The result of Empirical analysis on factors associated with international scientific collaboration

In recent years the number of academic papers being published on research and development has
been increasing more slowly in Japan than in other leading nations. One of the reasons for this
may be the low ratio of internationally-collaborated articles in Japan, which tend to have more
citations than those written by domestic groups or individuals. However, up until now two major
themes have still not been investigated: factors related to international co-authorship and the
international mobility of researchers, and the reasons behind the higher citation rates of
internationally-collaborated articles. Therefore, this study examines those two themes by
empirical analysis using two types of datasets created through use of the Web of Knowledge
provided by Thomson Reuters. One of the results gained from analysis on papers published in the
past 20 years in Nature and Science using a count data regression model indicated several factors
that have a positive relationship with the production of academic papers: investment on R&D and
the number of researchers, the number of international students, and European Union membership.
The mobility of international researchers had a positive relationship with the number of
researchers at the transfer locations. Analysis using a database of papers in the field of chemistry
showed positive correlations between research performance (number of papers and their times
cited) and the degree of international collaboration. These tendencies appear to be common in
Japan, as well as in the United States, the United Kingdom, and China.