Talk:Kendall's W

p-value for Kendall's W?
It would be very helpful if this page had some indication about how a p-value is derived for a given Kendall's W test... anyone? --mcld (talk) 11:18, 25 February 2008 (UTC)

Me too. The Legendre paper referenced suggests that a Chi-squared test is appropriate for large n and m, but I haven't figured out the details yet. --69.168.48.167 (talk) 08:57, 7 August 2010 (UTC)