The IRMA Community
Newsletters
Research IRM
Click a keyword to search titles using our InfoSci-OnDemand powered search:
|
Shopbot Market Coverage
Abstract
The increased availability of pricing data at shopbot websites has resulted in a proliferation of empirical researches on pricing behavior in electronic markets. The data obtained from these shopbots have been extensively used to make inferences on a variety of vendor pricing behaviors in many product markets. Some such studies have used multiple shopbots to increase market coverage (Kauffmann and Wood 2000, Clay et al. 2001). However, because the sampling process is not random, the effect that increasing coverage has on sample representativeness is unclear. As shopbots become popular data agents for making inferences on Internet pricing behavior, it is important to question: “To what degree do individual shopbots represent a market being studied?” To answer the question rigorously, we present a large scale study in which we tracked 459 books offered for sale by 84 online vendors as reported by eight shopbots for a four month time period (between March and July of 2002, totaling 2.2 million price observations. The books could be categorized as “Bestsellers” (hardcover and paperback), “Classics,” and “New Releases”. Bestsellers were selected from the New York Times best seller list, classic books (whose text is now public domain, such as A Tale of Two Cities and The Count of Monte Crsito) were selected from litrix.com and new releases were selected from the “Coming Soon” area of randomhouse.com.
|
|