New Leiden Ranking

Loet Leydesdorff loet at LEYDESDORFF.NET
Mon Dec 19 08:07:47 EST 2011 

Dear Gangan, 

One can also take the first moment: p*q. This is by definition equal to the number of papers in the top-10% most-highly cited papers. Following Rousseau's recent article, one can also consider two percentile rank classes. The top-10% is then counted with value = 1 and the others with value = 0.

I am more inclined to appreciate the tales of the distribution, and thus to value these two ranks with 2 and 1, respectively. I added these two options in red to your spreadsheet (attached). Alternatively, one could use the six percentile ranks of the NSF (Bornmann & Mutz, 2011). According to the authors in Leiden, the rankings in the various (top) ranks (5%, 10%, 20%) would be rather similar <Figure 3 at http://www.leidenranking.com/leidenranking.pdf >.

Perhaps, it is also useful to distinguish between an excellence indicator (such as the top-10%) and an impact indicator (such as I3).

Best wishes,
Loet

-----Original Message-----
From: gprathap at cmmacs.ernet.in [mailto:gprathap at cmmacs.ernet.in] 
Sent: Monday, December 19, 2011 12:51 PM
To: Loet Leydesdorff
Cc: sigmetrics at listserv.utk.edu
Subject: Re: [SIGMETRICS] New Leiden Ranking

Dear Loet,

I'm not sure the Leiden approach to rank entitites on a quality proxy alone does justice to the big fellas like Harvard. A better method may be to use what I call the second order indicators where quantity proxy (read P, the number of papers) is multiplied by the square of the quality proxy (say, PP for a very simplistic measure of quality).

Recently it was brought to my notice that Vinkler (Scientometrics,1988,
14,161-163) had advocated precisely this, arguing that by doing so, one "rewards" high quality and "punishes" low quality by such an approach.

If the Leiden rankings are re-worked in this manne, we get the results shown in the attached spreadsheet. I have compared the 4 Indian institutions in the Leiden list with the "top 100". Harvard is 64 times more "effective" than the 4 Indian institutions in the Leiden list.

