多元数据分析 第7版 英文版【2025|PDF|Epub|mobi|kindle电子书版本百度云盘下载】

多元数据分析 第7版 英文版PDF格式电子书版下载

注意：本站所有压缩包均有解压码： 点击下载压缩包解压工具

Chapter 1 Introduction:Methods and Model Building1

What Is Multivariate Analysis?3

Multivariate Analysis in Statistical Terms4

Some Basic Concepts of Multivariate Analysis4

The Variate4

Measurement Scales5

Measurement Error and Multivariate Measurement7

Statistical Significance Versus Statistical Power8

Types of Statistical Error and Statistical Power9

Impacts on Statistical Power9

Using Power with Multivariate Techniques11

A Classification of Multivariate Techniques11

Dependence Techniques14

Interdependence Techniques14

Types of Multivariate Techniques15

Principal Components and Common Factor Analysis16

Multiple Regression16

Multiple Discriminant Analysis and Logistic Regression16

Canonical Correlation17

Multivariate Analysis of Variance and Covariance17

Conjoint Analysis18

Cluster Analysis18

Perceptual Mapping19

Correspondence Analysis19

Structural Equation Modeling and Confirmatory Factor Analysis19

Guidelines for Multivariate Analyses and Interpretation20

Establish Practical Significance as Well as Statistical Significance20

Recognize That Sample Size Affects All Results21

Know Your Data21

Strive for Model Parsimony21

Look at Your Errors22

Validate Your Results22

A Structured Approach to Multivariate Model Building22

Stage 1:Define the Research Problem,Objectives,and Multivariate Technique to Be Used23

