10 institutions contribute to more than 10% index output in 2016

More than 9,500 institutions published at least one paper included in 2016 index.

18 July 2017

DATA TREND

eye35.pix / Alamy Stock Photo

The French National Centre for Scientific Research produced close to 5,000 articles in journals included in the index in 2016.

In 2016, close to 60,000 articles were published in the 68 high-quality journals included in the Nature Index. These articles were produced by more than 9,500 institutions in 163 countries. But how evenly were they distributed across institutions?

As shown in the two graphs below, three-quarters of institutions in the index produced 11 or fewer articles, measured as article count (AC), in 2016. On the other end of the scale, less than 3% of the institutions represented in the index produced 250 articles or more.

The 250 institutions that produced more than 250 articles accounted for close to 60% of the total fractional count (FC) in the index last year. While the more than 6,000 institutions that published 11 articles or fewer in the index account for less than 5% of the total FC. Indeed, just 10 institutions (listed below) contributed 10% of the index's total FC last year.

The index tracks affiliation data by AC and FC. An institution gets an AC of 1 for every article that includes at least one author from that institution, but its FC represents its share of authorship — the FC for all the authors on a paper sums up to 1. The graph below shows a close relationship between the two metrics: a higher AC is associated with a higher FC. Each point on the graph represents a single institution, with a cutoff for institutions that produced more than 2,000 high-quality papers in 2016.

But some institutions have a greater share of the authorship on their papers. Dividing an institution's FC by its AC can reveal the average share of authorship it contributes to papers. Multiple institutions contribute to articles published in the index. More than 85% of institutions contribute, on average, less than one third to paper authorship.

Data analysis by Alex Scherrmann and Aaron Ballagh