Scholarly communication in the social sciences differs from that in the pure sciences. Social scientists publish more often in monographs than journals, when compared to fundamental and applied science researchers. Monographs and their references are not systematically indexed in databases. It is estimated that journal articles account for 45-70% of research output in the social sciences, depending on the discipline (1). As a result, citation studies in these fields require additional care since they can give an incomplete and inaccurate representation of research output if they focus only on journal articles.

Professor Charles Oppenheim, an information scientist for 40 years, and currently Head of Information Science at Loughborough University, UK, has found, however, that there is a continuing trend in the social sciences to publish increasingly in journals. “After World War II, science was seen as successful, a paradigm: it cured diseases, created energy supplies and so on,” he says. “Social sciences felt a bit like Cinderella; they were left out of the funding and grammar of science. Subconsciously, social scientists thought that if they aped pure science, one way of which was to publish in journals, then they might be able to get a larger slice of the funding pie.”

Assessing social sciences output

He continues, “A much more conscious reason is things like the Research Assessment Exercise in the UK, the principal method by which university research funding decisions have been made since 1986. The RAE typically requires each individual who is being returned by a university for consideration to identify four of his/her publications for evaluation. If you’re working on a monograph between each assessment – which takes place roughly every four years - you won’t have four papers available, and producing four monographs in that time is unrealistic.” The RAE will take place in its present form for the last time this year, and it is expected that future assessments will be based, at least in part, on bibliometrics. This will make citation counts increasingly important for all the sciences.

So how does one analyze research output in the social sciences? In 2006, the Economic and Social Research Council (ESRC), a research funding and training agency in the UK, asked Professor Oppenheim to help it answer this question. “The ESRC was under pressure from the British government to come up with a measure of the quality of social sciences research conducted in the UK, compared to research done elsewhere. It could find no single database that supplied this information and so asked me to conduct research into the databases available and suggest which one would be the best to use for this study,” says Oppenheim. “Until quite recently, Thomson’s Web of Science (WoS) was the only credible database which had reasonable social sciences coverage and provided citation indexing. In the last years, CSA Illumina, Google Scholar and Scopus have also entered the market, offering a similar service. My research thus covered these four databases. Their holdings and citation records were assessed against two sets of data: one from the 2001 RAE, the other from the International Bibliography of the Social Sciences, a bibliography managed by the London School of Economics and Political Science.”

Analyzing the results

The results of the research have since been published in the Journal of Informetrics (2). They suggest that of the four databases studied, WoS and Scopus offer the best social sciences coverage at journal, article and cited reference level. Both have a comprehensive ‘cover-to-cover’ indexing policy, although Scopus’ coverage only captures references for documents published after 1995. In citation searches carried out for records published after 1995, Oppenheim found that there was a 5.4% advantage in Scopus’ favor. CSA Illumina fared best when it came to foreign language journal coverage.

“Despite Scopus’ limited coverage of foreign language journals, something I suggest it consider extending for goodwill purposes, my research concluded that Scopus, with good coverage and sufficient tools to analyze citation counts, is arguably the best choice of the four databases reviewed and could be used as an alternative to WoS to evaluate research impact in the social sciences.”

Professor Oppenheim’s full article and methodology can be found here.

References:

(1) Archambault, É., Vignola Gagné, É. (2004), “The use of bibliometrics in the social sciences and humanities”, Science-Metrix Report, pp.3
(2) Norris, M., Oppenheim, C. (2007) “Comparing alternatives to the Web of Science for coverage of the social sciences’ literature”, Journal of Informetrics, Vol. 1, No. 2, pp.161-169.
VN:F [1.9.22_1171]
Rating: 0.0/10 (0 votes cast)