ABSTRACT
Multivariate analysis is a class of analytical tools (mostly computer intensive) that provides great insight into the latent structure of the characteristics under study. It is a holistic study of multiple attributes of data which describe an instance. Several real world problems are multivariate in nature and decision making depends on a scientific summarization and visual understanding of the complex data. The classical approach of analysis in clinical studies is univariate to mean that variables are handled one at a time as if they are independent. The multi-variable approach, also known as multivariate approach takes into account the relationships among variables and permits simultaneous study of several variables.
This chapter introduces concepts like mean vectors, profile variables, summary matrices for variances, covariance and correlation. Data visualization using popular software tools is also addressed by illustrative examples to draw and understand charts like bivariate distributions, scatter matrix, Box-plot etc. An outline of important applications of multivariate statistical tools like MANOVA, repeated measures analysis, factor analysis and classification problems is provided. (168)
