摘要
一、引言传统的方差协方差分析技术尽管在数量遗传研究中获得了广泛的应用,但仍有很大的局限性。特别是随着电子计算机的应用和多元统计技术的发展,这种局限性日益暴露出来.其一是估算误差较大;其二是遗传协差阵与遗传相关阵有时非正定或接近奇异而出现所谓“病态”;其三是根据于方差协方差成分的有些估计最难于进行显著性检验。这些均给多元遗传分析和生物学解释带来根本性困难.因此。
This paper,first,presents some formulae to estimate the sampling variance, covariance and correlation coefficient of the variance and covariance under general circumstances.Then,the following basic quantitative indices that are used to estimate the precision of variance and covariance components are derived from the above results and basic principles of quantitative genetics:standard error of variance,standard error of covariance,sampling correlation coefficient between two variances,and sampling correlation coefficient between two cova- riances.There are five kinds of variance and covariance analysis models discu- ssed.There are fifty formulae or so in all.The paper gives two illustrative examples at the end.
出处
《生物数学学报》
CSCD
北大核心
1989年第2期132-143,共12页
Journal of Biomathematics