摘要
本文讨论了通径分析理论与应用中常见的几个问题.分析了剩余效应对结果的决定系数的理论组成,和剩余效应与观测的自变量间存在相关时对通径分析结果的影响.
Some theoretical and practical problems of path analysis are discussed in this paper.The so-called residual effect usually includes measure error and some unmeasured causes.The theoretical structure of coefficient of determination of residual effect are exposed.There are generally correlations among these un-measured causes and the measured causes.In this situation the estimatad path coefficients are incorrect.It is also notes that path analysis can only be used to deal with linear models.Any application to non-linear models may lead to the wrong result.
出处
《生物数学学报》
CSCD
北大核心
1995年第3期185-189,共5页
Journal of Biomathematics