For example, when simulation (n1 = 30, n2 = 30,
alpha = .05) used data from the Tromsø cohort
(n = 10,486), theWMWtest enjoyed an enormous power
advantage of more than 5 times the comparative statistical
power of the t-test (t-test power = .1914, WMW test
power = .9961) for the tolerance data set, and twice as
powerful (t-test power = .1944,WMWtest power = .3892)
for the pain intensity data set.