Sunday, November 18, 2012

Article in University World News

online hub
    s View Printable VersionEmail Article To a Friend
GLOBAL
Ranking’s research impact indicator is skewed

Saturday, November 03, 2012


Apology

In a recent article in University World News I made a claim that Times Higher Education in their recent World University Rankings had introduced a methodological change that substantially affected the overall ranking scores. I acknowledge that this claim was without factual foundation. I withdraw the claim and apologise without reservation to Phil Baty and Times Higher education.

Saturday, October 27, 2012

More on MEPhI

Right after putting up the post on Moscow State Engineering Physics Institute and its "achievement" in getting the maximum score for research impact in the latest THE - TR World University Rankings, I found this exchange on Facebook.  See my comments at the end.

  • Valery Adzhiev So, the best university in the world in the "citation" (i.e. "research influence") category is Moscow State Engineering Physics Institute with maximum '100' score. This is remarkable achivement by any standards. At the same time it scored in "research" just 10.6 (out of 100) which is very, very low result. How on earth that can be?
  • Times Higher Education World University Rankings Hi Valery,

    Regarding MEPHI’s high citation impact, there are two causes: Firstly they have a couple of extremely highly cited papers out of a very low volume of papers.The two extremely highly cited papers are skewing what would ordinarily be a very g
    ood normalized citation impact to an even higher level.

    We also apply "regional modification" to the Normalized Citation Impact. This is an adjustment that we make to take into account the different citation cultures of each country (because of things like language and research policy). In the case of Russia, because the underlying citation impact of the country is low it means that Russian universities get a bit of a boost for the Normalized Citation Impact.

    MEPHI is right on the boundary for meeting the minimum requirement for the THE World University Rankings, and for this reason was excluded from the rankings in previous years. There is still a big concern with the number of papers being so low and I think we may see MEPHI’s citation impact change considerably over time as the effect of the above mentioned 2 papers go out of the system (although there will probably be new ones come in).

    Hope this helps to explain things.
    THE
  • Valery Adzhiev Thanks for your prompt reply. Unfortunately, the closer look at that case only adds rather awkward questions. "a couple of extremely highly cited papers are actually not "papers": they are biannual volumes titled "The Review of Particle Physics" that ...See More
  • Valery Adzhiev I continue. There are more than 200 authors (in fact, they are "editors") from more than 100 organisation from all over the world, who produce those volumes. Look: just one of them happened to be affiliated with MEPhI - and that rather modest fact (tha...See More
  • Valery Adzhiev Sorry, another addition: I'd just want to repeat that my point is not concerned only with MEPhI - Am talking about your methodology. Look at the "citation score" of some other universities. Royal Holloway, University of London having justt 27.7 in "res...See More
  • Alvin See Great observations, Valery.
  • Times Higher Education World University Rankings Hi Valery,

    Thanks again for your thorough analysis. The citation score is one of 13 indicators within what is a balanced and comprehensive system. Everything is put in place to ensure a balanced overall result, and we put our methodology up online for
    ...See More
  • Andrei Rostovtsev This is in fact rather philosofical point. There are also a number of very scandalous papers with definitively negative scientific impact, but making a lot of noise around. Those have also high contribution to the citation score, but negative impact t...See More

    It is true that two extremely highly cited publications combined with a low total number of publications skewed the results but what is equally or perhaps more important is that  these citations occur in the year or two years after publication when citations tend to be relatively infrequent compared to later years. The 2010 publication is a biennial review, like the 2008 publication, that will be cited copiously for two years after which it will no doubt be superseded by the 2012 edition.

    Also, we should note that in the ISI Web of Science, the 2008 publication is classified as "physics, multidisciplinary". Papers listed as multidisciplinary generally get relatively few citations so if the publication was compared to other multidisciplinary papers it would get an even larger weighting. 
    Valery has an excellent point when he points out that these publications have over 100 authors or contributors each (I am not sure whether they are actual researchers or administrators). Why then did not all the other contributors boost their instutitions' scores to similar heights? Partly because they were not in Russia and therefore did not get the regional weighting but also because they were publishing many more papers overall than MEPhI.  

