September 30, 2008

Thomson Scientific still trying to predict the Nobel Prizes

Over the last couple of years (here, here, here), I've been razzing Thomson Scientific about their lame attempts to predict the science & economics Nobel Prizes. Last year, they predicted every single prize incorrectly.

Well, they're at it again. I won't go into detail again how misguided I think their attempts are to use their citation counts to predict the outcome of the prizes. However, I will list their predictions for this year so I can once again point to them as the prizes are awarded to show how useless something like citation counts can be for this kind of purpose.

Here goes.

Chemistry

  • Charles M. Lieber
  • Krzysztof (Kris) Matyjaszewski
  • Roger Y. Tsien


Economics
  • Lars P. Hansen and Thomas J. Sargent and Christopher A. Sims
  • Martin S. Feldstein
  • Armen A. Alchian and Harold Demsetz


Physics
  • Andre K. Geim and Kostya Novoselov
  • Vera C. Rubin
  • Sir Roger Penrose and Dan Shechtman


Physiology or Medicine
  • Shizuo Akira and Bruce A. Beutler and Jules A. Hoffmann
  • Victor R. Ambros and Gary Ruvkun
  • Rory Collins and Sir Richard Peto


So, let's see how they do this year. I predict about the same as previous years, in other words, pretty random. Some of the people they pick based on citation counts will be picked in the year Thomson guesses, some won't. Some will get picked in a later year. It's interesting, as a way a hedging their bets Thomson lists some of the previous years' "citation laureates" at the bottom of the page.

Once again, I would also like to emphasize that I have nothing against the scholars whom Thomson has "nominated" and wish them well. I certainly don't mean to cast a negative light on their contributions at all. My beef is not with them, but with Thomson's misuse of their citation data.

4 comments:

Anonymous said...

Forget their predictions, Thomson should be called out for what they're trying to do to Zotero, and the attempted lock-in of academics to EndNote:
http://www.boingboing.net/2008/09/30/reuters-sues-academi.html

John Dupuis said...

Hi David,

I agree that the Zotero issue is also very important and a lot of librarians and scholars are certainly posting and writing about it. I can feel a related post bubbling up myself.

Hopefully, the adverse publicity from something as popular as BoingBoing will get them to change their tune.

At my institution, we mostly support RefWorks but we certainly did used to push EndNote very hard. In fact for a few years I was the primary contact/trainer for EndNote so I'm well aware of it's strengths and weaknesses as a product.

Anonymous said...

Boy, talk about egg on your face.

Looks like they got Roger Tsein right.

John Dupuis said...

Hi Anon, thanks for standing up for your comment.

Sure, they got 1/3 of today's announcement right. I've never said that they'll never guess right, after all, all the people they highlight are extremely accomplished scientists.

In fact, to repeat myself: "I predict about the same as previous years, in other words, pretty random. Some of the people they pick based on citation counts will be picked in the year Thomson guesses, some won't. Some will get picked in a later year."

Yeah, some get picked in the same year Thomson guesses.

My point is that citation counts aren't valid predictors for the Nobels. There's a correlation between the two, but no causation.