Ars Physica

Another interesting article regarding P-values and its interpretation.

“If you are foolish enough to define “statistically significant” as anything less than P = 0.05 then, according to one argument (Sellke et al., 2001), you have a 29% chance (at least) of making a fool of yourself. Who would take a risk like that? Judging by the medical literature, most people would.”