    So basically, A. Romaniouk who contributed 1/173rd of one publication was considered as having more research impact than hundreds of researchers at Harvard, MIT, Caltech etc producing hundreds of papers cited hundreds of times.  Sorry, but is this a ranking of research quality or a lottery?

    The worse part of THE's reply is this:

    Thanks again for your thorough analysis. The citation score is one of 13 indicators within what is a balanced and comprehensive system. Everything is put in place to ensure a balanced overall result, and we put our methodology up online for all to see (and indeed scrutinise, which everyone is entitled to do).

    We welcome feedback, are constantly developing our system, and will definitely take your comments on board.

    The system is not balanced. Citations have a weighting of 30 %, much more than any other  indicator. Even the research reputation survey has a weighting of only 18%.  And to describe as comprehensive an indicator which allows a fraction of one or two publications to surpass massive amounts of original and influential research is really plumbing the depths of absurdity.

    I am just about to finish comparing the scores for research and research impact for the top 400 universities. There is a statistically significant correlation but it is quite modest. When research reputation, volume of publications and research income show such a modestcorrelation with research impact it is time to ask whether there is a serious problem with this indicator.

    Here is some advice for THE and TR.

    • First, and surely very obvious, if you are going to use field normalisation then calculate the score for discipline groups, natural sciences, social sciences and so on and aggregate the scores. So give MEPhI a 100 for physical or natural sciences if you think they deserve it but not for the arts and humanities.
    • Second, and also obvious, introduce fractional counting, that is dividing the number of citations by the number of authors of the cited paper.
    • Do not count citations to summaries, reviews or compilations of research.
    • Do not count citations of commercial material about computer programs. This would reduce the very high and implausible score for Gottingen which is derived from a single publication.
    • Do not assess research impact with only one indicator. See the Leiden ranking for the many ways of rating research.
    • Consider whether it is appropriate to have a regional weighting. This is after all an international ranking.
    • Reduce the weighting for this indicator.
    • Do not count self-citations. Better yet do  not count citations from researchers at the same university.
    • Strictly enforce your rule about  not including single subject institutions in the general rankings.
    • Increase the threshold number of publications for inclusion in the rankings from two hundred to four hundred.


Friday, October 26, 2012

Stellenbosch

Credit when due. Times Higher Education has now put Stellenbosch University in Africa.
Dancing in the Street in Moscow

"Jubilant crowds poured into the streets of Moscow when it was announced that Moscow State Engineering Physics Institute had been declared to be the joint top university in the world, along with Rice University in Texas, for research impact".

Just kidding about the celebrations.

But the Times Higher Education - Thomson Reuters World University Rankings have given the "Moscow State Engineering Physics Institute" a score of 100 for research impact, which is measured by the number of citations per paper normalised by field, year of publication and country.

There are a couple of odd things about this.

First, "Moscow State Engineering Physics Institute " was reorganised in 2009 and its official title is now  National Research Nuclear University MEPhI. It still seems to be normal to refer to MEPhI or Moscow State Engineering Physics Institute so I will not argue about this. But I wonder if there has been some confusion in TR's data collection.

Second, THE says that institutions are not ranked if they teach only a single narrow subject. Does the institution teach more than just physics?

So how did MEPhI do it ?  The answer seems to be because of a couple of massively cited review articles. The first was by C Amsler et (many many) alia in Physics Letters B of September 2008, entitled Review of Particle Physics. It was cited 1278 times in 2009 and 1627 times in 2010 according to the Web of Science, even more according to Google Scholar.

Here is the abstract.

"Abstract: This biennial Review summarizes much of particle physics. Using data from previous editions., plus 2778 new measurements from 645 papers, we list, evaluate, and average measured properties of gauge bosons, leptons, quarks, mesons, and baryons. We also summarize searches for hypothetical particles such as Higgs bosons, heavy neutrinos, and supersymmetric particles. All the particle properties and search limits are listed in Summary Tables. We also give numerous tables, figures, formulae, and reviews of topics such as the Standard Model, particle detectors., probability, and statistics. Among the 108 reviews are many that are new or heavily revised including those on CKM quark-mixing matrix, V-ud & V-us, V-cb & V-ub, top quark, muon anomalous magnetic moment, extra dimensions, particle detectors, cosmic background radiation, dark matter, cosmological parameters, and big bang cosmology".

