Saturday, 16 August 2008

Statistical errors

Did you know that there is statistical proof that storks deliver babies?

Robert Matthews gathered statistics on stork populations from 17 European counties, and compared them to national birth rates. Analysing the data, he found a significant correlation... and one that was also highly statistically significant.

Of course, the purpose of all this was to demonstrate how statistics can be misleading - especially where the data is badly selected (what about the countries that don't have any storks?)

And of course, it's important to remember that 56% of statistics are made up on the spot.

Image by Arpingstone of Wikimedia Commons
(Public domain)