Citation analysis for research evaluation: Citation analysis for research evaluation Wouter Gerritsma Information Specialist, Wageningen UR Library


Programme: Programme Introduction Citation data Benchmarking Journal quality Ranking University On publishing


Why research evaluation/citation analysis?: Why research evaluation/citation analysis? Research assement of graduate schools (6 year cycle, stipulated by the VSNU) Other research groups are reviewed in a similar way Benchmarking becomes more important Grant applications Job positions In the press Sooner of later, you will be subjected to such an analysis


Objective of this course: Objective of this course How are citation analyses performed, and what are their shortcomings or weaknesses? What can you do to enhance your results


Citation data: Citation data


Citation data: Citation data Paid subscription (wide coverage) Web of Science Scopus Free systems broad coverage Google Scholar Paid subscription (specialized coverage) SciFinder Psychindex Free systems (specialized coverage) Citeseer Citebase


Web of Science: Web of Science Thomson Scientific Founder Eugene Garfield Formerly: Institute of Scientific Information (ISI) Indexes articles of about 9000 journals, standard bibliographic information + reference lists Sell the same data in various databases to libraries and others (CC, WoS, JCR, ESI) They (still) have monopoly position for citation data; they are the Golden Standard


CC, WoS, JCR, ESI: CC, WoS, JCR, ESI CC: Current Contents, weekly updated bibliographic database covering the top of Sciences (no citation data) WoS: Web of Science, weekly updated database, that includes references and citation data as well (slower than CC) JCR: Journal Citation Report, yearly analysis of WoS data, presenting perfomance measures for journals. ESI: Essential Science Indicators, bimonthly updated analytical database giving performance measures for Countries, Institutes, Scientists and Journals next to research fronts, hot papers and baselines.


Database Coverage WoS/JCR/ESI/CC: Database Coverage WoS/JCR/ESI/CC Source: Moed (2005)


Scopus: Scopus (New) product of Elsevier Science Covers 15,000 serials Not all of them peer reviewed Coverage of citation data only from 1996 -> First analytical database: SJR http://www.scimagojr.com/


Google Scholar: Google Scholar Free database Serious concerns about data quality Unknown source base Very labour intensive to extract data But please have a look at http://www.harzing.com/pop.htm


Why Scopus doesn't add substantially to the number of citations found in WoS : Why Scopus doesn't add substantially to the number of citations found in WoS Source: http://www.wowter.nl/blog/2007/01/why-scopus-doesnt-add-substantially-to.html


Searching WoS: Searching WoS General search: Searches bibliographic data: Title, Abstract, keywords, Authors, Addresses Not all citations are found due to errors in citing, indexing etc. Cited reference search: searches in the references of source data Not all articles are found. Non cited articles are missing


Exercises !: Exercises !


Benchmarking: Benchmarking


Benchmarking: Benchmarking Essential Science Indicators (ESI) h-index


How do we compare numbers?: How do we compare numbers? Scientist Z. Math has a publication from 1996 with 17 citations Scientist M. Biology has a publication from 2003 with 24 citations


Baseline mathematics: Baseline mathematics


Baseline Molecular Biology: Baseline Molecular Biology


Essential Science Indicators (ESI): Essential Science Indicators (ESI) Analytical database of SCI, covering 10 years + current year building Comparison between Countries, Institutes, Scientists and Journals Hot papers / Highly cited papers Research fronts Baselines


Publication age : world averages citations (1): Publication age : world averages citations (1)


In practice: In practice


Relative impact of a research institute: Relative impact of a research institute


Evaluation of candidates : Evaluation of candidates


h-index: h-index A scientist has index h if h of his/her Np papers have at least h citations each, and the other (Np − h) paper have no more than h citations each.


h-index: h-index h-index tries to find a balance between productivity and citation impact Only published in 2005, has made a substantial impact in the world of bibliometrics. Implemented in Spires, WoS (and soon) in Scopus Applicable to authors, journals, research groups, compounds, subjects etc…


Exercises !: Exercises !


Journal quality: Journal quality


Journal quality: Journal quality JCR (Impact factors) JSR ESI h-index Or pagerank ?


Warning !: Warning ! It has often been attempted to use Journal Impact factors in some or other way to judge the performance of a group or individual scientists. This is proven very bad practice!


Impact factor: Impact factor Performance measure for journals “… it is also used for assessment of the quality of individual papers, scientists and departments. For the latter a scientific basis is lacking, as we will demonstrate in this contribution” (Opthof, 1997) Opthof, T. (1997). Sense and nonsense about the impact factor. Cardiovascular Research 33(1): 1-7. http://dx.doi.org/10.1016/S0008-6363(96)00215-5


Journal Citation Reports (JCR): Journal Citation Reports (JCR) Reports 3 measures Impact factor Immediacy Index Cited half life Cited half-life 50% citations 50% citations 3 1 Immediacy index Window Impact Factor Window


Definition of Impact factor: Definition of Impact factor The impact factor of a journal J in year T is defined as follows: “The number of citations received in year T by all documents published in J in the years T-1 and T-2” ÷ “The number of citable documents published in J in the years T-1 and T-2”


Citable items: Citable items Cited items (nominator) include: Research articles, review arrrticles, notes, letters, editorials, news items, corrections and meeting abstracts Citable items (by definition) Articles, notes and reviews


Lancet (2002): Lancet (2002) JCR Like IF = (b)/(a) = 10.2


50 % of articles generate 90% of all cites: 50 % of articles generate 90% of all cites Seglen, P. O. (1997). Why the impact factor of journals should not be used for evaluating research. BMJ 314(7079): 497-502. http://bmj.bmjjournals.com/cgi/content/full/314/7079/497


Some other points: Some other points IF measures citation impact of articles in 2nd & 3rd year after publication. Therefore biased towards journals showing rapid maturing. Reference practices, particularly the number of references per article and their age distribution, vary considerably among subfields. IF between subfields can not be compared.


When do you make use of IF: When do you make use of IF Answering the questions where do you want to publish your manuscript. Other questions, to be asked Reach colleagues ? In which article indexes included? Open access?


Exercises !: Exercises !


Ranking universities: Ranking universities Popular lists Shanghai Jiao Tong University THES Newsweek ISIhighly cited ESI Other reports EC, NOWT, OECD, USA


Winding down: Winding down


What’s in a name?: What’s in a name? A.G.J. Voragen or A. Voragen or F. Voragen or F.A.G.J. Voragen or F.G.J. Voragen