Stage 2:Develop the Analysis Plan23

Stage 3:Evaluate the Assumptions Underlying the Multivariate Technique23

Stage 4:Estimate the Multivariate Model and Assess Overall Model Fit23

Stage 5:Interpret the Variate(s)24

Stage 6:Validate the Multivariate Model24

A Decision Flowchart24

Databases24

Primary Database25

Other Databases27

Organization of the Remaining Chapters28

Section Ⅰ:Understanding and Preparing For Multivariate Analysis28

Section Ⅱ:Analysis Using Dependence Techniques28

Section Ⅲ:Interdependence Techniques28

Section Ⅳ:Structural Equations Modeling28

Summary28

Questions30

Suggested Readings30

References30

SECTION Ⅰ Understanding and Preparing For Multivariate Analysis31

Chapter 2 Cleaning and Transforming Data33

Introduction36

Graphical Examination of the Data37

Univariate Profiling:Examining the Shape of the Distribution38

Bivariate Profiling:Examining the Relationship Between Variables39

Bivariate Profiling:Examining Group Differences40

Multivariate Profiles41

Missing Data42

The Impact of Missing Data42

A Simple Example of a Missing Data Analysis43

A Four-Step Process for Identifying Missing Data and Applying Remedies44

An Illustration of Missing Data Diagnosis with the Four-Step Process54

Outliers64

Detecting and Handling Outliers65

An Illustrative Example of Analyzing Outliers68

Testing the Assumptions of Multivariate Analysis70

Assessing Individual Variables Versus the Variate70

Four Important Statistical Assumptions71

Data Transformations77

An Illustration of Testing the Assumptions Underlying Multivariate Analysis79

Incorporating Nonmetric Data with Dummy Variables86

Summary88

Questions89

Suggested Readings89

References90

Chapter 3 Factor Analysis91

What Is Factor Analysis?94

A Hypothetical Example of Factor Analysis95

Factor Analysis Decision Process96

Stage 1:Objectives of Factor Analysis96

Specifying the Unit of Analysis98

Achieving Data Summarization Versus Data Reduction98

Variable Selection99

Using Factor Analysis with Other Multivariate Techniques100

Stage 2:Designing a Factor Analysis100

Correlations Among Variables or Respondents100

Variable Selection and Measurement Issues101

Sample Size102

Summary102

Stage 3:Assumptions in Factor Analysis103

Conceptual Issues103

Statistical Issues103

Summary104

Stage 4:Deriving Factors and Assessing Overall Fit105

Selecting the Factor Extraction Method105

Criteria for the Number of Factors to Extract108

Stage 5:Interpreting the Factors112

The Three Processes of Factor Interpretation112

Rotation of Factors113

Judging the Significance of Factor Loadings116

Interpreting a Factor Matrix118

Stage 6:Validation of Factor Analysis122

Use of a Confirmatory Perspective122

Assessing Factor Structure Stability122

Detecting Influential Observations123

Stage 7:Additional Uses of Factor Analysis Results123

Selecting Surrogate Variables for Subsequent Analysis123

Creating Summated Scales124

Computing Factor Scores127

Selecting Among the Three Methods128

An Illustrative Example129

Stage 1:Objectives of Factor Analysis129

Stage 2:Designing a Factor Analysis129

Stage 3:Assumptions in Factor Analysis129

Component Factor Analysis:Stages 4 Through 7132

Common Factor Analysis:Stages 4 and 5144

A Managerial Overview of the Results146

Summary148

Questions150

Suggested Readings150

References150

SECTION Ⅱ Analysis Using Dependence Techniques153

Chapter 4 Simple and Multiple Regression155

What Is Multiple Regression Analysis?161

An Example of Simple and Multiple Regression162

Prediction Using a Single Independent Variable:Simple Regression162

Prediction Using Several Independent Variables:Multiple Regression165

Summary167

A Decision Process for Multiple Regression Analysis167

Stage 1:Objectives of Multiple Regression169

Research Problems Appropriate for Multiple Regression169

Specifying a Statistical Relationship171

Selection of Dependent and Independent Variables171

Stage 2:Research Design of a Multiple Regression Analysis173

Sample Size174

Creating Additional Variables176

Stage 3:Assumptions in Multiple Regression Analysis181

Assessing Individual Variables Versus the Variate182

Methods of Diagnosis183

Linearity of the Phenomenon183

Constant Variance of the Error Term185

Independence of the Error Terms185

Normality of the Error Term Distribution185

Summary186

Stage 4:Estimating the Regression Model and Assessing Overall Model Fit186

Selecting an Estimation Technique186

Testing the Regression Variate for Meeting the Regression Assumptions191

Examining the Statistical Significance of Our Model192

Identifying Influential Observations194

Stage 5:Interpreting the Regression Variate197

Using the Regression Coefficients197

Assessing Multicollinearity200

Stage 6:Validation of the Results206

Additional or Split Samples206

Calculating the PRESS Statistic206

Comparing Regression Models206

Forecasting with the Model207

Illustration of a Regression Analysis207

Stage 1:Objectives of Multiple Regression207

Stage 2:Research Design of a Multiple Regression Analysis208

Stage 3:Assumptions in Multiple Regression Analysis208

Stage 4:Estimating the Regression Model and Assessing Overall Model Fit208

Stage 5:Interpreting the Regression Variate223

Stage 6:Validating the Results226

Evaluating Alternative Regression Models227

A Managerial Overview of the Results231

Summary231

Questions234

Suggested Readings234

References234

Chapter 5 Canonical Correlation235

What Is Canonical Correlation?237

Hypothetical Example of Canonical Correlation238

Developing a Variate of Dependent Variables238

Estimating the First Canonical Function238

Estimating a Second Canonical Function240

Relationships of Canonical Correlation Analysis to Other Multivariate Techniques241

