Page 5 - IDEA Study 2016 01 Publication Output

Basis of Comparison

Most international bibliometric comparisons are based on bibliometric indicators that are
easy to calculate but often they may be misleading. A generally unsuitable, yet still widely
used publication output indicator is the simple total number of articles published in all
scientific fields, even though they may vary greatly as to their publication practice and
frequency (i.e. the usual number of articles produced by an author per year). Such
numbers are then affected by the field structure of research, which varies from country to
country. Therefore, we have decided to compare publication output strictly within
individual, quite narrowly defined fields9. This has significantly reduced (albeit not
completely eliminated) the impact of differences in publication practice.10

Using the simple number of published articles, albeit within a single scientific discipline,
does not reflect the potentially vast differences in the articles’ quality, importance and
contribution to scientific knowledge.11 We have therefore attempted to offer comparisons
based on the ability to publish first-rate research. The quality of publications is often
measured by their average number of citations. A typical tool used for this purpose is the
relative citation index (RCI). The RCI compares the average citation rate of articles
published in the given country during a given year, in scientific journals for the given
discipline, with the average citation rates of all articles published in that year and
discipline worldwide. Such an indicator, however, cannot explain whether its low value
stems from the lack of frequently cited works, or from an unusually high number of rarely
cited articles in a given country (e.g. if the country’s national journals have been included
in the WoS database). The RCI value may also be affected by a mere few WoS articles that
are not typical for the discipline, or by thousands of articles included in the WoS: this
makes it a significant factor in R&D management. Finally, we have analyzed recently
published articles (up until the year 2014), which also limits the possibility of tracking
their citation impact.

Our comparison has therefore focused on a number of first-rate publications that is easy to
interpret. This means that apart from the overall publication output within a given field,

9 We classify disciplines using the WoS Categories classification, distinguishing 173 fields in the Sciences
group, and 55 fields in the Social Sciences group.
10 In many science fields, for example, theoretical researchers have a significantly lower publishing frequency
than empirical researchers in the same field do, since empirical research focuses primarily on measurement
results.
11 A number of studies have proven that research published in a relatively small number of the most
prestigious journals has a significantly higher typical impact than research results published in other
journals. See, for example, Garfield (1996), Ioannidis (2006), Meho (2007).

3