A statistical technique for testing for differences in the means of several data populations.
(London, United Kingdom)