I have not counted the number of authors but there are113 institutional affiliations of which MEPhI is 84th.

The second paper is by K. Nakamura et alia.  It is also entitled Review of Particle Physics and was published in the Journal Of Physics G-Nuclear and Particle Physics in July 2010 . It was cited 1240 times in 2011. This is the abstract.
 
"This biennial Review summarizes much of particle physics. Using data from previous editions, plus 2158 new measurements from 551 papers, we list, evaluate, and average measured properties of gauge bosons, leptons, quarks, mesons, and baryons. We also summarize searches for hypothetical particles such as Higgs bosons, heavy neutrinos, and supersymmetric particles. All the particle properties and search limits are listed in Summary Tables. We also give numerous tables, figures, formulae, and reviews of topics such as the Standard Model, particle detectors, probability, and statistics. Among the 108 reviews are many that are new or heavily revised including those on neutrino mass, mixing, and oscillations, QCD, top quark, CKM quark-mixing matrix, V-ud & V-us, V-cb & V-ub, fragmentation functions, particle detectors for accelerator and non-accelerator physics, magnetic monopoles, cosmological parameters, and big bang cosmology".

There are 119 affiliations of which MEPhI is 91st..

Let me stress that there is nothing improper here. It is normal for papers in the physical sciences to include summaries or reviews of research at the beginning of a literature review. I also assume that the similarity in the wording of the abstracts would be considered appropriate standardisation within the discipline rather than plagiarism.

TR 's method counts the numbers of citations of a paper compared to the average for that field in that year in that country. MEPhI would not get very much credit for a publication in physics which is a quite highly cited discipline, but it would get some for being in Russia where citations in English are relatively sparse and a massive boost for exceeding the average for citations within one or two years of publication many times over.

There is one other factor. MEPhI was only one of more than 100 institutions contributing to each of these papers but it got such an unusually massive score because its citations, which were magnified by region and period of publication, were divided by a comparatively small number of publications.

This is not as bad as Alexandria University being declared the fourth best university for research impact in 2010. MEPhI is a genuinely excellent institution which Alexandria, despite a solitary Nobel laureate and an historic library, was not.  But does it really deserve to be number one for research impact or even in the top 100? TR's methods are in need of very thorough revision.

And I haven't heard about any celebrations in Houston either.

Tuesday, October 09, 2012

What are the latest THE rankings telling us?


 This post has been taken down pending some fact checking.
.

Saturday, October 06, 2012

What Happened in Massachusetts?

The University of Massachusetts has fallen from 64th place in the Times Higher Education World University Rankings to 72nd.. Was this the result of savage spending cuts, the flight of international students or the rise of Asian universities?

Maybe, but there is perhaps another, less interesting reason.

Take a look at the THE reputation rankings for 2011 and 2012. The University of Massachusetts has fallen from 19th  position in 2011 to 39th in 2012. How did that happen when there is no other remotely comparable change in the reputation rankings?

The obvious answer is that THE, like QS before them, are sometimes a bit hazy about American state university systems. They most probably counted votes for all five campuses of the university system last year but only the flagship campus at Amherst  this year. The decline in the reputation score produced a smaller fall in the overall score. If there is another explanation I would be happy to hear it.
Update on My Job Application to Thomson Reuters

Looking at the Times Higher Education Reputation Rankings for 2012, I have just noticed that number 31 has gone missing, leaving 49 universities in the rankings.

Maybe not very important, but if you go around telling everybody how sophisticated and robust you are, a little bit more is expected.
My Job Application to Thomson Reuters

Looking at the 2011 Times Higher Education Reputation Ranking, prepared by Thomson Reuters, purveyors of robust and dynamic datasets, I noticed that there are only 48 universities listed (there are supposed to be 50) because numbers 25 and 34 have been omitted.



Friday, October 05, 2012

Observation on the THE Ranking

There will be some comments on the latest Times Higher Education World University Rankings over the next few days.

For the moment, I would just like to point to one noticeable feature of the rankings. The scores have improved across the board.

In 2011 the top university (Caltech) had an overall score of 94.8. This year it was 95.5.

In 2011 the 50th ranked university had a score of 64.9. This year it was 69.4.

In 2011 the 100th ranked university had a score of 53.7. This year it was 57.5 for the universities jointly ranked 99th..