Regards,

Gangan Prathap



> Adminstrative info for SIGMETRICS (for example unsubscribe):
> http://web.utk.edu/~gwhitney/sigmetrics.html
>
> Dear Chudamani,
>
> It seems to me that an indicator or ranking which allows for 
> statitistical testing of the significance of differences and positions 
> is far superior to one which does not (such as the h-index). 
> Fortunately, these two major teams in our field (Granada and Leiden 
> University) have agreed on using such an indicator for the Scopus and 
> WoS databases, respectively.
>
> Of course, there remains the problem of 
> interdisciplinarity/multidisciplinarity in institutional units such as 
> universities. One might say that a university can improve its average 
> citation score by closing the mathematics department. :-) This can be 
> counteracted a bit by field-normalization and perhaps by fractionation 
> of the citations (1/NRef) in the citigng papers.
>
> Best wishes,
> Loet
>
> On Mon, Dec 12, 2011 at 11:30 AM, K S Chudamani
> <ksc at library.iisc.ernet.in>wrote:
>
>> Adminstrative info for SIGMETRICS (for example unsubscribe):
>> http://web.utk.edu/~gwhitney/sigmetrics.html
>>
>> >
>>
>>
>>
>> dear all,
>>
>> I would like to point out that the elaborate method used is simply an 
>> extension of regular method. We have used h index and ranked 
>> universitires in India.it has been published in pearl journal. This 
>> also gives similar results. Then why break our head by complicatihg 
>> matters
>>
>> chudamani
>>
>>
>> Adminstrative info for SIGMETRICS (for example unsubscribe):
>> > http://web.utk.edu/~gwhitney/sigmetrics.html
>> >
>> > Dear colleagues,
>> >
>> >
>> >
>> > Using the newly introduced indicator for impact in the Leiden 
>> > Rankings
>> > 2011/2012 <http://www.leidenranking.com/ranking.aspx> —the
>> Proportion
>> > top-10% publications (PPtop 10%)—one can test differences between 
>> > institutions statistically using the z-test. Furthermore, one can 
>> > test whether each university performs above or below expectation.
>> >
>> >
>> >
>> > An Excel sheet with the test embedded is made available at 
>> > http://www.leydesdorff.net/leiden11/leiden11.xls and an example is 
>> > elaborated in a short introduction at 
>> > http://www.leydesdorff.net/leiden11/index.htm (coauthored with Lutz 
>> > Bornmann).
>> >
>> >
>> >
>> > The test was previously used analogously for the Excellence 
>> > Indicator
>> in
>> > the SCImago Institutions Rankings
>> > <http://www.scimagoir.com/pdf/sir_2011_world_report.pdf> ; cf.
>> > http://www.leydesdorff.net/scimago11; Bornmann 
>> > <http://arxiv.org/ftp/arxiv/papers/1110/1110.2305.pdf> et al., in
>> press),
>> > and can be considered as additional to the stability intervals
>> provided
>> at
>> > the webpages of the Leiden Ranking
>> > <http://www.leidenranking.com/ranking.aspx> . The SCImago Rankings 
>> > are based on Scopus data, and the Leiden Ranking on Web-of-Science data.
>> >
>> >
>> >
>> > With best wishes,
>> >
>> > Loet
>> >
>> >
>> >
>> >   _____
>> >
>> > Amsterdam School of Communications Research (ASCoR), 
>> > Kloveniersburgwal 48, 1012 CX Amsterdam.
>> > Tel.: +31-20- 525 6598; fax: +31-842239111  
>> > <mailto:loet at leydesdorff.net> loet at leydesdorff.net ; 
>> > <http://www.leydesdorff.net/> http://www.leydesdorff.net/ ; 
>> > <http://scholar.google.com/citations?user=ych9gNYAAAAJ&hl=en>
>> > http://scholar.google.com/citations?user=ych9gNYAAAAJ&hl=en
>> >
>> >
>> >
>> > From: ASIS&T Special Interest Group on Metrics 
>> > [mailto:SIGMETRICS at LISTSERV.UTK.EDU] On Behalf Of Paul Wouters
>> > Sent: Friday, December 09, 2011 11:58 AM
>> > To: SIGMETRICS at LISTSERV.UTK.EDU
>> > Subject: [SIGMETRICS] New Leiden Ranking
>> >
>> >
>> >
>> > Adminstrative info for SIGMETRICS (for example unsubscribe):
>> > http://web.utk.edu/~gwhitney/sigmetrics.html Dear colleagues
>> >
>> > US still dominates high impact publications in science
>> >
>> >
>> >
>> > The US are still the dominant scientific world power, but new 
>> > centres
>> of
>> > science are emerging. MIT is the university which has the highest
>> citation
>> > impact of its publications in the world. Princeton and Harvard take 
>> > positions two and three. These are some of the findings of the new
>> Leiden
>> > Ranking 2011 – 2012 which has been published on the website:
>> > www.leidenranking.com <http://www.leidenranking.com/> . The top 
>> > fifty
>> list
>> > consists of 42 US based universities, 2 Swiss (Lausanne at 12 and 
>> > ETH Zurich at 18), 1 Israeli (Weizmann Institute of Science), 4 
>> > British (Cambridge at 31, London School of Hygiene & Tropical 
>> > Medicine at 33, Oxford at 36 and Durham at 42), and one Danish 
>> > university (Technical University of Denmark). Aggregated to country 
>> > level, the US has 64 universities in the top 100 list, the UK 12, and the Netherlands 7.
>> The
>> > latter is remarkable given its small size.
>> >
>> >
>> >
