合理进行多元分析——非度量型多维尺度分析

Reasonably carry out multivariate analysis:nonmetric multidimensional scaling analysis

本文目的是介绍与非度量型多维尺度分析有关的基本概念、计算方法、两个实例以及SAS实现。基本概念包括非度量型、应用场合、基本思想和应力系数;计算方法涉及定义研究对象之间的次序关系和Kruskal算法的基本步骤;两个实例分别为“英文字母错误识别调查结果”和“6种糖果相似性的调查结果”;借助SAS软件,对两个实例中的数据分别进行了非度量型多维尺度分析,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 nonmetric multidimensional scaling analysis.Basic concepts included nonmetric,application occasions,basic ideas and stress coefficients.The calculation methods involved the definition of the order relationship between the research objects and the basic steps of the Kruskal algorithm.The data in the two examples were"survey results on misrecognition of English letters"and"survey results on the similarity of 6 candies".With the help of SAS software,nonmetric multidimensional scaling analysis was carried out on the data in the two examples,and an explanation was made for the output results of SAS.