The normalized citation impact of a publication is not available in the Web of Science as such, but in InCites (build on top of the Web of Science) to which the UM subscribes. In InCites, the normalized citation impact indicator is called the Category Normalized Citation Impact (CNCI). In Scopus (and SciVal build on top of Scopus – to which the UM does not subscribe), the normalized citation impact is called the Field-Weighted Citation Impact (FWCI). In Scopus, the FWCI is only available for individual publications, including for books and chapters when indexed.
How to interpret the CNCI or FWCI
The CNCI and the FWCI are unbiased indicators of impact irrespective of age, subject focus or document type, although they only become stable, two years after publication. For very recent publications, use other indicators like Percentile indicators that are not mean-based.
- In both cases, an FWCI or CNCI of 1 indicates that the publication has been cited at the world average for similar publications.
- If greater than 1, then the publication has been cited more than would be expected for the world average for similar publications. For example, an FWCI or CNCI of 2, indicates that the article has received double the citations than the world average for similar publications.
- If less than 1, this indicates that the publication has been cited less than would be expected based on the world average for similar publications.
- If the FWCI or CNCI is absent, this indicates zero citations.
When averaged over a set of publications one speaks of a Mean Normalized Citation Impact or Mean Normalized Citation Score (MNCI or MNCS), which can be strongly influenced by outliers as the distribution of citations across publications is often highly skewed *. Some document types – such as book reviews – are hardly cited, but when they are, they can receive very high CNCIs or FWCIs that can skew impact analyses. This is the reason why in citation impact analyses only articles and reviews are included when analyzing sets of documents. Other documents types should be analyzed as separate categories. Please note that in InCites, the mean normalized citation impact, or score, is not called MNCI or MNCS but that it carries the same name: CNCI.
Many scholars find that the FWCI and CNCI metrics for the same publication are different – sometimes significantly so. This is due to differences in:
- database coverage between Web of Science and Scopus, affecting the number of citations indexed in each system
- how subject areas are defined ** and assigned to publications
- how the CNCI or FWCI is calculated for a publication that is assigned to more than one subject area ***
Dimensions uses a normalized indicator as well, called the Field Citation Ratio (FCR). It is calculated in a similar way as the others, but is based on a different subject-area classification system. However, the calculation does not differentiate articles from reviews leading to an inflated value. It is recommended not use the FCR as a citation impact indicator.
- How to acquire normalized citation indicators from InCites
- Using InCites Responsibly: A Guide to Interpretation and Good Practice
- Using SciVal responsibly: a guide to interpretation and good practice
- The Leiden Manifesto for research metrics
* A reason why averages or sums of citations measured over more than one output, are often not allowed (e.g. in NWO grant applications).
** Both systems use a journal-level based subject-area classification system (Web of Science Subject Categories or the Scopus All Science Journal Classification (AJSC)), meaning that the subject area of a scholarly work is based on the journal in which it is published and may be inaccurate for some outputs.
Percentile indicators (Web of Science/ Incites and Scopus)
Percentiles are an alternative to mean-based indicators such as the FWCI or CNCI. A percentile indicates how a paper has performed relative to other indexed publications in its field, year, and document type, and is therefore a normalized indicator.
The percentile for a publication or in subject area is determined by creating a citation frequency distribution for all the publications in the same year, subject category and of the same document type. The percentile is the percent of items cited less often than the item of interest, and therefore a higher percentile indicates better relative performance.
Scopus displays the percentile under the bibliographic data of each record, and the Web of Science in the Author Impact Beamplot after an author search, displaying the percentiles for all records belonging to the author searched for. Further guidance:
The percentile in subject area – on author or publication level – is also available in InCites: https://vimeo.com/763350037.
Percentile-based indicators are based on citations, so they inherit all of citations’ limitations, including that a number of factors that can influence the citation rates for publications within subject areas and year of publication, including gender, the number of co-authors, the nationalities of authors, and outliers (very high or low cited publications) when sets of documents are analysed.