[mikraig]

In Bayesian analysis, it's possible to actually accept the null, rather than just the reject/not reject that is common to frequentist statistical inference.