Page 6 - IDEA Study 2016 01 Publication Output

we have also monitored the publication output in highly cited journals, with a generally
strict review process. For each discipline, we present detailed data describing the dynamics
of its output on an international level, as well as the quality of its scientific journals helping
to increase its output.

Methodology

For our analysis, we have only used works published within the years 2010–2014, included
in the WoS database as Articles. An article is credited to a certain country if at least one of
the authors listed in the WoS is affiliated with an institution located in that country. This
means that an article with multiple authors affiliated to different countries has been
credited to each country used in our comparison which is listed at least once. If a journal
publishing a certain article is listed under more than one WoS category, the article has
been credited to each of the disciplines.

As a publication quality benchmark, we have used the Article Influence Score citation
index (AIS). The AIS is derived from the article citation rate, giving greater weight to
citations in prestigious journals. The AIS of a certain journal therefore correlates
significantly with the journal’s scientific repute, which also results in a more stringent
review process. Citation indices such as the AIS are not particularly useful in measuring
the importance and quality of a single publication or a select few publications (e.g. when
evaluating individual scientists), since the citation rate of a specific article (i.e. its
importance) may significantly differ from the average citation rate of all articles published
in the given journal. However, when assessing a large number of articles, as is the case
here, the quality estimate bias is much reduced.12

We have included each article found in one of the four quartiles of the total number of
journals listed for each field and year based on their AIS value.13 Furthermore, we have
separately analyzed the number of articles in the highest AIS value decile.

Our comparison includes 11 small and medium-sized countries, only one of which has
English as its official language: 5 European countries from the former “Soviet bloc”,
4 Western European countries, and 2 non-European countries. For each country in the
comparison, we have normalized the total numbers of articles in each given discipline to
the population of the respective country, i.e. multiplied them by the ratio of POPcr/POPi,

12 The peer review process is another way of assessing quality of research. Compared to bibliometric
indicators, it has a number of advantages when evaluating research results, but may still fall short of correctly
assessing their true quality and importance.
13 Ranking is sorted by journal, not by articles included therein.

4