In 2011 the 150th ranked university had a score of 46.7. This year it was 51.6.

In 2011 the 200th ranked university had a score of 41.4. This year it was 46.2.

The overall score is a combination of 13 different indicators all of which are benchmarked against the highest scorer in each category, which receives a score of 100. Even if universities throughout the world were spending more money, improving staff - student ratios, producing more articles, generating more citations and so on, this would not in itself raise everbody's, or nearly everybody's score.

There are no methodological changes this year  that might explain what happened.

Wednesday, October 03, 2012

Dumbing Down Watch

Is any comment needed?

Students starting university for the first time this autumn will be given a detailed breakdown of their academic achievements, exam results, extra-curricular activities and work placements, it was revealed.

More than half of universities in Britain will issue the new "Higher Education Achievement Report", with plans for others to adopt it in the future.

University leaders said the document would initially list students’ overarching degree classification.

But Prof Sir Robert Burgess, vice-chancellor of Leicester University and chairman of a working group set up to drive the reforms, said it was hoped that first, second and third-class degrees would eventually be phased out altogether.
University Migration Alert

At the time of posting, The new THE rankings listed Stellenbosch University in South Africa as one of the top European universities.

Remind me to apply for a job with Thomson Reuters sometime.

THE Rankings Out

The Times Higher Education World University Rankings 2012 are out. The top ten are :

1.  Caltech  (same as last year)
2.  Oxford  (up 2 places)
3.  Stanford (down 1)
4.  Harvard  (down 2)
5.  MIT (up 2)
6.  Princeton  (down 1)
7.  Cambridge (down 1)
8.  Imperial College London (same)
9.  Berekeley (up 1)

At the top, the most important change is that Oxford has moved up two places to replace Harvard in the number two spot.

Monday, October 01, 2012

Well,  they would, wouldn't they?

Times Higher Education has published the results of a survey by IDP, a student recruitment agency:

The international student recruitment agency IDP asked globally mobile students which of the university ranking systems they were aware of. The Times Higher Education World University Rankings attracted more responses than any other ranking - some 67 per cent.This was some way ahead of any others. Rankings produced by the careers information company Quacquarelli Symonds (QS) garnered 50 per cent of responses, and the Shanghai Academic Rankings of World Universities (ARWU) received 15.8 per cent.Asked which of the global rankings they had used when choosing which institution to study at, 49 per cent of students named the THE World University Rankings, compared to 37 per cent who named QS and 6.7 per cent who named the ARWU and the Webometrics ranking published by the Spanish Cybermetrics Lab, a research group of the Consejo Superior de Investigaciones Científicas (CSIC).


At the top of the page is a banner about IDP being proudly associated with the THE rankings. Also, IDP, which is in the student recruitment trade, is a direct competitor of QS.

The data could be interpreted differently. More respondents were aware of the THE rankings and had not used them than knew of the QS rankings and had not used them.

Saturday, September 29, 2012

Forgive me for being pedantic...

My respect for American conservatism took a deep plunge when I read this in an otherwise enjoyable review by Matthew Walther of Kingsley Amis's Lucky Jim:
Its eponymous hero, Jim Dixon, is a junior lecturer in history at an undistinguished Welsh college. Dixon’s pleasures are simple: he smokes a carefully allotted number of cigarettes each day and drinks a rather less measured amount of beer most nights at pubs. His single goal is to coast successfully through his two-year probation period and become a permanent faculty member in the history department.


It is well known, or ought to be, that the institution in the novel was based on University College, Leicester, which is a long way from Wales. The bit about the "Honours class over the road", a reference to the Welford Road municipal cemetery, is a dead giveaway.

Walther can be forgiven though since he reminded me of this description of Lucky Jim's history article;

“It was a perfect title, in that it crystallized the article’s niggling mindlessness, its funereal parade of yawn-enforcing facts, the pseudo-light it threw upon non-problems. Dixon had read, or begun to read, dozens like it, but his own seemed worse than most in its air of being convinced of its own usefulness and significance.”

:
Dumbing Down Watch

The New York Fire Department has announced the results of a new entrance exam. The passmark of 70 was reached by 95.72% of applicants. Previous tests had been thrown out because insufficient numbers of African-Americans and Hispanics were able to pass.

