摘要
对于两个相互独立且来自不同正态总体的样本,推导出其方差齐性检验问题的似然比统计量。在犯第二类错误累积概率最小的意义下,推导出其最佳检验的临界值条件,并证明了最佳检验的存在唯一性,给出了显著性水平为0.05时,样本容量从(5,5)到(121,121)的最佳检验的临界值表。
For hypothesis test on equality of variances from two independent Gamma populations, the likelihood ratio statistic is derived. The best likelihood ratio test is found in the sense of minimal cumulative probability while making the second type of mistakes, and its existence and uniqueness are proved. We have also worked out the best choice of boundary values from (5,5) to (121,121) for level0.05.
出处
《大连民族学院学报》
CAS
2009年第3期235-238,共4页
Journal of Dalian Nationalities University
关键词
正态总体
方差
似然比检验
累积概率
normal population
variance
likelihood ratio test
eumulative probability