合理进行多元分析——定性资料对应分析和Shannon信息量分析

Reasonably carry out multivariate analysis:qualitative data correspondence analysis and Shannon information quantity analysis

本文目的是介绍与定性资料对应分析和Shannon信息量分析有关的基本概念、计算方法、两个实例以及SAS实现。基本概念包括列联表与Burt表、边缘概率、行剖面与列剖面、信息和信息量、熵;计算方法涉及定性资料对应分析和Shannon信息量分析;两个实例分别为“某医院3年间不同季节4种甲状腺疾病的检出情况”和“不同专业学生的4种气质类型分布”;借助SAS软件,对两个实例中的数据分别进行了定性资料对应分析和Shannon信息量分析,并对SAS输出结果做出了解释。

The purpose of this article was to introduce the basic concepts,calculation methods,two examples and SAS implementation related to the qualitative data correspondence analysis and Shannon information quantity analysis.Basic concepts included contingency table and Burt table,marginal probability,row profile and column profile,information and information volume,entropy.The calculation method involved the qualitative data correspondence analysis and Shannon information quantity analysis.The two examples were"detection of 4 kinds of thyroid diseases in different seasons in a hospital in 3 years"and"distribution of 4 temperament types of students in different majors".With the help of SAS software,qualitative data correspondence analysis and Shannon information quantity analysis were carried out on the data in the two examples,and an explanation was made for the SAS output results.