摘要
目的:介绍应用多重对应分析方法实现对大样本分类变量之间的相关性分析,旨在对大样本疾病信息采集资料的临床和基础科研工作者提供可借鉴的统计学分析方法。方法:应用SPSS11.5统计分析软件中Data Reduction菜单的Optimal Scaling过程对大样本分类变量进行多重对应分析,以研究它们之间的联系。结果:慢性胃炎的4种病理组织诊断与8种中医证型有一定的相关性,这对慢性胃炎的临床诊断及治疗具有指导意义,为慢性胃炎中医证候规范化研究提供一定的思路。结论:对于中医研究中常见的无序多分类或二分类变量,要同时研究它们之间的联系,用多重对应分析在结果的直观性和可解释性方面优于对数线性模型。SPSS11.5统计分析软件易于实现多个变量间的多重对应分析。
Obiective: To introduce the method of multiple corresponding analysis in correlation analysis of big sample,and offer a reference statistical method to clinical and basic medicine researchers who take the design of big sample categorical variable in TCM. Methods: Using optimal scaling process of the data reduction in SPSS 11.5 to apply multiple corresponding analysis in correlation analysis. Results: There is some relevance between the pathology organization diagnoses and Chinese medicine syndrome of chronic gastritis. It has guiding sense to clinical diagnosis and treatment, and provides the mentality to the standardized research of Chinese medicine syndrome. Conclusion:The multiple corresponding analysis is suitable for correlation analysis of big sample categorical variable. SPSS statistical package is available to fufil this process.
出处
《数理医药学杂志》
2010年第1期65-66,共2页
Journal of Mathematical Medicine
关键词
多重对应分析
相关性分析
大样本
分类变量
multiple corresponding analysis
correlation analysis
big sample
categorical variable