05 April 2006

Six degrees of reputation

Lengthy, but nevertheless fascinating, report of research by David Shay and Trevor Pinch, prompted by Pinch's discovery that a positive review on Amazon for a book he had co-authored was then almost replicated for another (competing) book on the same topic.

Shay and Pinch analysed Amazon book and CD reviews and found that roughly 1% were duplicates of other reviews, and that, more broadly, there were a range of reasons for bogus reviews. Some reasons were obvious: authors, their family and friends aiming to create a positive impression of a book; or people using book reviews as a way to publicise a specific agenda. But I found the most intriguing reason was the need of rated reviewers to keep their ratings high by posting multiple, positive reviews (a sort of back-firing of a mechanism intended to ensure reviewer integrity in an open field), presumably because there is an intrinsic reward (15 minutes of fame) in being a highly rated reviewer. There is more chance of a reviewer being rated highly (at least by the author and cohort) if they place a positive review than if they place a negative one.

Shay and Pinch point out if they could devise software to spot plagiarised reviews Amazon and other on-line services could certainly do the same. More generally they use this study as a demonstration of how people interpret technology and its use in new contexts.

No comments:

Post a Comment