
501
15
Multivariate Analysis II
Identifying and Developing
Relationships among
Observations and Variables
15.1 INTRODUCTION
In this chapter, we continue the study of multivariate methods by covering Multigroup Discriminant
Analysis (MDA), Canonical Correlation (CANCOR), Constrained Correspondence Analysis
(CCA), Cluster Analysis, and Multidimensional Scaling (MDS). Of these, CANCOR, MDA, and
CCA help nd relations between linear combinations of multiple independent and multiple depen-
dent variables. Using MDA we can employ these relationships to separate groups of observations.
MDS and cluster analysis deal with the issue of identifying relationships ...