Methodology

CWTS Journal Indicators offers a number of bibliometric indicators on scientific journals. These indicators have been calculated based on the Scopus bibliographic database produced by Elsevier. A key indicator offered by CWTS Journal Indicators is the SNIP indicator, where SNIP stands for source normalized impact per paper. The original version of the SNIP indicator was developed by Henk Moed in 2009 and is documented in a scientific paper (an open access preprint is available here). In 2012, SNIP was revised, leading to some changes in the way it is calculated. These changes are explained in another paper (an open access preprint is available here).

Indicators

CWTS Journal Indicators currently provides four indicators:

• P. The number of publications of a source in the past three years.
• IPP. The impact per publication, calculated as the number of citations given in the present year to publications in the past three years divided by the total number of publications in the past three years. IPP is fairly similar to the well-known journal impact factor. Like the journal impact factor, IPP does not correct for differences in citation practices between scientific fields. IPP was previously known as RIP (raw impact per publication).
• SNIP. The source normalized impact per publication, calculated as the number of citations given in the present year to publications in the past three years divided by the total number of publications in the past three years. The difference with IPP is that in the case of SNIP citations are normalized in order to correct for differences in citation practices between scientific fields. Essentially, the longer the reference list of a citing publication, the lower the value of a citation originating from that publication. A detailed explanation is offered in our scientific paper.
• % self cit. The percentage of self citations of a source, calculated as the percentage of all citations given in the present year to publications in the past three years that originate from the source itself.

In the calculation of the above indicators, only publications that are classified as article, conference paper, or review in Scopus are considered. Publications of other document types are ignored. Citations originating from such publications are ignored as well. Furthermore, citations are not counted if they originate from special types of sources (referred to as non-citing sources), in particular trade journals and sources with very few references to other sources (which includes many sources in the arts and humanities). For more details on the distinction between citing and non-citing sources, please see Section 4 of our paper.

Stability intervals

IPP and SNIP are provided with stability intervals. A stability interval reflects the stability or reliability of an indicator. The wider the stability interval of an indicator, the less reliable the indicator. If for a particular source IPP and SNIP have a wide stability interval, the indicators have a low reliability for this source. This for instance means that the indicators are likely to fluctuate quite significantly over time. CWTS Journal Indicators employs 95% stability intervals constructed using a statistical technique known as bootstrapping.

Differences with journal impact factor

The main differences between the indicators provided by CWTS Journal Indicators, in particular the IPP and SNIP indicators, and the journal impact factor (JIF) can be summarized as follows:

• Based on Scopus (IPP and SNIP) vs. based on Web of Science (JIF).
• Correction for field differences (SNIP) vs. no correction for field differences (IPP and JIF).
• Three years of cited publications (IPP and SNIP) vs. two years of cited publications (JIF).
• Citations from selected sources and selected document types only (IPP and SNIP) vs. citations from all sources and document types (JIF).
• Citations to selected document types only (IPP and SNIP) vs. citations to all document types (JIF).

Guidelines for interpretation

In the interpretation of the indicators provided by CWTS Journal Indicators, in particular the IPP and SNIP indicators, please take into account the following considerations:

• Review articles. IPP and SNIP do not distinguish between ordinary research articles and review articles. Review articles tend to be cited substantially more frequently than ordinary research articles. Journals that publish many review articles therefore tend to have higher IPP and SNIP values than journals that publish mainly ordinary research articles.
• Journal self citations. Some journals may try to increase their citation impact by increasing their number of self citations, sometimes in questionable ways (e.g., coercive citing). IPP and SNIP do not correct for this. However, the percentage of self citations of a journal is reported as a separate indicator in CWTS Journal Indicators. In the interpretation of this indicator, one should keep in mind that in general larger journals can be expected to have a higher percentage of self citations than smaller journals.
• Small journals. IPP and SNIP are less reliable for small journals with only a limited number of publications than for larger journals. For this reason, CWTS Journal Indicators by default displays statistics only for journals with at least 50 publications. Notice that smaller journals also tend to have wider stability intervals than larger journals.
• Skewness of citation distributions. The distribution of citations over the publications in a journal tends to be highly skewed, with many uncited and lowly cited publications and only a small number of highly cited ones. Because of this skewness, the average citation impact of a journal, as measured using indicators such as IPP and SNIP (but also the journal impact factor), is not very representative of the citation impact of individual publications in the journal. One should therefore be careful in assessing individual publications based on the journal in which they have appeared (see for instance the San Francisco Declaration on Research Assessment).
• Outliers. IPP and SNIP are sensitive to ‘outliers’, that is, these indicators may sometimes be strongly influenced by one or a few very highly cited publications. If for a particular journal IPP and SNIP are largely determined by a few very highly cited publications, this is reflected by wide stability intervals. In the interpretation of IPP and SNIP, it is therefore important to take into consideration not only the value of the indicator but also the width of the stability interval.
• Non-citing sources. As explained above, some journals have been classified as non-citing sources. This applies in particular to many journals in the arts and humanities. Citations originating from these journals are not counted in the calculation of IPP and SNIP, but these journals may have IPP and SNIP values themselves. These IPP and SNIP values should be interpreted with extreme caution. These values for instance do not include journal self citations, and therefore the values tend to be artificially low.