TCA Letters to the Editor
Press Release
Title: Scopus vs WoS: Additional Review in LJ
Date: February 2005
Organization: Palmer School of Librarianship
Letter:
In the Jan 15, 2005 issue of LJ http://www.libraryjournal.com/index.asp there is a review of Scopus and Web of Science by Cheryl LaGuardia of the Harvard University Library. She finds, for a number of topics, many more articles in Scopus than in WoS. Actually, the content is roughly comparable if comparably searched.
The author of the review examined listing for such phrases, as: spontaneously broken gauge field theories,
Web of Science searches such a search query as a query for the entire phrase, while Scopus interprets it with an "AND" between each word: spontaneously AND broken AND gauge AND field AND theories. Thus, by using the default interpretations, the reviewer found 0 results for WoS and 88 for Scopus.
To obtain comparable results, both must be searched with explicit quotation marks if one is look for the phrase, and both Scopus and WoS find zero. Alternatively,
one inserts explicit AND, then WoS finds 97 and Scopus 92.(The figures are slightly higher because we searched a few weeks later, on 2/16 and 2/17.)
To look at one more example, the reviewer searched for the better known subject: HIV protease inhibitors. In WoS, there were 1055 hits and in Scopus 5568. If one sets each to search as a phrase with quotation marks, one find 1082 in WoS and 1013 in Scopus; if one inserts the AND, there are 5461 in WoS and 5558 in Scopus.
These are not the only errors we find, but they are the most dramatic and the easiest to document unambiguously,
Fortunately, the complementary merits of the two services are so clear that we both agree in conclusion, that all large libraries should try to get both.
David Goodman (Palmer School of Library Science, LIU)
and
Louise Deis (Princeton University Library)
