Advanced multivariate analysis methods. An application oriented introduction (Q5899492)

The first edition of the textbook ``Multivariate Analysemethoden'', authored by \textit{C. Schuchard-Ficher} et al. [Zbl 0433.62003], appeared in 1980. Since then the book has become very popular in particular in the German marketing community so that 2016 saw already the 14th edition. Moreover, with the appearance of the 12th edition in 2008 [Zbl 1270.62085], the authors decided to divide the material into basic multivariate statistical methods and advanced multivariate statistical methods and consequently published two corresponding separate volumes, the volume on advanced methods being entitled ``Fortgeschrittene Multivariate Analysemethoden''. In 2015 appeared the 3rd edition of this part of the project, which is the subject of this review. Seven groups of multivariate statistical methods are presented in seven chapters entitled nonlinear regression, structural equation modeling, confirmatory factor analysis, choice based conjoint analysis, neural networks, multidimensional scaling, and correspondence analysis. As becomes clear from the subtitle of the book, ``Eine anwendungsorientierte Einführung'' -- an application-oriented introduction, the principle of the authors is to provide insight into and understanding of the presented methods with minimal mathematical prerequisites. In addition to the general description of the methods many examples are given to illustrate them, where one marketing example is used -- as far as feasible -- for the different methods. This helps considerably to understand that for dealing with a real world problem involving statistical data very often there is no single method of choice, but different methods may be applied successfully. For the numerical computations in the examples mainly IBM SPSS is used, but also MS Excel to some extent, and AMOS in the area of structural equation modeling. The instruction in the use of this statistical software is an important aim of the book. Under \url{http://www.multivariate.de}, the reader finds a lot of complementary material as well as a number of services (in German). This book is not meant for the mathematical statistician, and therefore not for the typical reader of Zentralblatt. It helps anyone who has to apply multivariate statistical methods in his or her work, or has to understand the implications of the application of such methods, to get into the field. It will always be debatable if a full understanding can really be achieved if the mathematical background is left out more or less completely. In any case, a thorough study of a book like this will prepare the reader for a successful cooperation with a statistical consultant.