationships between variables such as item characteristics, auction characteristics and review scores, and the final winning price. This integrated approach combines the advantages of both methods, and also overcomes their own limitations. In addition, we also present the insights gained from bidding behaviors utilizing each of the approaches.