Two Infinities


Tue, 03 May 2005

Statistics in the news

The Wall Street Journal, picked up here, ran a story on a report from the Insurance Institute for Highway Safety, which can be found here. It appears that car occupants are surviving crashes with SUV's more frequently. No solid reason was offered. Maybe airbags, maybe car passengers are using seatbelts more. But despite the cartoon in the report, and despite the surge in popularity of SUV's, no one seems to have mentioned that the SUV's themselves may have been driven less aggressively by the new demographic sitting behind their steering wheels.

In the same issue of the WSJ (see here) is an article regarding the association between religious practice and reduced mortality. Never mind the cause; just imagine the difficulty of teasing a conclusion like this "...studies showed a 25 percent lower mortality rate for those who attend religious services at least weekly," from noisy data. Google turns up a research proposal dated January of this year which claims

[O]nly very limited work has seriously considered whether selectivity is responsible for this association. Indeed, there are a number of factors that may influence both religious involvement and mortality that can account for this statistical association.

And finally the new statistics on obesity from the CDC:

Being overweight (BMI of 25-29.9) was not associated with excess mortality. The study found that 87,000 fewer deaths than expected were associated with being overweight.
Apparently, these statistics are pretty good. Perhaps there is a cautionary tale in how previous statistics were not so much bad as badly extrapolated.



posted at: 22:44 | path: | permanent link to this entry