Multivariable Modeling and Multivariate Analysis for the Behavioral Sciences (Hardback) book cover

Multivariable Modeling and Multivariate Analysis for the Behavioral Sciences

By Brian S. Everitt

© 2009 – CRC Press

320 pages | 113 B/W Illus.

Purchasing Options:$ = USD
Hardback: 9781439807699
pub: 2009-09-28
Hardback: 9781138464551
pub: 2017-08-15
eBook (VitalSource) : 9781439807705
pub: 2009-09-28
from $39.58

FREE Standard Shipping!
e–Inspection Copy


Multivariable Modeling and Multivariate Analysis for the Behavioral Sciences shows students how to apply statistical methods to behavioral science data in a sensible manner. Assuming some familiarity with introductory statistics, the book analyzes a host of real-world data to provide useful answers to real-life issues.

The author begins by exploring the types and design of behavioral studies. He also explains how models are used in the analysis of data. After describing graphical methods, such as scatterplot matrices, the text covers simple linear regression, locally weighted regression, multiple linear regression, regression diagnostics, the equivalence of regression and ANOVA, the generalized linear model, and logistic regression. The author then discusses aspects of survival analysis, linear mixed effects models for longitudinal data, and the analysis of multivariate data. He also shows how to carry out principal components, factor, and cluster analyses. The final chapter presents approaches to analyzing multivariate observations from several different populations.

Through real-life applications of statistical methodology, this book elucidates the implications of behavioral science studies for statistical analysis. It equips behavioral science students with enough statistical tools to help them succeed later on in their careers. Solutions to the problems as well as all R code and data sets for the examples are available at


Clarity and conciseness have always been the hallmarks of Everitt’s writing. This book is no exception. Anyone looking for a clearly written text on the subject that is also practitioner oriented needs to look no further.

—Chuck Chakrapani, Journal of the Royal Statistical Society, Series A, 2012

… a clear, well-orchestrated guide to multivariate statistics for the post-graduate and professional behavioural scientist who possesses basic statistical knowledge. … Everitt successfully crafts a well-integrated introductory text that obviates potential difficulties by including real problems and their data sets. … the book’s applied orientation introduces the behavioural scientist to both the use and rudimentary understanding of multivariate techniques. … The book would also serve well as a training guide for the practitioner less experienced in multivariate techniques. …

Psychometrika, June 2010

… The first two chapters give a magnificent introduction before approaching the modeling issues. Especially the second chapter, which shows how to look at data, is among the best I have ever seen in books on multivariate methods. … He also goes well beyond the typical graphs showing how to explore real insights of the data. … the book is extremely easy to browse and read. … Putting the R code in an appendix and on the website is an excellent choice. … the huge experience of the author … makes the presentation so clear and understandable. I’ll be happy to recommend this book to students and researchers.

International Statistical Review, 2010

Table of Contents

Data, Measurement, and Models


Types of Study

Types of Measurement

Missing Values

The Role of Models in the Analysis of Data

Determining Sample Size

Significance Tests, p-Values, and Confidence Intervals

Looking at Data


Simple Graphics—Pie Charts, Bar Charts, Histograms, and Boxplots

The Scatterplot and Beyond

Scatterplot Matrices

Conditioning Plots and Trellis Graphics

Graphical Deception

Simple Linear and Locally Weighted Regression


Simple Linear Regression

Regression Diagnostics

Locally Weighted Regression

Multiple Linear Regression


An Example of Multiple Linear Regression

Choosing the Most Parsimonious Model When Applying Multiple Linear Regression

Regression Diagnostics

The Equivalence of Analysis of Variance and Multiple Linear Regression, and An

Introduction to the Generalized Linear Model


The Equivalence of Multiple Regression and ANOVA

The Generalized Linear Model

Logistic Regression


Odds and Odds Ratios

Logistic Regression

Applying Logistic Regression to the GHQ Data

Selecting the Most Parsimonious Logistic Regression Model

Survival Analysis


The Survival Function

The Hazard Function

Cox’s Proportional Hazards Model

Linear Mixed Models for Longitudinal Data


Linear Mixed Effects Models for Longitudinal Data

How Do Rats Grow?

Computerized Delivery of Cognitive Behavioral Therapy—Beat the Blues

The Problem of Dropouts in Longitudinal Studies

Multivariate Data and Multivariate Analysis


The Initial Analysis of Multivariate Data

The Multivariate Normal Probability Density Function

Principal Components Analysis



Finding the Sample Principal Components

Should Principal Components Be Extracted from the Covariance or the Correlation


Principal Components of Bivariate Data with Correlation Coefficient r

Rescaling the Principal Components

How the Principal Components Predict the Observed Covariance Matrix

Choosing the Number of Components

Calculating Principal Component Scores

Some Examples of the Application of PCA

Using PCA to Select a Subset of the Variables

Factor Analysis


The Factor Analysis Model

Estimating the Parameters in the Factor Analysis Model

Estimating the Numbers of Factors

Fitting the Factor Analysis Model: An Example

Rotation of Factors

Estimating Factor Scores

Exploratory Factor Analysis and PCA Compared

Confirmatory Factor Analysis

Cluster Analysis


Cluster Analysis

Agglomerative Hierarchical Clustering

k-Means Clustering

Model-Based Clustering

Grouped Multivariate Data


Two-Group Multivariate Data

More Than Two Groups


Appendix: Solutions to Selected Exercises


A Summary and Exercises appear at the end of each chapter.

About the Author

Brian S. Everitt is Professor Emeritus at King’s College, London, UK.

About the Series

Chapman & Hall/CRC Statistics in the Social and Behavioral Sciences

Learn more…

Subject Categories

BISAC Subject Codes/Headings:
MATHEMATICS / Probability & Statistics / General