摘要
双重筛选逐步回归是一种多自变量对多因变量的多元线性逐步回归。本文着重介绍其计算原理、步骤及应用注意事项,用一计算实例对其进行说明.此法应用灵活、方便、方法较简单、分析功效较高。其用途主要三方前:1.对病因进行多因多果的分析;2.根据自变量与因变量的关系对因变量进行分类;3.研究综合指标的影响因素.在因变量个数为一时,它与一般的逐步回归是等价的。
Doube-slcreening stepwise regression(DSSR)isa form of multiple linear stepwise regression withmultiple ind ependent variables and multiple dependentvariables.The basic principles and procedures areintroduced.Its application is illustrated with an ex-ample using physical characteristics as predictors ofphysical development.The advantages of the modelare summarized as follows:1.It can be used to(?) tablish a multi-causal and multi-outcome modlefor etiological studies;2.It may be used to helpclassif ymultiple dependent variables;and 3.It issuitable for analyzing the factors associated withmultiple dependent variables.
出处
《中国卫生统计》
CSCD
北大核心
1989年第3期44-48,共5页
Chinese Journal of Health Statistics
关键词
双重筛选
逐步回归
多元回归
Multiple Regression
Stepwise Regression
Physique Development