李长平,胡良平.M∶N配对设计二值资料一水平多重Logistic回归分析[J].四川精神卫生杂志,2019,32(4):310-314.
M∶N配对设计二值资料一水平多重Logistic回归分析
One-level multiple Logistic regression analysis of the dichotomous choice data collected from the m∶n paried design
投稿时间:2019-08-01  
DOI:10.11886/j.issn.1007-3256.2019.04.004
中文关键词:  M∶N配对设计  派生变量  条件概率  似然函数  迭代计算
英文关键词:M∶N paired design  Derived variable  Conditional probability  Likelihood function  Iterative computation
基金项目:国家高技术研究发展计划课题资助(2015AA020102)
作者单位
李长平 天津医科大学公共卫生学院卫生统计学教研室世界中医药学会联合会临床科研统计学专业委员会 
胡良平 世界中医药学会联合会临床科研统计学专业委员会军事科学院研究生院 
摘要点击次数:
全文下载次数:
中文摘要:
      【摘要】 本文的目的是介绍m∶n配对设计二值资料一水平多重Logistic回归分析方法。首先,介绍了需要了解的基本概念;其次,介绍了构建此类回归模型的基本原理;最后,通过一个实例介绍了使用SAS实现计算的全过程。在此过程中,获得了如下四点启示:其一,有必要确保所获得的科研资料是值得分析的;其二,有必要基于定量自变量产生派生变量;其三,有必要同时采用“逐步法”“前进法”和“后退法”筛选自变量;其四,有必要采用多种方法评价不同回归模型对资料的拟合优度。
英文摘要:
      The purpose of this paper was to introduce the approach of multiple Logistic regression analysis for the binary data collected from the m∶n paired design. Firstly, the basic concepts that need to be understood were introduced. Secondly, the basic principle of constructing such regression model was presented. Finally, the whole process of computing with SAS was shown by an example. In this process, the following four useful enlightenments had been obtained: ① it was necessary to ensure that the scientific research data obtained were worth analyzing; ② it was necessary to generate the derived variables based on quantitative independent variables; ③ it was necessary to adopt the “stepwise method”“forward method” and “backward method” to screen the independent variables at the same time; ④ it was necessary to use several approaches to evaluate the goodness of fit of the different regression models.
      附件
查看全文  查看/发表评论  下载PDF阅读器
关闭
手机扫一扫