statistics and applied probability

national university of singapore

Month: September 2010

Two amusing things from Gelman’s blog

1. Check out the authors of this paper in Physics Letters B. (You need to scroll to the end.) This makes medicine look selective.

2. Learn how journalists structure science stories here.

Google searches and flu

You wait a while for a bus then two come at once. So it seemed it is with papers, when Emerging Infectious Diseases published two of my articles yesterday. One is a letter looking at google searches for influenza, H1N1, etc, against time, and then comparing these to an actual measure of influenza activity (acute respiratory infections in the polyclinics). Not really much relation: big spikes in the number of searches after newsworthy events, not much search activity during the outbreak itself. Previous papers had suggested internet searches are a good proxy for flu surveillance, but we suggest otherwise.

Disappointingly, I wasn’t able to sneak the simplified Chinese search terms we also tried into figure caption: the copyeditors were quick to remove this. 🙁

PS The other article was about H1N1 seroconversions in health care workers.
PPS A question was raised about how to abbreviate names with just one letter: often in acknowledgments, a name like Alex Richard Cook would be abbreviated A.R.C. for short, right? Well, one co-author is called Mark I-Cheng Chen. And they (the copy editor and Mark himself) wanted to abbreviate this M.I.-C.C. but I don’t think this is right, as the I isn’t short for anything, and so doesn’t need an abbreviating full stop. Comments on this trivium welcome.

Update on Japan’s missing centurians

Further to a previous post, the been report that almost a quarter of a million Japanese on the books as being 100 years old or older and alive are “unaccounted for”. Apparently 100s of these would be over 150 if they really were alive! What kind of impact would this have on the average lifespan if they were included in its calculation, I wonder?

74% of Economist readers like maths

A debate over at the Economist, on the motion: promoting maths and sciences education is the best way to stimulate future innovation. On day 1 of the debate, 74% of voting readers supported the motion!

Personally, I’m opposed to the motion (or agnostic to it). There’s no evidence in favour of maths/sciences being the optimal way to stimulate innovation (how about engineering? is their bang bigger per buck?), although surely it’s a pretty good route to innovation.

Perhaps independently, the Times Higher Education has an article reporting a “statistical analysis” by Prof Whiteley of Essex (a political scientist) suggesting that investing in science is associated with only a correlation of 11% with economic growth. The analysis has, however, been rubbished by many commentators at THE.

Skip to toolbar