Saturday, May 30, 2009

Asian University Rankings

QS Quacquarelli Symonds has come out with a ranking of the top 200 Asian universities. Here is the top ten.

1. University of Hong Kong
2. Chinese University of Hong Kong
3. University of Tokyo
4. Hong Kong University of Science and Technology
5. Kyoto University
6. Osaka University
7. Korean Advanced Institute of Science and Technology
8. Seoul National University
9. Tokyo Institute of Technology
10. National University of Singapore and Peking University


There are also rankings by disciplinary cluster and by indicator.

For every single disciplinary cluster, the University of Tokyo, not the University of Hong Kong is top. How strange.

For the indicators, the National University of Singapore is first for Employer Review and International Students, Tokyo University for Academic Peer Review, College of Medicine at Pochon Cha University (Korea) for faculty student ratio [I’m wondering about that as well], Gwangju Institute of Science and Technology for papers per faculty, Yokohama City University for Citations per Paper, Hong Kong University of Science and Technology for International Faculty, Kansai Gaidai for Inbound Exchange Students and City University of Hong Kong for Outbound Exchange Students.


These rankings seem to be a shrewd marketing move. Universities that have no chance of getting anywhere in the World University Rankings will now be able to boast that they came in the top 50 Asian universities for outbound exchange students or top 100 for citations per paper. A glance at the indicator rankings, for example, shows some Malaysian universities that one would not have thought had any chance of being in any sort of ranking. On the other hand, these rankings have been able to identify rising stars such as the Multi Media University.

There are two methodological innovations, both of which are questionable. They need to be discussed since this regional ranking could be a tryout for the global rankings. The first is the addition of two further measures of internationalization, inbound and outbound exchange students.

If internationalization is going to be a criterion, then having more measures might be a good idea. However, it is time to consider whether internationalization is actually a valid indicator of quality. Measures of internationalization do not correlate very well if at all with any other indicator and they also give an unfair advantage to the European Union and Hong Kong.

If we want to measure faculty quality, which internationalization supposedly underlies, a better method might be calculate the percentage of a random sample of teaching and research staff on university web pages who obtained degrees from the top 100 universities (on the Shanghai rankings?).

However, since QS get a lot of their bread and butter from facilitating students moving across national boundaries we are unlikely to see the end of this indicator.

The addition of number of inbound and outbound exchange students might also be very easily manipulated. If it were included in the world rankings it is likely that we will see universities setting up branch campuses a few miles away across some increasingly irrelevant frontier and then moving everybody there for their second year and calling them exchange students. So we might expect to see Queens University Belfast setting up a branch in Dundalk in the Irish Republic or the National University of Singapore in Johore in Malaysia and so on.

The other innovation is that research is measured by citations per paper, which measures the average impact of papers, and papers per faculty which measures the quantity of research in a very basic sense. This represents an improvement over the previous policy of using a single indicator. However, the problem remains that both are based on the Scopus database which aims to be as inclusive as possible. Scopus is an excellent research tool but inclusion in its database is an indicator of quality only in the broadest sense. To be credible, QS should consider finding some measure of research that measures genuine excellence.

These rankings have some surprises, the most noticeable and one lacking in face validity, is that the University of Hong Kong and not the University of Tokyo is the top university in Asia. Or perhaps this should not really be a surprise. Tokyo actually outperforms Hong Kong on all indicators except the internationalization ones and is ahead in all of the disciplinary rankings. Again, a lot of South Korean universities do very well.

It is good that QS are prepared to experiment with different indicators but the methodological innovations of these rankings do not seem to help very much.

Wednesday, May 13, 2009

Jawaharlal Nehru University has been returned to India. (see previous post)

Tuesday, May 12, 2009

Don't They Teach Geography Any More?

