| 关于混杂效应、交互效应的探讨 |
| |
| 引用本文： | 李丽霞,郜艳晖,张敏,周舒冬,叶小华. 关于混杂效应、交互效应的探讨[J]. 数理医药学杂志, 2011, 24(5): 510-513 |
| |
| 作者姓名： | 李丽霞 郜艳晖 张敏 周舒冬 叶小华 |
| |
| 作者单位： | 广东药学院公共卫生学院卫生统计学教研室,广州,510006 |
| |
| 摘 要： | 目的:探讨在实际处理数据时对混杂效应、交互效应的处理方法。方法:通过具体实例说明混杂效应、交互效应的概念及正确应用。结果:判断变量间是否存在交互效应,需要在模型中纳入变量间的乘积项并通过统计学检验来评价;判断某变量是否为混杂因素,需比较模型中没有纳入该变量时得到的粗估计值与该变量纳入模型时得到的校正估计值的差别是否具有实际意义的不同,而不是通过统计学检验来评价;结论:在对数据进行统计分析时,当某变量可能与其它变量存在交互作用,同时又考虑其可能为混杂因素时,应先考虑其是否存在可能的交互作用,因为研究因素的效应随其他变量的取值不同而变化,如交互作用无统计学意义,进一步评价其是否为混杂因素。
|
| 关 键 词： | 混杂效应 交互效应 条件效应图 |
The Exploration of Confounding and Interaction Effect |
| |
| Li Lixia,et al. The Exploration of Confounding and Interaction Effect[J]. Journal of Mathematical Medicine, 2011, 24(5): 510-513 |
| |
| Authors: | Li Lixia et al |
| |
| Affiliation: | Li Lixia,et al(Department of Medical Statistics,Guangdong College of Pharmacy,Guangzhou 510006) |
| |
| Abstract: | Objective: To explore the application of confounding and interaction effect in data analysis.Method: The definition and correct application of confounding and interaction are illustrated by specific examples.Results: Interaction effect exists when the relationship of interest differs at different levels of extraneous variables and is evaluated by using statistical test;confounding is present when the effect of interest differs depending on whether an extraneous variable is ignored or included in the data an... |
| |
| Keywords: | confounding interaction effect conditional effects plot |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
