Thursday, 28 February 2008

Feedback on the blog so far ...

You may recall some focus groups we ran late last year on the theme of what researchers want/need from the library. One element of the feedback was the need for better communication. To that end we have set up this blog. Themes we have addressed so far are BURO; EndNote; Metrics and Resources to give it some 'critical mass'. Please have a look. We hope with your feedback we can refine the topics to meet your expectations. Don't forget to leave comments and suggestions of future posts.

Wednesday, 27 February 2008

Comparing Citation Tools

A number of studies compare citation tools, ISI Web of Science [WOS], Scopus and Google Scholar [GS]. In fact there is a sub-genre of research analysing the impact of using each of these tools to measure citations in different discipline areas. Taking one citation as an example, a simple exercise illustrates the key points.

Bakkalbasi, N., Bauer, K., Glover, J. and Wang, L., 2006. Three options for citation tracking: Google Scholar, Scopus and Web of Science. Biomedical Digital Libraries. 3 (7).

Using WOS to find who has cited this article identifies 11 citations, with 4 different versions of the publication title, in separate academic journals.

Searching for the same title using a phrase/exact match search in GS identifies 31 citations: 15 in academic journals, 4 'unpublished' articles in institutional repositories [IR's], 3 unpublished articles on the web, 3 'unclassifiable, 2 pre-prints in IR's, 1 conference paper in an IR, 1 conference paper published on the web, 1 thesis and 1 promotional item. 3 items are not in English.


  • GS searches a wide range of material compared to the carefully selected scholarly journals included in WOS, with the obvious consequences of higher citation counts from a variety of publications forms.


  • GS searches material that is published quickly on the web or in IR's, for example conference papers and pre-prints. GS is likely to record citations sooner than WOS.


  • GS searches are hard to refine, with relatively rudimentary search tools. Citations may be missed, duplicated or include false citations. Author searches are relatively ineffective. Searching for an "exact match" on the tile is the most effective method.


  • The set of data searched by GS is not pre-selected - WOS of course only includes material from selected peer reviewed titles - the results need careful checking to identify 'legitimate' citations. Even if the criteria for what is included is drawn more widely than WOS.



Note that the University does not currently subscribe to Scopus.

Links

Meho, L.I. and Yang, K., 2007. Impact of Data Sources on Citation Counts and Rankings of LIS Faculty: Web of Science Versus Scopus and Google Scholar. Journal of the American Society for Information Science and Technology, 58(13), 2105–2125.