Either rank correlation could be used here. The sample size is so large that
the advantages that Kendall's has in dealing with ties will not matter. We
cannot see any advantage in quoting both. Indeed, there is a disadvantage, in
general, as were they to give different answers we would have to choose between
them. We should decide on our statistical method and stick to it.
Back to question.
This page maintained by Martin Bland.
Last updated: 8 April, 2004.