On Comparing Variances in the Paired Case with Incomplete Data
通过蒙特卡洛模拟比较了配对数据不完整时几种方差相等检验方法，推荐中等相关性时使用Ekbohm检验，高相关性时使用本文提出的新检验。
Several tests for the equality of variances in the paired case with incomplete data on both responses are compared using Monte Carlo simulations. The main result is the recommendation to use a test due to Ekbohm (1981) for cases with moderate correlations. A new test proposed in this paper may be the best choice when the correlation is high.