Page 194 - Becoming Metric Wise

184 Becoming Metric-Wise

the number of citations received by each journal by all documents published
in the two previous years. Note though that these numbers are not the same
as those one can retrieve oneself in the WoS. Thomson Reuters (nowadays,
Clarivate Analytics) claimed that the numbers shown in the JCR are more
accurate as they are derived from a cleaned version of the WoS. This means
that, most of the time, the official JIF is higher than one an outsider would
calculate based on the WoS (Wu et al., 2008). The JCR also provides the
number of review articles and the number of other citable documents for
each journal. This is interesting information as reviews generally receive
more citations than other publications and hence review journals (journals
that publish—almost—exclusively reviews) have generally higher impact fac-
tors than journals which publish no or much less reviews.
Yet, also here there is a problem in the sense that Clarivate Analytics
uses its own definition of a review. It may happen that the authors state
in the title or abstract that they wrote a review but that this article is not
classified as a review in the WoS or in the JCR (Colebunders &
Rousseau, 2013).
The JCR provides even more detailed information shown under the
cited journal data or the citing journal data for journal J. The cited jour-
nal table provides detailed information about journals, and other sources,
citing journal J; their names and the publication year(s) they have cited.
This information covers a 10 year period. Cited older volumes are
brought together under rest. There is a lower bound for the number of
citations (usually 2). All sources citing journal J just once (if the lower
bound is 2) are brought together as all others. As expected (see Chapter 9:
The Informetric Laws) all others is always very high in the list, as infor-
metric data such as these are very skew with a long tail. Yearly totals are
shown on top of the detailed table. The citing journal table is constructed
in a similar way. We note that a journal usually cites itself a considerable
amount of time. Hence a journal is usually ranked high, i.e., having a
small rank, in these tables and this in the cited as well as the citing data
list.
Recall that data taken from the cited journal table (top line) are
needed in case one wants to calculate a diachronous impact factor.

6.14 STRUCTURE OF THE SCIMAGO DATABASE

The SCImago database for journals is freely available at http://www.
scimagojr.com/journalrank.php. SCImago uses Scopus data for these