Open Access News

News from the open access movement


Sunday, February 26, 2006

Google Scholar citation counts nearly as good as Thomson's

Léo Charbonneau, Google Scholar service matches Thomson ISI citation index, University Affairs, March 2006. (Thanks to Dean Giustini.) Excerpt:
The free Google Scholar service does as good a job as Thomson ISI’s science citation index for performing citation counts and could be used as a cheap substitute to the costly Thomson service, says a University of British Columbia professor. Thomson’s citation databases are accessible through the company’s Web of Science portal only by subscription, which can cost a university tens of thousands of dollars a year. Daniel Pauly, director of the Fisheries Centre at UBC, and Konstantinos Stergiou, of Aristotle University of Thessaloniki in Greece, compared the two methods using 114 papers from 11 disciplines published between 1925 and 2004. For papers published before 1990, the authors found that the citation counts were proportional. In other words, if Thomson ISI found that a particular paper was cited 10 times as often as another, Google Scholar found the same ratio. However, for these older papers, the actual citation counts with Google were about half that of Thomson. But, for papers published from 1990 on, not only were the citation counts proportional, the actual number of citations was nearly the same. The result is “very surprising,” said Dr. Pauly. “I didn’t expect Google to perform so well. . . . I expected some vague proportionality, but I did not expect that it would be roughly one to one.”

Update. See Gary Price's short review of Pauly's article on ResourceShelf, February 25, 2006.

Update (2/27/06). Anurag Acharya, the man behind Google Scholar, responded to a few questions from Dean Giustini about the Daniel Pauly article. Excerpt:

[Google Scholar presents] the same index and the same ranking [to all users regardless of location]....We cannot share update information, but our long-term goal is to update every day.