Data can be extremely valuable if we are able to extract information from them. This is why multivariate data analysis is essential for business and science. This book offers an easy-to-understand introduction to the most relevant methods of multivariate data analysis. It is strictly application-oriented, requires little knowledge of mathematics and statistics, demonstrates the procedures with numerical examples and illustrates each method via a case study solved with IBM’s statistical software package SPSS. Extensions of the methods and links to other procedures are discussed and recommendations for application are given. An introductory chapter presents the basic ideas of the multivariate methods covered in the book and refreshes statistical basics which are relevant to all methods. For the 2nd edition, all chapters were checked and calculated using the current version of IBM SPSS. Contents Introduction to empirical data analysis Regression analysis Analysis of variance Discriminant analysis Logistic regression  Contingency analysis Factor analysis Cluster analysis Conjoint analysis The original German version is now available in its 17th edition. In 2015, this book was honored by the Federal Association of German Market and Social Researchers as “the textbook that has shaped market research and practice in German-speaking countries�. A Chinese version is available in its 3rd edition. On the website www.multivariate-methods.info, the authors further analyze the data with Excel and R and provide additional material to facilitate the understanding of the different multivariate methods. In addition, interactive flashcards are available to the reader for reviewing selected focal points. Download the Springer Nature Flashcards App and use exclusive content to test your knowledge.

Anhand einer Simulationsstudie vergleicht Sonja Gensler die Conjoint-Analyse und die Choice-Based-Conjoint-Analyse als Methoden zur Erfassung und Analyse von Konsumentenpraferenzen und untersucht die Moglichkeiten zur Berucksichtigung heterogener Praferenzen durch Finite-Mixture-Modelle und hierarchische Bayes-Modelle.