Skip to content
Dec 19 / Clayton Hayes

CiteScore, a new journal metric from Elsevier

Before diving into CiteScore, it’s a good idea to briefly discuss the current journal metric it most closely resembles, the Impact Factor. Those of you familiar with the world of scholarly journals are surely familiar with Impact Factor, a metric which ranks scientific journals based on (roughly speaking) the average number of citations received by articles in that journal. It has been more or less an industry standard since it was first introduced by Eugene Garfield in a 1972 paper. The Impact Factor was originally based on the Science Citation Index, but now relies on citation information harvested from the Web of Science database. Below is an explanation of how a journal’s impact factor is calculated, borrowed from the library’s own guide on measuring research impact:

Impact Factor Calculation


Elsevier’s CiteScore is broadly similar to the Impact Factor, with a few key differences:

  1. CiteScore pulls data from the Scopus database and considers about 22,000 different items, about double that of the Impact Factor
  2. CiteScore pulls citation data from the three previous years, as opposed to Impact Factor’s two
  3. Impact Factor only looks at what it considers to be citable items, meaning articles or reviews. CiteScore, on the other hand, pulls citation data from any available items in the journal, including front matter
  4. CiteScore is provided free of charge, and is openly available on the web
  5. CiteScore metrics are calculated monthly, whereas Impact Factors are calculated annually

Impact Factor (and journal metrics in general) have never received total acceptance from the scientific community (with good reason), but the reaction to CiteScore has been a bit more hostile than may be expected. The openness and transparency of their methods has generally been praised, as has the fact that it is provided at no charge. Many criticize item number 3 above, the fact that CiteScore pulls citation data from any and all available documents. This can be a problem because many prestigious journals include non-citable items, like editorials, letters from researchers, or subject-specific news, which increase the number of items appearing in the journal without providing any additional citations.

Since CiteScore is calculated on a monthly basis, Eslevier hopes, perhaps, that it can provide a bit more currency in subject areas where this is important. I’m not convinced, though, that this is necessary. Monthly updates seems to be more than could be considered useful. Perhaps if a journal produces one or two very impactful articles, or if a journal adjusts is publication schedule or practices, CiteScore will reflect this a bit sooner than Impact Factor would. Aside from situations like that, most metrics change only gradually, and as many journals publish four or fewer issues a year this is to be expected.

In the end, the greatest strength of CiteScore is that it is free. Journal Citation Reports, the service that provides Impact Factor data, is an expensive subscription service that is out of reach of many. CiteScore provides an alternative that is accessible by all, and is (I think) to be commended for that if nothing else. For further discussion, see posts on the NFAIS blog and on the Scholarly Kitchen.