The QS Asian University Rankings are now out. I hope to comment in a while. For the time being. I've noticed that in the "International Students Review" (I'm not sure that it's a review but never mind), Jawaharlal Nehru University is listed as being in South Korea? I'm wondering how long it will remain there?
Asian University Rankings

QS will shortly release their Top 100 Asian Universities Rankings. It seems from the bits and pieces released so far that Australia, the Pacific and South West Asia are not included. There appear to be two innovations -- a trial run for the global rankings? -- namely counting student exchanges and including citations per paper as a measure of quality of research.

Monday, May 11, 2009

The Man with the Midas Touch

Those who are familiar with Malaysian gymnastics – admittedly not a large group -- will know who I am talking about.

There is a gymnastics coach in Malaysia who has achieved remarkable results with the women’s team of a very small state. For the last few years this state has won gold after gold at national competition. Hence the title bestowed by the Malaysian press. There does not seem to be anything special about the state – the men’s team has never done very much. Nor is there anything unusual about his training methods apart from their rigour and his intolerance for poor performance.

There must then be something about his selection methods. The state is so small that it does not have a gymnastics association and so this coach is free to select anyone he wants for training.

Recently, there was an opportunity to see just what these selection methods might be. The coach has now moved up to the national team and held a selection for young gymnasts, one of whom was my daughter, to train at the national sports centre.

The selection was quite revealing. The gymnasts were required to run, jump, do a bridge, do as many chin-ups as they could and a few simple exercises. Their height and weight were measured. There was no interest in their competition records, team spirit, motivation, leadership qualities or ability to respond to adversity.

At the end there was an interesting moment. The gymnasts were told to line up with their parents behind. There was much scurrying into position as people assumed they were being summoned for a group photograph, something without which no Malaysian event of any sort is complete. But after surveying the lineup, the coach said thank you and waved everyone away. What he was in fact doing was checking to see what the gymnasts would look like in a few years

So that was the secret of the man with the Midas touch. Assessment of basic physical skills and characteristics and reference to inherited traits. It was as though someone selected for elite universities by a simple test of general intelligence and a check on parental academic performance.

So here are two proposed experiments. Asian universities – and others – should scrap the proliferating complex of personality tests, language tests, co-curricular activities, interviews to test for leadership, politeness, sensitivity, appearance, interest and so on, profiles, course work and essays and just test for general cognitive ability.

The second experiment is that this coach and others should take heed of the global consensus and introduce personality tests, interviews and the whole paraphernalia of holistic assessment to choose future athletes.

I have a horrible feeling that the first will never happen but that one day the second will.
Comments on an Article by Andrew Oswald

I have already referred to an article by Andrew Oswald (14/12/2007) that contained some acerbic comments on claims to excellence by British universities derived from their showing on the THE-QS rankings.

He recently had an article in the Independent suggesting three measures that would allow universities to rise in the rankings. I will skip the first and third and just look at the second. He proposes that universities should be free to pay the market rate for highly rated researchers and offers Dartmouth College, which built up a top economics faculty by paying outstanding researchers appropriate salaries.

Dr Oswald is quite right but a lot of people are going to be depressed after reading his article. How can universities in Africa, Latin America and Asia get anywhere if they cannot afford to pay Ivy League level salaries? Is there anything that a university without a big pot of money can do?

The answer is that there is. In Moneyball Michael Lewis described how the Oakland As baseball team performed dramatically well even though they had only a comparatively small amount of money with which to buy players. They did it by simply ignoring the intuitions of talent scouts and looking at crude statistical data.

Basically, Oakland did the equivalent of a university scrapping search committees, interviews, personality tests, references and looking at the research done by applicants. Or choosing undergraduates by testing cognitive skills, literacy and numeracy rather than holistic assessment of leadership, communication skills, response to adversity, community involvement and so on.

At least one American university has done something like this. George Mason University has built up an excellent economics department by recruiting academics specializing in unfashionable fields that were undervalued by the academic marketplace.

Asian universities and others might consider systematically recruiting researchers whose personal characteristics and choice of unpopular research topics put them at a disadvantage when applying for academic positions. They might end up with a collection of unpleasant eccentrics but they might also see their ranking scores inching upwards.