Page 209 - Becoming Metric Wise

199
Journal Citation Analysis

In this chapter we focused on the classical citation- based journal indi-
cators. Yet, all methods known as altmetrics can, of course, also be
applied to journals. This remark holds true also for download data which
are typically collected for journals, see e.g. (Bollen et al., 2005). We
should also mention that multidisciplinary journals such as Nature, Science
or PLoS One do not lead themselves to the application of most journal
indicators, as their content crosses the borders between many fields. Such
journals should be considered as sets of articles, each of which to be con-
sidered on their own merit.
Journals with a high impact factor attract a lot of attention, and hence,
one may expect that a Matthew effect comes at play, increasing the impact
factor. Yet, it is very difficult to prove this as many factors determine the
citedness of an article, besides the journal in which it was published. In
order to find out, one should theoretically publish the same article at the
same time in different journals to determine the influence of the impact
factor on citedness. This is, of course, normally not possible, but Perneger
(2010) and Shanahan (2016) found a case where the same text was pub-
lished in different journals, namely the case of biomedical reporting
guidelines. As they are not scientific publications, but rather recommen-
dations for authors, reporting guidelines are often published simulta-
neously to encourage wider adoption and dissemination. Perneger and
Shanahan found that the number of citations received by the same publi-
cation was moderately strongly correlated to the JIF. Shanahan (2016) also
found a strong positive correlation between the number of citations and
the number of article accesses. Bornmann and Leydesdorff (2017) investi-
gated and confirmed the skewness of the journal citation distribution.
They found though that among all factors that may have an effect on
resulting article citedness (number of cited references, number of authors,
number of pages, JIF) the JIF has the strongest correlation with citation
scores. A note of caution should be added here as this is not necessarily a
causal effect. It might simply be that authors submit their best articles to
high impact journals.