Stage 1:Objectives of Canonical Correlation Analysis242

Selection of Variable Sets242

Evaluating Research Obiectives242

Stage 2:Designing a Canonical Correlation Analysis243

Sample Size243

Variables and Their Conceptual Linkage243

Missing Data and Outliers244

Stage 3:Assumptions in Canonical Correlation244

Linearity244

Normality244

Homoscedasticity and Multicollinearity244

Stage 4:Deriving the Canonical Functions and Assessing Overall Fit245

Deriving Canonical Functions246

Which Canonical Functions Should Be Interpreted?246

Stage 5:Interpreting the Canonical Variate250

Canonical Weights250

Canonical Loadings250

Canonical Cross-Loadings250

Which Interpretation Approach to Use251

Stage 6:Validation and Diagnosis251

An Illustrative Example252

Stage 1:Objectives of Canonical Correlation Analysis253

Stages 2 and 3:Designing a Canonical Correlation Analysis and Testing the Assumptions253

Stage 4:Deriving the Canonical Functions and Assessing Overall Fit253

Stage 5:Interpreting the Canonical Variates254

Stage 6:Validation and Diagnosis257

A Managerial Overview of the Results258

Summary258

Questions259

References260

Chapter 6 Conjoint Analysis261

What Is Conjoint Analysis?266

Hypothetical Example of Conjoint Analysis267

Specifying Utility,Factors,Levels,and Profiles267

Gathering Preferences from Respondents268

Estimating Part-Worths269

Determining Attribute Importance270

Assessing Predictive Accuracy270

The Managerial Uses of Conjoint Analysis271

Comparing Conjoint Analysis with Other Multivariate Methods272

Compositional Versus Decompositional Techniques272

Specifying the Conjoint Variate272

Separate Models for Each Individual272

Flexibility in Types of Relationships273

Designing a Conjoint Analysis Experiment273

Stage 1:The Objectives of Conjoint Analysis276

Defining the Total Utility of the Object276

Specifying the Determinant Factors276

Stage 2:The Design of a Conjoint Analysis277

Selecting a Conjoint Analysis Methodology278

Designing Profiles:Selecting and Defining Factors and Levels278

Specifying the Basic Model Form283

Data Collection286

Stage 3:Assumptions of Conjoint Analysis293

Stage 4:Estimating the Conjoint Model and Assessing Overall Fit294

Selecting an Estimation Technique294

Estimated Part-Worths297

Evaluating Model Goodness-of-Fit298

Stage 5:Interpreting the Results299

Examining the Estimated Part-Worths300

Assessing the Relative Importance of Attributes302

Stage 6:Validation of the Conjoint Results303

Managerial Applications of Conjoint Analysis303

Segmentation304

Profitability Analysis304

Conjoint Simulators305

Alternative Conjoint Methodologies306

Adaptive/Self-Explicated Conjoint:Conjoint with a Large Number of Factors306

Choice-Based Conjoint:Adding Another Touch of Realism308

Overview of the Three Conjoint Methodologies312

An Illustration of Conjoint Analysis312

Stage 1:Objectives of the conjoint Analysis313

Stage 2:Design of the Conjoint Analysis313

