As a generalization of simple correspondence analysis, multiple correspondence analysis (MCA) is a powerful technique for handling larger, more complex datasets, including the high-dimensional categorical data often encountered in the social sciences, marketing, health economics, and biomedical research. Until now, however, the literature on the subject has been scattered, leaving many in these fields no comprehensive resource from which to learn its theory, applications, and implementation. Multiple Correspondence Analysis and Related Methods gives a state-of-the-art description of this new field in an accessible, self-contained, textbook format. Explaining the methodology step-by-step, it offers an exhaustive survey of the different approaches taken by researchers from different statistical "schools" and explores a wide variety of application areas. Each chapter includes empirical examples that provide a practical understanding of the method and its interpretation, and most chapters end with a "Software Note" that discusses software and computational aspects. An appendix at the end of the book gives further computing details along with code written in the R language for performing MCA and related techniques. The code and the datasets used in the book are available for download from a supporting Web page. Providing a unique, multidisciplinary perspective, experts in MCA from both statistics and the social sciences contributed chapters to the book. The editors unified the notation and coordinated and cross-referenced the theory across all of the chapters, making the book read seamlessly. Practical, accessible, and thorough, Multiple Correspondence Analysis and Related Methods brings thetheory and applications of MCA under one cover and provides a valuable addition to your statistical toolbox.

Drawing on the author’s 45 years of experience in multivariate analysis, Correspondence Analysis in Practice, Third Edition, shows how the versatile method of correspondence analysis (CA) can be used for data visualization in a wide variety of situations. CA and its variants, subset CA, multiple CA and joint CA, translate two-way and multi-way tables into more readable graphical forms — ideal for applications in the social, environmental and health sciences, as well as marketing, economics, linguistics, archaeology, and more.

Michael Greenacre is Professor of Statistics at the Universitat Pompeu Fabra, Barcelona, Spain, where he teaches a course, amongst others, on Data Visualization. He has authored and co-edited nine books and 80 journal articles and book chapters, mostly on correspondence analysis, the latest being Visualization and Verbalization of Data in 2015. He has given short courses in fifteen countries to environmental scientists, sociologists, data scientists and marketing professionals, and has specialized in statistics in ecology and social science.

Table of Contents

Preface

Scatterplots and Maps

Profiles and the Profile Space

Masses and Centroids

Chi-Square Distance and Inertia

Plotting Chi-Square Distances

Reduction of Dimensionality

Optimal Scaling

Symmetry of Row and Column Analyses

Two-Dimensional Maps

Three More Examples

Contributions to Inertia

Supplementary Points

Correspondence Analysis Biplots

Transition and Regression Relationships

Clustering Rows and Columns

Multiway Tables

Stacked Tables

Multiple Correspondence Analysis

Joint Correspondence Analysis

Scaling Properties of MCA

Subset Correspondence Analysis

Analysis of Matches Matrices

Analysis of Square Tables

Correspondence Analysis of Networks

Data Recoding

Canonical Correspondence Analysis

Co-Inertia and Co-Correspondence Analysis

Aspects of Stability and Inference

Permutation Tests

Appendix A: Theory of Correspondence Analysis

Appendix B: Computation of Correspondence Analysis

Appendix C: Bibliography of Correspondence Analysis

Appendix D: Glossary of Terms

Appendix E: Epilogue

Index

Compositional Data Analysis in Practice is a user-oriented practical guide to the analysis of data with the property of a constant sum, for example percentages adding up to 100%. Compositional data can give misleading results if regular statistical methods are applied, and are best analysed by first transforming them to logarithms of ratios. This book explains how this transformation affects the analysis, results and interpretation of this very special type of data. All aspects of compositional data analysis are considered: visualization, modelling, dimension-reduction, clustering and variable selection, with many examples in the fields of food science, archaeology, sociology and biochemistry, and a final chapter containing a complete case study using fatty acid compositions in ecology. The applicability of these methods extends to other fields such as linguistics, geochemistry, marketing, economics and finance. R Software The following repository contains data files and R scripts from the book https://github.com/michaelgreenacre/CODAinPractice. The R package easyCODA, which accompanies this book, is available on CRAN -- note that you should have version 0.25 or higher. The latest version of the package will always be available on R-Forge and can be installed from R with this instruction: install.packages("easyCODA", repos="http://R-Forge.R-project.org").

Visualization and Verbalization of Data shows how correspondence analysis and related techniques enable the display of data in graphical form, which results in the verbalization of the structures in data. Renowned researchers in the field trace the history of these techniques and cover their current applications. The first part of the book explains the historical origins of correspondence analysis and associated methods. The second part concentrates on the contributions made by the school of Jean-Paul Benzecri and related movements, such as social space and geometric data analysis. Although these topics are viewed from a French perspective, the book makes them understandable to an international audience. Throughout the text, well-known experts illustrate the use of the methods in practice. Examples include the spatial visualization of multivariate data, cluster analysis in computer science, the transformation of a textual data set into numerical data, the use of quantitative and qualitative variables in multiple factor analysis, different possibilities of recoding data prior to visualization, and the application of duality diagram theory to the analysis of a contingency table.