The new exam appears to be extremely easy and seems to assume that firefighting is a job that requires minimal intelligence. Effectively, the new policy for the New York Fire Department is to select at random from those able to get themselves to a testing center and answer questions that should pose no challenge to the average junior high school student.

The change was in response to the directives of Judge Nichols Garaufis, a graduate of Columbia Law School, which would seem to be as lacking in diversity as the NYFD. One suspects that the judge's disdain for the skills and knowledge, not to mention physical courage, of the firefighters is rooted in blatant class prejudice.

When is someone going to file a disparate impact suit against the top law schools?
Dumbing Down Watch

David Cameron, Old Etonian and Oxford graduate, apparently does not know the meaning of Magna Carta.

But, looking at the reports of the Letterman interview, he did not actually say he didn't know.

So, perhaps he was just pretending to be dumb.

Tuesday, September 25, 2012

Are Indian Universities Really Good at English?

Indian universities have never done well in any of the international university rankings. The problem is that although there are many talented Indian students and researchers they seem, like Indian entrepreneurs, to leave India as soon as they can.

But some observers have been taking comfort from a good showing in QS's subject rankings for English Language and Literature, which actually consists largely of the 2011 academic survey. One observer notes:

But DU's English department has made the day special as it figured in the top 100 list. Consequently, the celebratory mood was not restricted to North Campus alone but spilled over to South and off-campus colleges. Smiles could be observed on the faces of teachers as well as their pupils as their hard work paid off. "It is always pleasing to know you have done well at international level," said Ratnakar Kumar from Khalsa College. Another student from Hansraj added jokingly, "So we haven't been reading just stories because they don't fetch you such accolades." Verily, students were high on emotions.

A number of reasons can be attributed for the success that Department of English at Delhi University has witnessed. The significant factor is that students are encouraged and pushed to think outside the box, making bizarre to impressive attempts in their process of learning. So in a country in which rote-learning is the norm, DUDE has adopted a strategy to keep such evil at bay. Beside, the world-class faculty has also contributed to the improving standards of English Department.

The Times of India comments:

Teachers cite a number of reasons for the success of DU's English department. "First, it's the profile of the department in terms of research and publications. We are on top in subaltern studies, in post-colonial studies. Then, the numbers — we have 600 MA students, of whom 10-15 are as good as anybody," says a professor. He adds that India is not considered modern for technology but for the ideas of democracy and freedom, and those belong to the domain of humanities. The department is a part of the University Grants Commission's Special Assistance Programme.

The QS subject rankings for English are 90 per cent based on the academic survey so research and publications had nothing to do with it. My suspicion is that graduates of Delhi University have fanned out across the world for postgraduate studies and have signed up for the QS survey giving an American, British or whatever university as their affiliation. When they fill out the form, they probably put the Ivy League and Oxbridge down first and then Delhi university after about 20 or 30 other names. The QS methodology does not take account of the order of the responses so Harvard would get the same weighting as Delhi or much less if the respondent gave an American affiliation making Harvard a domestic vote and Delhi an international one.
Merging Universities

There is talk of merging Trinity College Dublin and University College Dublin. I am not sure whether there would be any benefit for faculty or students but inevitably there is a ranking angle.


The report says a UCD-TCD merger would give the merged college the critical mass and expertise needed to secure a place among the world’s best-ranked universities. At present, Ireland is not represented among the top 100 universities in the prestigious Times Higher Education World Reputation Ranking.

Monday, September 24, 2012

On the Right

I am reviving the papers section on the right hand side of this blog and hope to add to it from time to time. The number of citations in brackets refers to Google Scholar.
Are All Disciplines Equal?

Rankers, evaluators and assessors of all sorts have to face the problem that academic disciplines do not go about writing, publishing and citing in exactly the same way. Researchers in some disciplines write shorter papers that  have more authors, are more easily divided into smaller units and get more citations sooner after publication than others. Medical research papers tend to be short, frequent,  co-authored and cited very soon after publication compared to history or philosophy.

Different ranking organisations follow different approaches when dealing with this diversity of practice. Scimago just counts the total number of publications. ARWU of Shanghai counts publications but gives a double weighting to social sciences and none to the humanities. Thomson Reuters, who power the Times Higher Education world rankings, normalize by field and by country.