Stage 3:Assumptions in Conjoint Analysis316

Stage 4:Estimating the Conjoint Model and Assessing Overall Model Fit316

Stage 5:Interpreting the Results320

Stage 6:Validation of the Results324

A Managerial Application:Use of a Choice Simulator325

Summary327

Questions330

Suggested Readings330

References330

Chapter 7 Multiple Discriminant Analysis and Logistic Regression335

What Are Discriminant Analysis and Logistic Regression?339

Discriminant Analysis340

Logistic Regression341

Analogy with Regression and MANOVA341

Hypothetical Example of Discriminant Analysis342

A Two-Group Discriminant Analysis:Purchasers Versus Nonpurchasers342

A Geometric Representation of the Two-Group Discriminant Function345

A Three-Group Example of Discriminant Analysis:Switching Intentions346

The Decision Process for Discriminant Analysis348

Stage 1:Objectives of Discriminant Analysis350

Stage 2:Research Design for Discriminant Analysis351

Selecting Dependent and Independent Variables351

Sample Size353

Division of the Sample353

Stage 3:Assumptions of Discriminant Analysis354

Impacts on Estimation and Classification354

Impacts on Interpretation355

Stage 4:Estimation of the Discriminant Model and Assessing Overall Fit356

Selecting an Estimation Method356

Statistical Significance358

Assessing Overall Model Fit359

Casewise Diagnostics368

Stage 5:Interpretation of the Results369

Discriminant Weights369

Discriminant Loadings370

Partial F Values370

Interpretation of Two or More Functions370

Which Interpretive Method to Use?373

Stage 6:Validation of the Results373

Validation Procedures373

Profiling Group Differences374

A Two-Group Illustrative Example375

Stage 1:Objectives of Discriminant Analysis375

Stage 2:Research Design for Discriminant Analysis375

Stage 3:Assumptions of Discriminant Analysis376

Stage 4:Estimation of the Discriminant Model and Assessing Overall Fit376

Stage 5:Interpretation of the Results387

Stage 6:Validation of the Results390

A Managerial Overview391

A Three-Group Illustrative Example391

Stage 1:Objectives of Discriminant Analysis391

Stage 2:Research Design for Discriminant Analysis392

Stage 3:Assumptions of Discriminant Analysis392

Stage 4:Estimation of the Discriminant Model and Assessing Overall Fit392

Stage 5:Interpretation of Three-Group Discriminant Analysis Results404

Stage 6:Validation of the Discriminant Results410

A Managerial Overview412

Logistic Regression:Regression with a Binary Dependent Variable413

Representation of the Binary Dependent Variable414

Sample Size415

Estimating the Logistic Regression Model416

Assessing the Goodness-of-Fit of the Estimation Model419

Testing for Significance of the Coefficients421

Interpreting the Coefficients422

Calculating Probabilities for a Specific Value of the Independent Variable425

Overview of Interpreting Coefficients425

Summary425

An Illustrative Example of Logistic Regression426

Stages 1,2,and 3:Research Objectives,Research Design,and Statistical Assumptions426

