北京大学出版社内容简介本书致力于从社会学的角度,系统介绍多元分析的三类主要方法原理:回归、聚类与判别分析法。书中首先介绍了三变量中控制变量的各种模式,进而对多元回归做了详尽的介绍,其中包括不同变量层次的回归模型。结合主成分分析法和因子分析法,介绍了因果模型的两种方法:路径分析和LISREL法。并设有独立章节讨论多重共线,对测量误差、量度误差理论、忽略有关变量、引入无关变量等问题都有所涉及。对多元分析中 Peking University Press Content brief introduction This book is devoted to systematically introduce three kinds of principle methods of multivariate analysis: regression, clustering and discriminant analysis from the perspective of sociology. The book first introduces the various modes of control variables in the three variables, and then introduces the multiple regression in detail, including the regression models of different levels of variables. In combination with principal component analysis and factor analysis, two methods of causal model are introduced: path analysis and LISREL. There are separate chapters discussing multicollinearity, which have implications for measurement error, measurement error theory, neglect of related variables, and introduction of extraneous variables. In multivariate analysis