>>431
If we you CI(confidence interval) and ROPE(range of practival equivalence) to reject the null hypothesis,
the false rejection rate gets much lower.

http://i.imgur.com/7RNHwz9.jpg