Applied multivariate data analysis. (Q2778895)

[For the review of the first edition from 1991 see Zbl 0788.62001.]NEWLINENEWLINENEWLINEThis book deals with multivariate analysis including methods both for describing and exploring data and for more formal inferential procedures. Multivariate analysis plays an important role in the understanding of complex data sets requiring simultaneous examination of all variables. The aim of its techniques is, generally, to display or extract the signal in the data in the presence of noise, and to extract the underlying patterns and structures.NEWLINENEWLINENEWLINEThe book consists of 13 chapters. The first one provides a general introduction to multivariate data and multivariate statistics. The next five chapters concentrate on what might loosely be described as exploratory multivariate techniques including graphical methods, principal components and correspondence analysis, multidimensional scaling and cluster analysis. Often, these techniques are graphical in nature and provide the means to unravel patterns in the considered data.NEWLINENEWLINENEWLINEThe following five chapters consider what might be viewed as a more formal approach, in which specific models are fitted to the data and hypotheses of interest are tested. These chapters include linear modelling, regression and variance analysis, log-linear and logistic models, models for multivariate response variables, discrimination, classification and pattern recognition techniques. Within the final two chapters, models are considered that involve latent variables for which no direct methods of measurement exist. More specifically, these two chapters deal with exploratory and confirmatory factor analyses. Each of the aforementioned chapters is accompanied by a number of exercises that allow for the best understanding of the presented subjects.NEWLINENEWLINENEWLINEThe book aims both at students in applied statistics courses, and researchers dealing with multivariate data. The readers need to have some basic background in statistical issues such as estimation, inference, regression, analysis of variance and so on, while the main mathematical requirement is a degree of familiarity with matrix algebra.