http://www.newscientist.com/article/dn1748
As for the statistical significance, here's another view of how to do it: http://www.inference.phy.cam.ac.uk/mackay/abstracts/euro.htm...
http://www.newscientist.com/article/dn1748