Here are some fairly simple things that rankers might try to overcome disparities between disciplines. They could  count total pages rather than the number of papers. They could look into counting citations of conference papers or books. Something worth doing might be giving a  reduced weighting to co-authored papers, which would shift the balance a little bit towards the arts and humanities and might also help to discourage dubious practices like supervisors adding their names to papers written by their graduate students.

We should also ask whether there are limits to how far field and country normalization should go. Do we really believe that someone who has received an average number of citations for political science in Belarus deserves the same score as someone with an average number of citations for chemical engineering in Germany?

It does seem that there are substantial and significant variations in the cognitive skills required to compete effectively in the academic arena. Here are combined verbal and quantitative GRE scores for selected disciplines by intended majors for 2011-2012.

Physics and Astronomy                          317
Economics                                              313
Biology                                                   308
Philosophy                                              310
English Language and Literature           305
Education: Secondary                             305
History                                                    304
Education: Higher                                  304
Psychology                                             300
Sociology                                               300
Education: Administration                     299
Education: Curriculum                           299
Education: Early Childhood                   294


The scores look as they are very close together but this is largely a a consequence, perhaps intended, of the revision (dumbing down?) of the GRE.

Is it possible that one reason why physicists and economists publish more papers which are read more than those by education specialists is simply that the former have more ability and interest than the latter?

Friday, September 21, 2012

Dumbing Down Watch

There is a "cheating" scandal at Harvard. Apparently, students were given an open-book, open-anything take-home exam for 'Introduction to Congress' and were not expected to consult each other in any way.
 
Harvard College’s disciplinary board is investigating nearly half of the 279 students who enrolled in Government 1310: “Introduction to Congress” last spring for allegedly plagiarizing answers or inappropriately collaborating on the class’ final take-home exam.
Dean of Undergraduate Education Jay M. Harris said the magnitude of the case was “unprecedented in anyone’s living memory.”
Harris declined to name the course, but several students familiar with the investigation confirmed that Professor Matthew B. Platt's spring government lecture course was the class in question.
The professor of the course brought the case to the Administrative Board in May after noticing similarities in 10 to 20 exams, Harris said. During the summer, the Ad Board conducted a review of all final exams submitted for the course and found about 125 of them to be suspicious.

Presumably this is not the only take-home exam in Harvard and presumably not the first for this course. So why now has half the class felt compelled to plagiarise or to collaborate inappropriately?

Has the course become more difficult than it used to be? Or are the students less capable? Or have admission standards become less rigorous?

Maybe QS and Times Higher were on to something after all when they dethroned Harvard from the number one spot in their world rankings.


Wednesday, September 19, 2012

New Canadian Research Rankings

Higher Education Strategy Associates recently published their Canadian Research Rankings, which are based on the award of grants and H-indexes. I am not sure about counting grants since it is likely that the skills needed to lobby for grants and those needed to actually do research are not always the same.

The rankings do not include medical research.

The top five for science and engineering are:

1.  University of British Columbia
2.  Montreal
3.  Toronto -- St. George
4.  Ottawa
5.  McGill

The top five for  social sciences and humanities are:

1.  University of British Columbia
2.  Mcgill
3.  Toronto -- St George
4.  Alberta
5.  Guelph


These rankings, like the Times Higher Education (THE) World University Rankings are based on field normalisation. In other words they do not simple count the number of grants and H-index scores but compare them with the average for the field. The rationale for this is that there are enormous differences between disciplines so that it would, for example, be unfair to compare a physicist who has won a grant of 10,000 dollars, which is below average for physics, with an education researcher who has won a similar award, which is above average for education. Equally, does it make sense to rank a physicist with an average H-index for physics well above a linguist with an average one for linguistics?

Here are the average grants for various fields:

biological engineering 84,327
physics  42,913
linguistics 13,147
history 6,417
education 5,733

Here are the average H-indexes for discipline clusters:

science  10.6
social sciences 5.2
humanities 2.3


HESA (and THE ) do have a point. But there are problems. One is that as we drill down to smaller units of analysis there is a greater risk of outliers. So a single large grant or a single much cited author in a field with few grants or citations could have a disproportionate impact.

The other is that field normalisation implies that all disciplines are equal. But is that in fact the case?