>> > The Leiden Ranking 2011-2012 is based on an advanced methodology 
>> > which compensates for distorting effects due to the size of the 
>> > university,
>> the
>> > differences in citation characteristics between scientific fields, 
>> > differences between English and non-English publications, and
>> distorting
>> > effects of extremely high cited publications. Publications authored 
>> > by researchers at different universities are attributed to the
>> universities
>> > as fractions. This prevents distortion of the ranking by counting
>> these
>> > publications multiple times (for each co-authoring university). 
>> > This distorting effect is often overseen in other global university
>> rankings,
>> > which leads to a relative advantage of clinical research and some
>> physics
>> > fields in these rankings. This makes clear how sensitive global
>> rankings
>> > are to the nitty-gritty of the calculations.
>> >
>> >
>> >
>> > The Leiden Ranking enables users to choose the criteria on which 
>> > they
>> wish
>> > to compare university performance. The menu offers 3 indicators of
>> impact
>> > and 4 indicators of scientific collaboration. When scored on the 
>> > percentage of their papers produced in collaboration with 
>> > institutes
>> in
>> > different countries, the London School of Hygiene & Tropical 
>> > Medicine
>> tops
>> > the list with more than 50 % of its publications co-authored with
>> other
>> > countries.
>> >
>> >
>> >
>> > Although in terms of impact, US universities are still strongest, 
>> > it
>> is
>> > clear that other countries are emerging as centres of science by
>> looking
>> > at the total production (number of publications in the Web of
>> Science).
>> In
>> > this ranking, Harvard University is number one. But in the top 25 
>> > we
>> also
>> > see universities from Canada (Toronto at 2, British Columbia at 
>> > 22),
>> Japan
>> > (Tokyo at 4, Kyoto at 11, Osaka at 25), Brazil (Sao Paulo at 8),
>> United
>> > Kingdom (Cambridge at 13, Oxford at 14, University College at 17),
>> South
>> > Korea (Seoul at 19), and China (Zhejiang at 20).
>> >
>> >
>> >
>> > The Leiden Ranking is the first global university ranking which has 
>> > published the details of its methodology and indicators. The
>> indicators
>> > are presented in combination with stability intervals, an advanced 
>> > statistical method to measure to what extent the differences in
>> rankings
>> > between universities are significant.
>> >
>> >
>> >
>> > If one wishes to compare the university citation impact in a global 
>> > context, it is best to take the percentage of papers in the top 10 
>> > % highly cited papers together with the calculation method 
>> > “fractional counting†. This is the method which compares across 
>> > institutions and fields in the fairest way.
>> >
>> >
>> >
>> > The Leiden Ranking is based on data of the Web of Science. Data on 
>> > the arts and humanities are not included since these fields are not 
>> > well represented in the Web of Science. The Leiden Ranking 
>> > exclusively
>> measures
>> > the citation impact of research of the 500 largest universities in 
>> > the world. This prevents an arbritrary combination of performance 
>> > in education, valorization and research, a disadvantage of many 
>> > global university rankings.
>> >
>> >
>> >
>> > More information about the ranking results and its methodology:
>> > www.leidenranking.com <http://www.leidenranking.com/> .
>> >
>> >
>> >
>> > With regards
>> >
>> > Paul Wouters
>> > Professor of Scientometrics
>> > Director Centre for Science and Technology Studies Leiden 
>> > University
>> >
>> > Visiting address:
>> > Willem Einthoven Building
>> > Wassenaarseweg 62A
>> > 2333 AL Leiden
>> > Mail address: P.O. Box 905
>> > 2300 AX Leiden
>> > T: +31 71 5273909 (secr.)
>> > F: +31 71 5273911
>> > E: p.f.wouters at cwts.leidenuniv.nl
>> >
>> > CWTS home page: www.cwts.nl
>> > Blog about Citation Cultures: http://citationculture.wordpress.com/
>> > <http://citationculture.wordpress.com/>
>> > Research Dreams: www.researchdreams.nl
>> >
>> >
>> > --
>> > This message has been scanned for viruses and dangerous content by 
>> > MailScanner, and is believed to be clean.
>> >
>> >
>>
>>
>>
>>
>> --
>> This message has been scanned for viruses and dangerous content by 
>> MailScanner, and is believed to be clean.
>>
>>
>
>
> --
> Professor Loet Leydesdorff
> Amsterdam School of Communications Research (ASCoR) Kloveniersburgwal 
> 48, 1012 CX Amsterdam
> Tel.: +31-20- 525 6598; fax: +31-20- 525 3681 loet at leydesdorff.net ; 
> http://www.leydesdorff.net/
>
>

-------------- next part --------------
A non-text attachment was scrubbed...
Name: Leiden Rankings.GPrathap.19Dec11.xls
Type: application/vnd.ms-excel
Size: 138752 bytes
Desc: Leiden Rankings.GPrathap.19Dec11.xls
URL: <http://mail.asis.org/pipermail/sigmetrics/attachments/20111219/c6965c1a/attachment.xls>

More information about the SIGMETRICS mailing list