A Practical Solution to the Multivariate Behrens-Fisher Problem
提出一种基于置信椭球几何条件的多元Behrens-Fisher问题解法，是Welch近似自由度方法的扩展，并通过蒙特卡洛模拟与Yao方法比较了显著水平和检验功效。
A practical solution to the multivariate Behrens-Fisher problem is proposed. The new procedure is constructed using the geometry of confidence ellipsoids for the two mean vectors and the condition under which two ellipsoids have no points in common. The new procedure is another extension of the Welch approximate degrees-of-freedom solution as is Yao's (1965) procedure. The new procedure is compared with Yao' s procedure in terms of the level of significance and power using Monte Carlo simulation.