Traditional indicators inflate some countries’ scientific impact over 10 times

Abstract

At the center of this research lies the issue of properly counting international collaborations when assessing countries’ scientific productivity. Much of country-level scientometric research still uses the traditional “total counting” approach, wherein a country receives full credit for its international collaborations, as if it had produced every publication alone. For over a decade, various researchers have been showing how total counting distorts country outputs. However, the alternative, fractional counting methods, designed to eliminate the problem still have not prevailed. Hence more discussion and quantitative evidence is needed. In this article I study 40 average-productivity countries and find that total counting can result in even bigger distortion than the previous studies have shown. Namely, I show that total counting inflates some countries’ scientific impact as much as 12–13 times, rather than about 2 times, as observed with higher productivity countries. I also show that the degree of overcounting varies sharply across countries, often even resulting in a more productive country appearing behind a less productive one or vice versa. Based on the accumulated evidence, I suggest that total counting should be replaced with fractional counting more decisively, in most if not all of the research concerned with scientific productivity of countries.