Stage 4:Estimation of the Logistic Regression Model and Assessing Overall Fit426

Stage 5:Interpretation of the Results432

Stage 6:Validation of the Results433

A Managerial Overview434

Summary434

Questions437

Suggested Readings437

References437

Chapter 8 ANOVA and MANOVA439

MANOVA:Extending Univariate Methods for Assessing Group Differences443

Multivariate Procedures for Assessing Group Differences444

A Hypothetical Illustration of MANOVA447

Analysis Design447

Differences from Discriminant Analysis448

Forming the Variate and Assessing Differences448

A Decision Process for MANOVA449

Stage 1:Objectives of MANOVA450

When Should We Use MANOVA?450

Types of Multivariate Questions Suitable for MANOVA451

Selecting the Dependent Measures452

Stage 2:Issues in the Research Design of MANOVA453

Sample Size Requirements—Overall and by Group453

Factorial Designs—Two or More Treatments453

Using Covariates—ANCOVA and MANCOVA455

MANOVA Counterparts of Other ANOVA Designs457

A Special Case of MANOVA:Repeated Measures457

Stage 3:Assumptions of ANOVA and MANOVA458

Independence458

Equality of Variance-Covariance Matrices459

Normality460

Linearity and Multicollinearity Among the Dependent Variables460

Sensitivity to Outliers460

Stage 4:Estimation of the MANOVA Model and Assessing Overall Fit460

Estimation with the General Linear Model462

Criteria for Significance Testing463

Statistical Power of the Multivariate Tests463

Stage 5:Interpretation of the MANOVA Results468

Evaluating Covariates468

Assessing Effects on the Dependent Variate468

Identifying Differences Between Individual Groups472

Assessing Significance for Individual Dependent Variables474

Stage 6:Validation of the Results475

Summary476

Illustration of a MANOVA Analysis476

Example 1:Difference Between Two Independent Groups477

Stage 1:Objectives of the Analysis478

Stage 2:Research Design of the MANOVA478

Stage 3:Assumptions in MANOVA479

Stage 4:Estimation of the MANOVA Model and Assessing the Overall Fit480

Stage 5:Interpretation of the Results482

Example 2:Difference Between K Independent Groups482

Stage 1:Objectives ofthe MANOVA483

Stage 2:Research Design of MANOVA483

Stage 3:Assumptions in MANOVA484

Stage 4:Estimation of the MANOVA Model and Assessing Overall Fit485

Stage 5:Interpretation of the Results485

Example 3:A Factorial Design for MANOVA with Two Independent Variables488

Stage 1:Objectives of the MANOVA489

Stage 2:Research Design of the MANOVA489

Stage 3:Assumptions in MANOVA491

Stage 4:Estimation of the MANOVA Model and Assessing Overail Fit492

Stage 5:Interpretation of the Results495

Summary496

A Managerial Overview of the Results496

Summary498

Questions500

Suggested Readings500

References500

SECTION Ⅲ Analysis Using Interdependence Techniques503

Chapter 9 Grouping Data with Cluster Analysis505

What Is Cluster Analysis?508

Cluster Analysis as a Multivariate Technique508

Conceptual Development with Cluster Analysis508

Necessity of Conceptual Support in Cluster Analysis509

How Does Cluster Analysis Work?510

A Simple Example510

Objective Versus Subjective Considerations515

Cluster Analysis Decision Process515

Stage 1:Objectives of Cluster Analysis517

Stage 2:Research Design in Cluster Analysis518

Stage 3:Assumptions in Cluster Analysis526

Stage 4:Deriving Clusters and Assessing Overall Fit527

Stage 5:Interpretation of the Clusters538

Stage 6:Validation and Profiling of the Clusters539

An Illustrative Example541

Stage 1:Objectives of the Cluster Analysis541

Stage 2:Research Design of the Cluster Analysis542

Stage 3:Assumptions in Cluster Analysis545

Employing Hierarchical and Nonhierarchical Methods546

Step 1:Hierarchical Cluster Analysis(Stage 4)546

Step 2:Nonhierarchical Cluster Analysis(Stages 4,5,and 6)552

Summary561

Questions563

Suggested Readings563

References563

Chapter 10 MDS and Correspondence Analysis565

What Is Multidimensional Scaling?568

Comparing Objects568

Dimensions:The Basis for Comparison569

A Simplified Look at How MDS Works570

Gathering Similarity Judgments570

Creating a Perceptual Map570

Interpreting the Axes571

Comparing MDS to Other Interdependence Techniques572

Individual as the Unit of Analysis573

Lack of a Variate573

A Decision Framework for Perceptual Mapping573

Stage 1:Objectives of MDS573

Key Decisions in Setting Objectives573

Stage 2:Research Design of MDS578

Selection of Either a Decompositional(Attribute-Free)or Compositional(Attribute-Based)Approach578

