Table 3.

Comparing intervals and Bayes for interpreting a non-significant result.

	What does it tell you?	What do you need to link data to theory?	Amount of data needed to obtain evidence for the null?	What would be a useful stopping rule to guarantee sensitivity?
Intervals	How precisely a parameter has been estimated; a reflection of data rather than theory.	A minimal value below which the theory is refuted.	Enough to make sure the width of the interval is less than that of the null region; considerable participant numbers will typically be needed in contrast to Bayes factors.	Interval width no more than null region width and interval either completely in or completely out of the null region.
Bayes factors	The strength of evidence the data provide for one theory over another; specific to the two theories contrasted.	A rough expected value or maximum value consistent with theory.	Bayes factors ensure maximum efficiency in use of participants, given a Bayes factor measures strength of evidence.	Bayes factor either greater than three or less than a third.