What does this index measure?
The Nature index’s source is based only on the number of articles published in Nature titles. The index does not contain citations, and only covers the publisher’s own journals. It is therefore neither a measure of impact, nor of productivity, but of a university’s ability to produce articles in one highly reputable source. It therefore only covers chemistry, life science, earth and environmental sciences and physics. Each of these individually are presented on a list of the 50 highest publishing universities, and one list of 175 based on all of them. The young universities categorisation is 50 years – longer than for Times Higher’s comparable ranking (25 years).
|University||Pos.||Foundation Year||FC||AC||Highest position||Global Pos. (all universities)|
The data show that UNESP publishes slightly more in Nature journals than UFABC, and also tends to feature slightly more prominently in research teams than its federal counterpart. In part, this is reflective of the fact that in its short history UFABC has been very successful in participating in international research groups, but does not have as much leadership of these networks as UNESP. Despite this, UFABC is higher ranked in physics. This is the only category that they both appear in. It should be noted that this is UFABC’s biggest research strength, whereas UNESP has a number of other high impact departments in life sciences. These help Unesp to a higher overall position in the ranking.
There is evidence of the bias introduced by fractional counting (a percentage of involvement in each article) in comparing areas of knowledge in this way. We can see that each of UNESP’s papers in life sciences earned an average of 0.21 points (indicating around total author 21% involvement in each paper), compared to 0.075 per paper for physics (indicating 7.5% total author involvement). In other words, UNESP’s papers in life sciences were almost three times as effective on this methodology, because life sciences tend to have fewer authors.
|UNESP||Earth & Environmental Sciences||6||1.03|
|UFABC||Earth & Environmental Sciences||1||0|
UNESP publishes 86.21% of its articles in Nature in international coauthorship, and UFABC 89.46%, (compared to around 40% in broader bases such as Web of Science). This shows the importance of internationalization in top level research.
There are two indicators in this index. The first is full article count (AC), the number of articles published in Nature titles. The other is the fractional count (FC), which gives a relative score of the participation in an article. On this count, all authors are considered to have contributed equally, and each paper is given a contribution of 1.0. If there are two authors from two different institutions, each will be given 0.5, if there are five, each is given 0.2. Although in practice this is rarely true, attribution conventions vary from discipline to discipline so it is not possible to create a single indicator equally relevant to different areas of knowledge on this basis. That means that there is a natural bias against areas that tend to have large numbers of authors (physics) and towards those that tend towards smaller ones (particularly biological sciences). Furthermore, the fractional count appears to run counterintuitively to many of the universities’ priorities; to publish in coauthorship with government and industry, for example.
Benchmark for UFABC
For UFABC, universities founded up to 10 years prior (1995) and with a specific focus on physics have been selected.
|University||Position||Foundation Year||FC||AC||Highest position||Global Position|
|Ulsan National Institute of Science and Technology (UNIST)||10||2009||68.88||161||P8||189|
|Hong Kong Polytechnic||24||1994||34.63||116||P16||329|
|Skolkovo Institute of Science and Technology||97||2011||9.85||64||P66||682|
There are not many institutions as young as UFABC and specialized in physics as highly positioned in this ranking. Ranking above Skolkovo is particularly impressive, given that Russia has invested millions of dollars into the project, which it is creating in conjunction with MIT, and placing above the IITs (of which Indore here is illustrative) shows that UFABC has, in a very short space of time, established outstanding research capabilities in physics. The two Hong Kong universities are a decade further ahead in development, and while HKUST is much stronger, UFABC should seek to study further both UNIST and Hong Kong Polytechnic as universities benchmarks.
This benchmark aims to demonstrate how UNESP could enter into the Life Sciences Ranking. Most of the universities do not publish significantly more than UNESP in Nature titles, but have significantly higher numbers of authorship within each paper. In order to feature in this ranking, UNESP should seek to encourage its researchers to include more colleagues in their publications, as currently other universities are managing to score significantly higher per paper in the fractional count.
|Ulsan National Institute of Science and Technology (UNIST)||36||2009||5.41||19|
|Pohang University of Science and Technology (POSTECH)||24||1986||7.31||25|
|Miguel Hernández University of Elche (UMH)||46||1996||4.54||15|
|Southern University of Science and Technology (SUSTech)||47||2012||4.52||34|
|University of Wollongong (UOW)||49||1975||4.44||14|
How the universities can improve in this ranking
UNESP’s biggest listed research collaborator is FAPESP, which takes 2.58 of their score in the fractional count. Given that the support FAPESP offers is not usually with provision of human capital itself, if this were removed as a collaborator, UNESP would score 18.95 and move up to 53rd position. MCTIC takes a further point – taking UNESP to 20 points and a subsequent position of 50th. Ensuring correct attribution of articles would help UNESP to rise in this index.
Because UNESP’s papers in life sciences are nearly 3 times higher scoring than their physics papers, encouraging these departments to publish in Nature journals would have a proportionally larger impact on their performance in this ranking.
Because the fractional count is based on the number of authorships (i.e. if UNESP has 2 from 5 authors, they would score (0.2 + 0.2 = 0.4). This means that encouraging researchers to include their departmental colleagues in Nature articles they are publishing is beneficial to performance in this ranking. In order to gain entry to the list of the top 50 worldwide for life sciences, UNESP publishes almost enough articles to be included, the universities with less than 30 papers all have around 0.3 per paper (30% of authorships), whereas for UNESP it is 0.21 (21%).
UFABC is extremely successful on this methodology given how young the university is, and how much the methodology is biased against physics intensive institutions. The majority of the top 20 is occupied by institutions intensive in chemistry, which typically has the lowest number of authors of the areas included in this ranking.