Sunday, January 06, 2008

Ranking of Excellent European Graduate Programmes

The Centre for Higher Education Development (CHE) has published a Ranking of Excellent European Graduate Programmes in the natural sciences and mathematics. For their analysis they use four general indicators, and according to this ranking they then distribute 'gold', 'silver' and 'bronze' medals to the institutions. Achieving a gold (or silver) medal means that the respective institution belongs in the group of institutions which cover at least 25% (or 50%, if silver) of the total amount of publications. The share of medals therefore depends on the shape of the distribution: If there are few institutions with a large number of publications, there will be fewer medals than for a smoother distribution.

The general indicators for the classification are

Those institutions who made it into the `excellence group' were then further examined in an in-depth analysis based on institutional questionnaires and online surveys. According to their summary:
"The results show that Europe in general provides a very high level of research and of graduate teaching in the academic fields that were analysed."

The figures below show the medal distribution in physics and mathematics (p. 17 and 18 of the full report)





I wasn't aware Germany is a maths paradise.

This project is intended to be transferred to other academic fields in the near future.

You can download the full report as a PDF here.

9 comments:

Phil Warnell said...

Hi Bee,

This data seems to in the general sense be consistent with that of PISA’s I offered up in your earlier chart regarding arXiv submissions posted on Dec. 30th. It should be reminded however that Europe is becoming somewhat homogenous, in terms of its student body as it is in everything else. That is to say this is more reflective of the locations of institutions and does not serve to indicate where the students themselves originate from. For top notch quality if you were a graduate student it appears your best bet would be the U.K. followed by Germany. That is not too surprising as it has traditionally been this way for more then a century. What does stand out in such considerations is that although Switzerland would have fewer universities, as per total available, its quality is very high. Of course one would have to put up with all the ups and downs this would entail. (I’m sorry Bee I just couldn’t resist)

Best,

Phil

stefan said...

Dear Bee,

interesting... and yes, I would not have expected that math result either!

However, I do not quite get that distribution of medals. Does that mean that in the list of top universities, irrespective of the country and according to the four criteria, they start at the top and look for those institution who have produced 25% of all publications (gold) or 50% of all publications (silver), and then show the results split according to country? But should there not be some normalisation by total population? Maybe it is in the complete report.


Hi Phil,

to be sincere, I do not see a direct way to compare this ranking to the PISA rankings. Moreover, there are probably much bigger variations within the countries under study than in between.

What I am sceptic about in these studies in general is that some places may be top-ranking for some subfield of, say, physics, but not in the subject as a whole, so these places my not show up in the list, but nevertheless be the best place to go if you want to study a specific topic.

Best, Stefan

Phil Warnell said...

Hi Stefan,

“To be sincere, I do not see a direct way to compare this ranking to the PISA rankings.”

In as you are closer to the situation both geographically and culturally, I would of course default to your take on this. However, although I probably again am confused, this graph does not reflect the information on the relative quantitative level. That is to say for instance, where the Netherlands show less in terms of total quantity if the relative quantity/quality were taken into account this would be more born out. The PISA data relates to the acumen of the average student and needs to be adjusted in terms of the relative number of students to have the two comparable. Also, I qualified this in pointing out that this is more relevant to the location of the institutes, rather then the graduate student’s origins. It would then be of interest to learn when all is considered, if the two would indicate a closer correlation. As to the general spread in terms of specialty, I can form no opinion and would in this case totally default to your.

Best,

Phil

Robert said...

And France is not even on the plot in physics? Oh, statistics.

Bee said...

Dear Stefan:

What I am sceptic about in these studies in general is that some places may be top-ranking for some subfield of, say, physics, but not in the subject as a whole, so these places my not show up in the list, but nevertheless be the best place to go if you want to study a specific topic.

Yes, if you are searching for such information, you are better off with a university ranking, like e.g. the one for Germany here (look how the Frankfurt University improved in Physics).

What an analysis like the above shows is more the general research environment in the country, maybe the efficiency of governmental investment, funding generally, technological resources, programs, international competitiveness etc. Also, I think the factors they have examined in that study are very specific to PhD candidates, people who are still in their educational phase. I.e. if I were to pick a place for a senior research position I would put more emphasis on the potential of that place, the amount to which my research could be realized and would be supported, how much I could influence the future directions etc (as opposed to the cite index of my colleagues). Best,

B.

stefan said...

Dear Bee,

Look how the Frankfurt University improved in Physics.

Thanks for the link, that's nice to know :-)

BTW, I liked your example of university rankings to point out the problems of ranking everything according to monetary value very much - it shows clearly the caveats of such an approach.

Best, Stefan

Anonymous said...

This is completely off-topic, but I have to say that it pains me to see that Penrose diagram at the top of your page. I suggest that you replace it with the penrose diagram of a Schwarzschild-de Sitter black hole and then perform a topological identification so that the two of you can be together by proceeding sufficiently far in the opposite direction from the black hole.

Bee said...

Hi Anonymous:

Thanks, I will think of something less depressing, I don't believe singularities actually exist anyway. Also, since I've repeatedly mentioned we're about to move the blog, I'm currently having second thoughts. a) I haven't yet had the time to struggle with wordpress, but I am reasonably sure (me being me) I will probably fiddle around with some scripts and render them dysfunctional until my husband comes and saves me b) if you're frequently checking this blog, you will have noticed that our server in Germany isn't exactly reliable lately. it seems to be a hardware problem that should be fixed at some point, but I think we'll wait till then.

Best,

B.

Francis said...

Pity. My university didn't get any medals for physics, just one for biology.

Oh well :(.