Citations and citation databases

Citations are most often associated with citations in scholarly publications (books and journal articles) tracked by citation databases such as the Web of ScienceScopusDimensionsLens, and Google Scholar.[1]

There may be differences in the citation counts between the citation databases as each only counts the citations that appear in the journals and books that they index, and the scope of each database varies. The Web of Science being the most selective (in terms of the quantity of journals and disciplines covered) and Google Scholar being the least selective (indexing a great deal of non-peer-reviewed content and various research output types). Even Google Scholar is not all-inclusive as a research output has to be online available, which is not always the case. This means that you might have to use more than one citation database to find the citing documents of your research output.

Due to the diverse reasons to cite works of others (i.e. citation context) one cannot interpret citations in an unambiguous way. Nevertheless, many scholars consider citations a measure of influence amongst scholars. Especially in journal-oriented disciplines, whose journals have a good coverage in the Web of Science or in Scopus. In other disciplines where books are the main outputs, or whose journals are not well indexed in the citation databases (humanities, arts, and social sciences), citations play a minor role and other forms of output evaluation are used.

The search for citations in all the citation databases is straightforward:

  • Do a title search for your article
  • Look for ‘Times Cited’, ‘Cited by’, or ‘Citing Works’
  • Link to the list of citing sources

When using the Web of Science to find citations on a book or a book chapter you will have to use the ‘Cited Reference search’ option because books are not indexed in the Web of Science directly as a source item[2], but only findable when other Web of Science-indexed publications cites the book or chapter.

[1] OpenAlex, the successor of Microsoft Academic is not addressed because the web-interface is not (yet) available and information is only accessible through an API.
[2] Clarivate indexes books as source items in the Book Citation Index, to which the UM does not subscribe.

Who is citing?

Citations should be seen as the starting point to uncover who is citing your research, and preferably the reasons why, as these (may) point to qualitative aspects of your work.
To uncover who is citing your research output the citing publications have to be analysed further. When using Scopus or the Web of Science to retrieve citing publications (books, chapters, journal articles, or proceeding papers) they can be analysed and filtered on disciplines, authors, affiliations, keywords and more.

  • How to analyse search results (based on topic search) in Scopus view the video tutorial.
  • How to analyse research result in the Web of Science: Analyse Results

When the Web of Science is used to collect the citing publication, this set can be exported to InCites (max. 150.000 citing publications) to analyze in what other subject areas, or countries your research is cited, by which authors, and by which academic, corporate, or governmental institutes.

Citation context

Citation counts are not a direct measure of impact of an output or a measure of positive reputation for individual academics because you need to know the citation context of each citation. Citations are not always positive, and publications can have impact without citations. Nevertheless, many academics consider citations a measure of influence amongst scholars. Especially in journal-oriented disciplines, whose journals have a good coverage in the Web of Science or in Scopus.
To investigate the context of a citation presupposes access to the citing outputs, which might not always be the case, or hindered by the sheer volume of citing publications when the citation context has to be investigated manually.

Two automated systems that might help with a citation context analysis of your article are Semantic Scholar and Scite. These two systems use artificial intelligence to identify if a citation to your article supports, contradicts or just mentions your article. The Scite and Semantic Scholar databases are still growing; meaning that not all citations of every paper are available and that you may not be able to identify all citations to your paper. The systems work best if you search for a specific title or DOI.

In the Web of Science, the citation context is available in the “Citing items by classification” panel next to each publication record where the citations (linking to the citing documents) are classified as; Background, Basis, Support, Differ, and Discuss. Each citing item may mention the source article multiple times, and each mention is classified by the purpose of the citation. Note that the number of classified citations may be less than the total number of citing articles.

Normalized citation indicators

Besides the citation context, the main reason why raw citations counts cannot be used as a direct measure for the impact or influence of a research output is that the number of citations depends on its age (publication year), its subject area, and the document type (book, article, proceeding paper, review). To compensate for these differences in the citations rates of different kinds of publications, the citations need to be compared to a benchmark, called the expected or average number of citations typical for papers of the same age, of the same document type, and belonging to same subject area

For extensive information on this topic, go to the Normalized Citation Indicators subject guide.