Saturday, June 19, 2010

Lies, damned lies and...

This is a bit of a random post, but as "voodoo statistics" have been something of a recurring theme on this blog, I thought I'd draw attention to a textbook example of an area in which no-one seems to have the slightest clue which numbers are accurate or what they actually mean. A few weeks ago, there was a batch of stories about a possible method to slow down the aging process, and they all confidently asserted one central 'fact' - that, as things stand, an individual only has a one in 10,000 (0.01%) chance of living to 100. Those seemed startlingly long odds, given that most people probably know at least one person who's lived well into their 90s. But I just automatically assumed the figures must be authoritative - until, out of curiosity, I did a quick search engine check. It turned up these curiously contradictory nuggets of "information" -

According to the BBC, a baby boy has an 18.1% chance of living to 100, and a baby girl has a 23.5% chance.
BUT a 40-year-old man has only an 8% chance, and a 40-year-old woman has an 11.7% chance. (So the longer you live, the less chance you have?)

According to (catchy title) an individual has a 2% chance of making it to 100.

This American actuarial table seems to imply a 0.7% chance for men, and a 2.2% chance for women.

So what is the actual truth? Is there an actual truth? Answers on a postcard...

1 comment:

  1. I don't know the answer, but 1-2% intuitively sounds most plausible. It could be that journalists noticed that 1 in 10,000 of the population were centenarians, and jumped to the conclusion that it meant everyone had a 1 in 10,000 chance of living to 100. If so that really is a voodoo statistic!