Does knowing Stuff help?
By DaveLevy on Jun 30, 2009
How important are Universities to the software industry productivity. One would hope fairly high. For various reasons, I have been considering this question and some collaborators pointed me at the Academic Ranking of World Universitiesis which is referenced at Wikipedia as well and I first referred to in this blog last November. This is produced by Shanghai Jiao Tong University, in China.
I know that a discussion on ranking methodology may not be very helpful when considering economic growth issues, but there are some quite interesting and surprising results. One of the things that pointed me there is the domination of the USA, which has over 50% of the top 100 places as it was quoted for this reason.
Sadly I haven't kept in touch with this issue since I was asked to work on other things since Xmas. I am sure that basic research drives innovation and productivity; I think that research quality and output is part of an institution's organic capability and therefore its undergraduate body and its ability to attract top students is important. I have come to the conclusion that Joy's law
"Clever People work elsewhere"
applies to academia as well, and that a lot of innovation in, and production of, software happens, outside the research institutes and departments, and also outside the traditional software industry. This is one of the reasons why public policy makers need to look at their procurement policies as well as their subsidy policies.
The rest of this article looks at the 2007 results, specifically at the UK University positions and compares them with some data points from the Guardian's Guide to Universities 2007, together with some personal prejudice, some of it informed. BTW, I can't find reference to the 2007 Guide on the web, so you might like to use this link Guardian University Guide 2006, and the 2008 results are also available. If you're planning to apply to a UK University presumably for a 2010 entry, I'd recommend getting a copy of the next book, which should be published later in the year.
Shanghai Jiao Tong University have documented their methodology on their site, or at the Wikipedia page. It is based on Nobel prize winners and the publication record of alumni and staff. One thing from observation is that Universities with large medical faculties seem to do well. It seems to have been designed with a scientific bias and for the purpose of public policy planning. From my current research, I am not able to determine the role of ICT or Software Engineering in these results. It seems that this may be a piece of research yet to be done. i.e. the creation of a ranking table for ICT teaching.
The 2007 national results are published http://www.arwu.org/rank/2007/ARWU2007Statistics.htm.
I was surprised by the fact that the UK comes a good second to the USA. The UK has 11, which are
- Cambridge, Oxford, Imperial, UCL, Manchester, Edinburgh, Bristol, Sheffield, Nottingham, Kings College London and Birmingham
Another view of UK University ranking comes from "Blackadder goes forth".
Blackadder:I then leapt on the opportunity to test you. I asked if he'd been to one of the great universities: Oxford, Cambridge, or Hull.
Blackadder:You failed to spot that only two of those are great universities!
Nurse Fletcher-Brown:You swine!
Melchett: That's right! Oxford's a complete dump! [elsewhere]
Looking at the Guardian University Guide 2007's Computer Sciences and IT page, gives a quite different view. One of the most important things to say is that the Guardian's ranking methodology is optimised for undergraduate choice and the relationship between undergraduate choice and the wealth creation activities of a university are not well understood, or at least not by me. The Guardian's score is based on assessing the staff's qualifications, what it takes to get in, spend, pupil/teacher ratios, a value add score, post graduate job prospects and inclusiveness. The methodology is discussed in the book, and in the newspaper. Their 2009 Methodology notes are on the Guardian web site. The 2007 Computer & IT top ten were,
- Imperial, St. Andrews, Cambridge, Edinburgh, Oxford, York, Surrey, Durham, Bristol, & Glasgow.
with Nottingham 11th. It interests me that the Guardian, doesn't (didn't) take the research grade of the departments into account, or maybe it does within the calculation of the teaching quality index. Its not easy to produce a Guardian fact based 'Best University' since the book is aimed at helping Undergraduates discover the best courses for themselves and the analysis is both institution and subject driven.
Personally I am surprised at how low KCL scores in the World Rankings compared with the other UK universities. It's also curious to me that the LSE, Warwick and Sussex are missing. (I may look into the numbers and see where they are and try and see why these are as they are it is likely to be methodology based, and tell us something about the methodology.) I am most curious as to where the LSE sits, which from its high numbers of overseas students, and its ability to ask for very high entry grades seems to be internationally and domestically very popular. I suppose that it might be a reflection of the science focus of the methodology, or the biases of potential students in the UK.
Since the question I am looking at is how do or can Universities add to the value of the software industry, I wonder if under-graduate students are the raw material of universities. It seems reasonable to assume that good researchers and teachers want to work at renowned (& rich) Universities, and that a University's social agenda is harder to sustain in the UK than in the primary or secondary sector. My theory is that as students and their families take more financial responsibility for their education, an assessment of life-time earnings comes into the decision framework and traditional economic criteria such as returns on investment and payback horizons are consider in more or less formal terms.
In my regional chart above, Europe includes Russia & Israel, and the obvious non EU countries (Norway & Switzerland), otherwise they're EU member states, with the UK contributing 11.
Both Canada and Sweden are punching above their weight in terms of population and even GNP, although Sweden is the host nation for the Nobel panel, which may have some relevance.
The Wikipedia page, Academic Ranking of World Universities has a sort button so you can see the institutions in order of excellence, and now has the 2008 figures, and there are other ranking methodologies and publishers.
This was been written over a number of months, and the UK fact finding over a number of years as I helped and hindered my family choose their university courses. The article was originally planned to be about the value of research to industry, but has evolved into some thoughts about the UK higher education system. I hope its useful to someone.