Exploring bidder characteristics in online auctions
MetadataShow full item record
The recent advancements in network data modeling such as bilinear mixed models have opened doors to many other social researches that were not possible to explore earlier. In this thesis, we demonstrate an application of a bilinear mixed model for a complex human behavior such as overbidding in auctions, i.e. placing a bid of higher value than his or her preset valuation of the item.We use an innovative approach of illustrating auction data in the form of a network. The rich network framework allows us to consider bidder interdependence and examine overbidders (termed in the study as Reactors) in the presence of bidders who have influenced them to do so (termed here as Influencers). Results show that Reactors bid on fewer items, but on those with high pre-auction estimated values. They also tend to bid more in the second half of the auction as compared to the first half. Implications for the auction house managers were also presented.