This volume contains revised versions of selected papers presented at the biennial meeting of the Classi?cation and Data Analysis Group (CLADAG) of the Italian Statistical Society, which was held in Macerata, September 12-14, 2007. Carlo Laurochairedthe Scienti?c ProgrammeCommitteeand FrancescoPalumbochaired the Local Organizing Committee. The scienti?c programme scheduled 150 oral presentations and one poster s- sion. Sessions were organised in ?ve plenary sessions, 10 invited paper specialised sessions and 24 solicited paper sessions. Contributed papers and posters were 54 and 12, respectively. Five eminent scholars, who have given important impact in the Classi?cation and Data Analysis ?elds, were invited as keynote speakers, they are H. Bozdogan, S. R. Masera, G. McLachlan, A. Montanari, A. Rizzi. Invited Paper Specialised Sessions focused on the following topics: Knowledge extraction from temporal data models Statistical models with errors-in-covariates Multivariate analysis for microarray data Cluster analysis of complex data Educational processes assessment by means of latent variables models Classi?cation of complex data Multidimensional scaling Statistical models for public policies Classi? cation models for enterprise risk management Model-based clustering It is worth noting that two of the ten specialised sessions were organised by the French (Classi?cation of complex data) and Japanese (Multidimensional scaling) classi?cation societies. The SPC is grateful to professorsOkada (Japan) and Zighed (France), who took charge of the Japanese and French specialised session org- isation, respectively.

Drawing on the author’s 45 years of experience in multivariate analysis, Correspondence Analysis in Practice, Third Edition, shows how the versatile method of correspondence analysis (CA) can be used for data visualization in a wide variety of situations. CA and its variants, subset CA, multiple CA and joint CA, translate two-way and multi-way tables into more readable graphical forms — ideal for applications in the social, environmental and health sciences, as well as marketing, economics, linguistics, archaeology, and more.

Michael Greenacre is Professor of Statistics at the Universitat Pompeu Fabra, Barcelona, Spain, where he teaches a course, amongst others, on Data Visualization. He has authored and co-edited nine books and 80 journal articles and book chapters, mostly on correspondence analysis, the latest being Visualization and Verbalization of Data in 2015. He has given short courses in fifteen countries to environmental scientists, sociologists, data scientists and marketing professionals, and has specialized in statistics in ecology and social science.

Table of Contents

Preface

Scatterplots and Maps

Profiles and the Profile Space

Masses and Centroids

Chi-Square Distance and Inertia

Plotting Chi-Square Distances

Reduction of Dimensionality

Optimal Scaling

Symmetry of Row and Column Analyses

Two-Dimensional Maps

Three More Examples

Contributions to Inertia

Supplementary Points

Correspondence Analysis Biplots

Transition and Regression Relationships

Clustering Rows and Columns

Multiway Tables

Stacked Tables

Multiple Correspondence Analysis

Joint Correspondence Analysis

Scaling Properties of MCA

Subset Correspondence Analysis

Analysis of Matches Matrices

Analysis of Square Tables

Correspondence Analysis of Networks

Data Recoding

Canonical Correspondence Analysis

Co-Inertia and Co-Correspondence Analysis

Aspects of Stability and Inference

Permutation Tests

Appendix A: Theory of Correspondence Analysis

Appendix B: Computation of Correspondence Analysis

Appendix C: Bibliography of Correspondence Analysis

Appendix D: Glossary of Terms

Appendix E: Epilogue

Index

Compositional Data Analysis in Practice is a user-oriented practical guide to the analysis of data with the property of a constant sum, for example percentages adding up to 100%. Compositional data can give misleading results if regular statistical methods are applied, and are best analysed by first transforming them to logarithms of ratios. This book explains how this transformation affects the analysis, results and interpretation of this very special type of data. All aspects of compositional data analysis are considered: visualization, modelling, dimension-reduction, clustering and variable selection, with many examples in the fields of food science, archaeology, sociology and biochemistry, and a final chapter containing a complete case study using fatty acid compositions in ecology. The applicability of these methods extends to other fields such as linguistics, geochemistry, marketing, economics and finance. R Software The following repository contains data files and R scripts from the book https://github.com/michaelgreenacre/CODAinPractice. The R package easyCODA, which accompanies this book, is available on CRAN -- note that you should have version 0.25 or higher. The latest version of the package will always be available on R-Forge and can be installed from R with this instruction: install.packages("easyCODA", repos="http://R-Forge.R-project.org").

Drawing on the author's experience in social and environmental research, Correspondence Analysis in Practice, Second Edition shows how the versatile method of correspondence analysis (CA) can be used for data visualization in a wide variety of situations. This completely revised, up-to-date edition features a didactic approach with self-contained chapters, extensive marginal notes, informative figure and table captions, and end-of-chapter summaries. New to the Second Edition * Five new chapters on transition and regression relationships, stacked tables, subset correspondence analysis, analysis of square tables, and canonical correspondence analysis * Substantially more figures and tables than the first edition * A computational appendix that provides the R commands that correspond to most of the analyses featured throughout the book, making it easy for readers to reproduce the analyses With 33 years of CA experience, the expert author demonstrates how to use uncomplicated, relatively nonmathematical techniques to translate complex tabular data into more readable graphical forms. CA and its variants multiple CA (MCA) and joint CA (JCA) are suitable for analyses in various fields, including marketing research, the social and environmental sciences, biochemistry, and more.