NORMALIZING AND VARIANCE STABILIZING TRANSFORMATIONS FOR INTRACLASS CORRELATIONS
NORMALIZING AND VARIANCE STABILIZING TRANSFORMATIONS FOR INTRACLASS CORRELATIONS
复制标题
DOI：
10.1007/bf02481082
复制
发表时间：
1985-01-01
影响因子：
1
通讯作者：
KONISHI, S
中科院分区：
文献类型：
--
作者：
KONISHI, S
统计归一化变换的概念是建立在收敛到正态的速度的基础上的。应用这一概念推导了类内相关系数在ap变量正态样本中最大似然估计的归一化变换。通过数值比较来检验所提出的变换是否能有效地达到正态性。还考虑了归一化与方差稳定之间的关系。
A concept of normalizing transformations of statistics is constructed on the basis of the rate of convergence to normality. The concept is applied to derive a normalizing transformation of a maximum likelihood estimate of intraclass correlation coefficient in ap-variate normal sample. Numerical comparisons are made to examine whether the proposed transformation is efficient to achieve normality. The relationship between normalization and variance stabilization is also considered.