Saturday, 27 January 2007

The Causality of Obesity

The worldwide shortage of pirates is causing global temperatures to rise! As pirate numbers have decreased since the halcyon, pirate friendly days of the 18th and 19th centuries global temperatures have risen, damaging ecosystems, causing sea-levels to rise and even threatening coastal cities with utter annihilation. Surely a bit of good-natured piracy is preferable? It certainly worked for Johnny Depp.
There are, as Benjamin Disraeli said, three types of lies: Lies, damn lies, and statistics. Just because there is a correlation between decreased swashbuckling and increased temperatures, does that mean that there is a causal relationship? It seems pretty clear that there is not. The same logic, however, is used often in the popular media and even in serious scientific publications. Recently, in articles about obesity, "scientific articles" have been cited as stating that urban sprawl causes obesity. Look!, they say, there are fat people living in the suburbs, so the suburbs must cause people to be fat. It's a tempting jump to make. People in the suburbs do spend more time in their cars, and they do tend to walk less. But does living in the suburbs make people fat? Probably not, or at the very least there is not enough evidence to prove it. Several studies, in fact, have come to the conclusion that low-density areas merely attract inactive people who are prone to obesity.
It is extremely tempting to make statements which are based on a correlation and phrase them in such a way as to imply a causal relationship. It makes for interesting reading and adds a veneer of respectability, but science it is not. It pays to use a skeptical eye when reading the news, just remember the pirates.

1 comment:

apple said...

It looks like a pirate shaped balloon.
Hot air balloon rides