RT @mattjhodgkinson: At p<0.05 threshold "17-25% of findings are probably false" says Bayesian analysis. @NatureNews http://t.co/wkQFQptb

2:41pm November 13th 2013 via Hootsuite