Objects:Their Number and Selection580

Nonmetric Versus Metric Methods581

Collection of Similarity or Preference Data581

Stage 3:Assumptions of MDS Analysis584

Stage 4:Deriving the MDS Solution and Assessing Overall Fit584

Determining an Object's Position in the Perceptual Map584

Selecting the Dimensionality of the Perceptual Map586

Incorporating Preferences into MDS587

Stage 5:Interpreting the MDS Results592

Identifying the Dimensions593

Stage 6:Validating the MDS Results594

Issues in Validation594

Approaches to Validation594

Overview of Multidimensional Scaling595

Correspondence Analysis595

Distinguishing Characteristics595

Differences from Other Multivariate Techniques596

A Simple Example of CA596

A Decision Framework for Correspondence Analysis600

Stage 1:Objectives of CA601

Stage 2:Research Design of CA601

Stage 3:Assumptions in CA602

Stage 4:Deriving CA Results and Assessing Overall Fit602

Stage 5:Interpretation of the Results603

Stage 6:Validation of the Results604

Overview of Correspondence Analysis604

Illustrations of MDS and Correspondence Analysis605

Stage 1:Objectives of Perceptual Mapping606

Identifying Objects for Inclusion606

Basing the Analysis on Similarity or Preference Data607

Using a Disaggregate or Aggregate Analysis607

Stage 2:Research Design of the Perceptual Mapping Study607

Selecting Decompositional or Compositional Methods607

Selecting Firms for Analysis608

Nohmetric Versus Metric Methods608

Collecting Data for MDS608

Collecting Data for Correspondence Analysis609

Stage 3:Assumptions in Perceptual Mapping610

Multidimensional Scaling:Stages 4 and 5610

Stage 4:Deriving MDS Results and Assessing Overall Fit610

Stage 5:Interpretation of the Results615

Overview of the Decompositional Results616

Correspondence Analysis:Stages 4 and 5617

Stage 4:Estimating a Correspondence Analysis617

Stage 5:Interpreting CA Results619

Overview of CA621

Stage 6:Validation of the Results622

A Managerial Overview of MDS Results622

Summary623

Questions625

Suggested Readings625

References625

SECTION Ⅳ Structural Equations Modeling627

Chapter 11 SEM:An Introduction629

What Is Structural Equation Modeling?634

Estimation of Multiple Interrelated Dependence Relationships635

Incorporating Latent Variables Not Measured Directly635

Defining a Model637

SEM and Other Multivariate Techniques641

Similarity to Dependence Techniques641

Similarity to Interdependence Techniques641

The Emergence of SEM642

The Role of Theory in Structural Equation Modeling642

Specifying Relationships642

Establishing Causation643

Developing a Modeling Strategy646

A Simple Example of SEM647

The Research Question647

Setting Up the Structural Equation Model for Path Analysis648

The Basics of SEM Estimation and Assessment649

Six Stages in Structural Equation Modeling653

Stage 1:Defining Individual Constructs655

Operationalizing the Construct655

Pretesting655

Stage 2:Developing and Specifying the Measurement Model656

SEM Notation656

Creating the Measurement Model657

Stage 3:Designing a Study to Produce Empirical Results657

Issues in Research Design658

Issues in Model Estimation662

Stage 4:Assessing Measurement Model Validity664

The Basics of Goodness-of-Fit665

Absolute Fit Indices666

Incremental Fit Indices668

Parsimony Fit Indices669

Problems Associated with Using Fit Indices669

Unacceptable Model Specification to Achieve Fit671

Guidelines for Establishing Acceptable and Unacceptable Fit672

Stage 5:Specifying the Structural Model673

Stage 6:Assessing the Structural Model Validity675

Structural Model GOF675

Competitive Fit676

Comparison to the Measurement Model676

Testing Structural Relationships677

Summary678

Questions680

Suggested Readings680

Appendix 11A:Estimating Relationships Using Path Analysis681

Appendix 11B:SEM Abbreviations683

Appendix 11C:Detail on Selected GOF Indices684

References685

Chapter 12 Applications of SEM687

Part 1:Confirmatory Factor Analysis693

CFA and Exploratory Factor Analysis693

A Simple Example of CFA and SEM694

A Visual Diagram694

SEM Stages for Testing Measurement Theory Validation with CFA695

Stage 1:Defining Individual Constructs696

Stage 2:Developing the Overall Measurement Model696

Unidimensionality696

Congeneric Measurement Model698

Items per Construct698

Reflective Versus Formative Constructs701

Stage 3:Designing a Study to Produce Empirical Results702

Measurement Scales in CFA702

SEM and Sampling703

Specifying the Model703

Issues in Identification704

Avoiding Identification Problems704

Problems in Estimation706

Stage 4:Assessing Measurement Model Validity707

Assessing Fit707

Path Estimates707

Construct Validity708

Model Diagnostics711

Summary Example713

CFA Illustration715

Stage 1:Defining Individual Constructs716

Stage 2:Developing the Overall Measurement Model716

Stage 3:Designing a Study to Produce Empirical Results718

Stage 4:