Factors to consider before data collection for bibliometric analyses
- are the publications to be analysed indexed in the data source, e.g. Web of Science or Scopus (the language och subject coverage should be at least 60%)
- are there enough publications produced within the research group and subject area
- is it possible to identify the publications of the reseach group to be analysed - requires knowledge of organisations and their staff
- which authors should be included in the analysis - new staff and former staff
- who will receive the result of the analysis - the interpretation requires knowledge of bibliometrics
Examples of data sources
Scopus
Natural sciences, medicine, technology, social sciences, 15,000 peer-reviewed journals, over 1,000 Open Access journals, 500 conference proceedings, 40 million references, citations from 1996. The Scopus Affiliation Identifier automatically identifies and matches an organisation with all its research output.
Web of Science
Natural sciences, humanities, social sciences, over 11,000 peer-reviews journals, 45 million references. The Swedish Research Council has acquired data from Web of Science to a publication database with 24 million references (yearly updated with 1.2 million references) and over 400 million citations. Author names and addresses are not verified or normalised. On average 7% of the matches between a reference and its cited source is missing in Web of Science. Thomson Reuters also publish ResearcherID.com - a place for scientists to manage their professional profile with automatic update of citation data. Search for ResearcherID or Top Keywords.
Google Scholar
Natural sciences, medicin, technology dominates over humanites and social sciences, scholary journals and reports but also cited non-scholary publications, unknown number of references. Free software (Publish or Perish) available for analyses of citation data.
PubMed/Medline
Medicine, biomedicine, 20 million references. Author names and addresses are not verified or normalised.
Institutional repositories
Verified data, no citation data, no world average data.. Supplementary subject indexing might be needed.
Hybrid solutions
Institutional repository and citation data from a commercial vendor. Buying data from commercial data sources (Web of Science, Scopus) give access to huge data sets. Great effort is needed to verify the references (match authors to articles, citations and subject).
Analyse and visualise
Free software available to analyse and visualise cooperation between scientists, institutions, countries
Network Workbench - analyses and visualisation of networks
Pajek
WoS2Pajek and Excel2Pajek - software used to prepare Web of Science downloads and Excel files for Pajek analysis
Cite-Space utilizes Web of Science files in text format
SCImago Journal & Country Rank - is a portal that includes the journals and country scientific indicators developed from the information contained in the Scopus database
See also e-ref > Bibliometrics > Data Sources