ABSTRACT
This chapter explores several additional multivariate statistical analysis techniques. A common goal of multivariate statistical analysis is to determine and explain how groups of variables are related and ultimately to develop theories of causation that can be traced to those relationships (Bernard 2000). Among the many multivariate procedures developed for data analysis are multiple regression analysis, partial regression, path analysis, multiple dimensional scaling, multiple analyses of variance and covariance, multiple discriminant analysis, principal component and factor analysis, cluster analysis, and more. This chapter will explore three major families of multivariate statistics: (1) multiple discriminant analysis (MDA), a group-membership prediction and classification tool; (2) factor analysis; and (3) cluster analysis. Both factor analysis and cluster analysis can be use for reduction of large data sets and statistically summarizing data sets. Multiple analysis of variance (MANOVA) and multiple analysis of covariance (MANCOVA) are mentioned, but not discussed in any detail. Multiple regression analysis (MRA), an extension of the simple regression procedure, was discussed in Chapter 14.
