Physics Today
... an occasional look at foibles and funnies of the arxiv, mainly in cases where natural scientists do silly things while reinventing the wheel in social science statistics. Here we see three intrepid adventurers into "scientometrics", that being the statistical analysis of scientific journals.
In principle, this is a sensible thing to do. In execution ... perhaps not so much. Apparently, the frequency with which the words "simulation" and "Monte Carlo" appear in physics journals has a statistically significant correlation with average incomes in Argentina, the number of doctoral degrees granted in the USA and the number of collaborations at CERN. What they don't publish is that, via the spurious regression phenomenon, their dataset will also show a statistically significant correlation with anything else that has gone up over the period in a sufficiently smooth fashion.
(NB: physicists should only feel a little bit ashamed of this; most econometrics papers before the early 70s also suffered from the problem of regressing two nonstationary series on each other and then patting yourself on the back for getting a high r^2).
I don't understand why, but might it not be a spoof?
ReplyDeleteI suppose so but if so it's bone dry and I was fooled. I think they're just basically trying to back up someone's assertion in a lunchtime conversation somewhere that the reason so many Monte Carlo simulations are done these days is that they're cheaper than experimental work.
ReplyDeleteColour me incredibly unsurprised that as computing power becomes exponentially cheaper, the usage of Monte Carlo models goes up.
ReplyDeleteAgreed it is bone dry, but why would you look for the conclusion Richard mentions and not actually try run a regression on 'computing power'?
ReplyDeleteHah, these people clearly need blogs to publish this sort of shit on.
ReplyDeleteYou'll be relieved to know that economics or sociology are far from the only fields to be subject to this sort of attention from physicists.
Yes, physicists should take maths lessons from economists. Yes.
ReplyDeleteAnd everybody should give a fuck what you think about law, presumably for similar reasons.
Yes, physicists should take maths lessons from economists. Yes.
ReplyDeleteHard theoretical maths: no.
The sensible use of statistics, with specific reference to "traps that data mining can set for you": yes, obviously, and only a blinkered nutter who thinks that Physics, Maths and Engineering are real and everything else is some pansy nonsense would say otherwise.
Yes, physicists should take maths lessons from economists. Yes.
ReplyDeleteTo be honest, physicists can take maths lessons from who they like, but if they want to get things right, they're going to have to learn about spurious regression of nonstationary series from someone. Learn it from a biostatistician if you're worried about the social science cooties.