2nd Edition

Analysis of Multivariate Social Science Data

384 Pages 87 B/W Illustrations
by Chapman & Hall

384 Pages
by Chapman & Hall

384 Pages
by Chapman & Hall

Also available as eBook on:

Drawing on the authors’ varied experiences working and teaching in the field, Analysis of Multivariate Social Science Data, Second Editionenables a basic understanding of how to use key multivariate methods in the social sciences. With updates in every chapter, this edition expands its topics to include regression analysis, confirmatory factor analysis, structural equation models, and multilevel models.

After emphasizing the summarization of data in the first several chapters, the authors focus on regression analysis. This chapter provides a link between the two halves of the book, signaling the move from descriptive to inferential methods and from interdependence to dependence. The remainder of the text deals with model-based methods that primarily make inferences about processes that generate data.

Relying heavily on numerical examples, the authors provide insight into the purpose and working of the methods as well as the interpretation of data. Many of the same examples are used throughout to illustrate connections between the methods. In most chapters, the authors present suggestions for further work that go beyond conventional exercises, encouraging readers to explore new ground in social science research.

Requiring minimal mathematical and statistical knowledge, this book shows how various multivariate methods reveal different aspects of data and thus help answer substantive research questions.

Preface
Setting the Scene
Structure of the book
Our limited use of mathematics
Variables
The geometry of multivariate analysis
Use of examples
Data inspection, transformations, and missing data
Cluster Analysis
Classification in social sciences
Some methods of cluster analysis
Graphical presentation of results
Derivation of the distance matrix
Example on English dialects
Comparisons
Clustering variables
Further examples and suggestions for further work
Multidimensional Scaling
Introduction
Examples
Classical, ordinal, and metrical multidimensional scaling
Assessing fit and choosing the number of dimensions
A worked example: dimensions of color vision
Further examples and suggestions for further work
Correspondence Analysis
Aims of correspondence analysis
Carrying out a correspondence analysis: a simple numerical example
Carrying out a correspondence analysis: the general method
The biplot
Interpretation of dimensions
Choosing the number of dimensions
Example: confidence in purchasing from European Community countries
Correspondence analysis of multiway tables
Further examples and suggestions for further work
Principal Components Analysis
Introduction
Some potential applications
Illustration of PCA for two variables
An outline of PCA
Examples
Component scores
The link between PCA and multidimensional scaling and between PCA and correspondence analysis
Using principal component scores to replace the original variables
Further examples and suggestions for further work
NEW! Regression Analysis
Basic ideas
Simple linear regression
A probability model for simple linear regression
Inference for the simple linear regression model
Checking the assumptions
Multiple regression
Examples of multiple regression
Estimation and inference about the parameters
Interpretation of the regression coefficients
Selection of regressor variables
Transformations and interactions
Logistic regression
Path analysis
Further examples and suggestions for further work
Factor Analysis
Introduction to latent variable models
The linear single-factor model
The general linear factor model
Interpretation
Adequacy of the model and choice of the number of factors
Rotation
Factor scores
A worked example: the test anxiety inventory
How rotation helps interpretation
A comparison of factor analysis and principal components analysis
Further examples and suggestions for further work
Software
Factor Analysis for Binary Data
Latent trait models
Why is the factor analysis model for metrical variables invalid for binary responses?
Factor model for binary data using the item response theory approach
Goodness-of-fit
Factor scores
Rotation
Underlying variable approach
Example: sexual attitudes
Further examples and suggestions for further work
Software
Factor Analysis for Ordered Categorical Variables
The practical background
Two approaches to modeling ordered categorical data
Item response function approach
Examples
The underlying variable approach
Unordered and partially ordered observed variables
Further examples and suggestions for further work
Software
Latent Class Analysis for Binary Data
Introduction
The latent class model for binary data
Example: attitude to science and technology data
How can we distinguish the latent class model from the latent trait model?
Latent class analysis, cluster analysis, and latent profile analysis
Further examples and suggestions for further work
Software
NEW! Confirmatory Factor Analysis and Structural Equation Models
Introduction
Path diagram
Measurement models
Introduction to structural equation models with latent variables
The linear structural equation model
A worked example
Extensions
Further examples
Software
NEW! Multilevel Modeling
Introduction
Some potential applications
Comparing groups using multilevel modeling
Random intercept model
Random slope model
Contextual effects
Multilevel multivariate regression
Multilevel factor analysis
Further examples and suggestions for further work
Further topics
Estimation procedures and software
References
Index
Further reading sections appear at the end of each chapter.

Biography

David J. Bartholomew, Fiona Steele, Fiona Steele, Irini Moustaki

"… Written by some of the leaders in the field, the second edition expands the horizon of the first edition by three new chapters. The new edition enabled the authors to deal with two equally important types of methods—those for data summarization and those that are model based. … The book should provide a superb introduction to these methods for graduate students who are without substantial statistical or mathematical training … Good examples abound [and] … so do worked-out applications. … I also like the authors’ effort to compare related methods across the chapters … The website is a treasure trove … the book is essential to read … ."
—Tim Futing Liao, University of Illinois, Journal of the Royal Statistical Society, Series A, 2010

"The strength of this book lies in the right mixture of simple mathematical expressions, comprehensive non-mathematical descriptions of various multivariate approaches, numerous interesting real-life data examples, and detailed interpretation of the results. … The comprehensive web resource the authors provide is also commendable. … Overall, this is an outstanding book on multivariate statistics in the field of social sciences, with a strong focus on categorical data. It can be recommended without reservations for quantitative graduate courses in psychology, sociology, education, and related areas. …"
Journal of Statistical Software, February 2009

"…I am pleased that the authors emphasise that the book is in no sense a cookbook. … the presentation is well matched to its intended audience, relying on only the minimal necessary mathematics and driving the development with examples, figures, and verbal descriptions. …This is the sort of book from which I would have liked to have learnt multivariate statistics."
International Statistical Review, 2008