ABSTRACT
In Chapters 1-6, we investigated statistical methods assuming that each outcome or observation of interest had a single value. In this chapter, we will explore methods that accommodate situations where each outcome has several values which may be correlated. We first introduce the multivariate normal or multinormal distribution followed by multivariate analogs to t-tests, analysis of variance (ANOVA), and regression. We also explore how multivariate methods can be used to discriminate or classify observations to different populations.
