Page 198 - Becoming Metric Wise

188 Becoming Metric-Wise

Vanclay (2012) published an article which was very critical for
Thomson Reuters and the way the standard impact factor is calculated.
Colleagues, including some working for Thomson Reuters (Pendlebury &
Adams, 2012) got the opportunity to react (positively or negatively) to his
memorandum.
We think that citations in editorials should not be counted when cal-
culating the JIF. As editorials are very rarely highly cited (Rousseau,
2009c) this illustrates the fact that such references hardly ever support
high level arguments.
A simple solution to some of the problems related to the calculation
of the JIF is using a median instead of an average. This approach has
recently been applied by the Nature Publishing Group (NPG) as
announced by EIC Philip Campbell and Sowmya Swaminathan, head of
editorial policy, in an anonymous editorial (Anonymous, 2016). The
median impact of a journal for the year Y is the median number of cita-
tions that articles published in the years Y 2 2 and Y 2 1 received during
the year Y. NPG used only genuine articles or reviews for the calculation
of the median impact.

6.16 CITESCORE INDEX
On December 8, 2016 Elsevier launched the CiteScore index, a journal
indicator serving the same purpose as the JIF, but calculated in a slightly
different way.
The first difference is, of course, the used database. The CiteScore
index makes use of the journals covered by Scopus. The second is the dif-
ferent way in which a journal’s impact is calculated. The CiteScore index
is a 3-year synchronous indicator (while JIF is a 2-year one). Finally, and
this makes a huge difference for some journals, while not affecting others,
the CiteScore index divides by all published items, not just the so-called
“citable” ones. For journals such as the Lancet, Science or Nature this makes
quite a difference.

6.17 WHO MAKES USE OF BIBLIOMETRIC INDICATORS?
(STOCK, 2009)
• Journal editors and publishers.
They use impact factors as marketing tools. Journals’ impact factors
and especially their increase are prominently displayed on journals’
webpages. Note that, as databases tend to increase and reference lists