Answers are grounded in this paper and cite their sources.
It is standard practice that a positive reversal test is claimed on the basis of inability to reject the hypothesis that two distributions share a common mean direction, and thus the claim of a positive reversal test is in fact often based on a lack of information. This is unsatisfactory. Therefore it is suggested that positive reversal tests should be classified according to the amount of information that was available for the test. This amount of information is readily indicated by the critical angle (e.g., at the 95 per cent confidence level) between the two sample mean directions at which the hypothesis of common mean direction for the distributions would be rejected. It is recommended that 5°, 10° and 20° be used as the breakpoints in the classification.
No takes yet. Share an insight, caveat, or question.
McFadden et al. (1990) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: