TCA Letters to the Editor
Press Release
Title: Letter to Editor re: Web of Science (2004 version) and Scopus
Date: February 2005
Organization: New Jersey Institute of Technology
Letter:
Dear Editor,
I am the University Librarian and my colleague is a Reference Librarian at New Jersey Institute of Technology (NJIT) and we have read the article “Web of Science (2004 version) and Scopus” in the January 2005 issue of The Charleston Advisor. We think that it presents many valuable points and is useful, notwithstanding a few nit picky errors. However, one might conclude inaccurately from the article that WOS is a better product for all universities. We believe that Scopus, indeed, is better for many universities. The best way to evaluate both databases is to do it as we did at NJIT, and that means obtaining simultaneous demonstration access for three or more months to both databases and give them the local “driving” test. We had our faculty and librarians conduct searches in all of our NJIT disciplines, and carefully looked at all needs. This required a fairly large effort but was worth it since the price of both databases was significant. When NJIT selected Scopus, we did so because, in certain areas, the coverage was better than WOS, because the usability was far better, and because the long term (1945+) WOS retrospective backfiles were less important and far too costly overall to most NJIT needs.
In an ideal university world, where money is no object, and where a university has almost every known human discipline represented, then perhaps you would subscribe to both databases. There definitely were some subject disciplines better covered with titles in Scopus, and others in WOS, so any institution choosing just the WOS or Scopus would not have the complete picture. This was certainly a point to which the authors (David Goodman and Louise Deis) must have agreed when they said, “Quick summary in one sentence: keep Web of Science and buy Scopus if you can-once the publisher gets the data loaded”.
We would assume from the 1.5 stars given Scopus for coverage, that it was inferior in every way compared to WOS but that is just not true. Quality is determined by all of the criteria as weighted to meet your own local needs. It does not matter that a “sample university” (page 9) has more item counts in one or more subject disciplines, what matters is whether a database has more citations in the subject disciplines that meet your university needs.
Cost is more of an issue, no doubt, for NJIT, than Princeton. However, when cost is an issue, the institution must choose the best affordable database that meets local user needs. In short, cost is given a higher weight at some institutions. For example, at least one real ARL library, (not a small library as in p. 20 paragraph 2) has only been able to afford a 10 year WOS backfile which is roughly the same time scope coverage as Scopus. Therefore, there is no backfile advantage (dates of coverage) to that ARL institution’s purchase decision which was weighted as one star. This clearly contradicts the statement made in that paragraph that “;there is probably no ARL library for which it would be reasonable.” It’s not just small libraries that might benefit from Scopus.
The weighting system of stars in this article is certainly subjective given the NJIT experience and criteria, and, we assume, that many readers of your article will arrive at a totally different set of criteria and weights. What works for one ARL library may also not work for all libraries, even all ARL libraries.
Richard T. Sweeney
University Librarian
R.W. Van Houten Library
New Jersey Institute of Technology
University Heights, Newark, NJ 07102-1982
sweeney@njit.edu
Haymwantee Singh
University Librarian
R.W. Van Houten Library
New Jersey Institute of Technology
University Heights, Newark, NJ 07102-1982
singhh@adm.njit.edu
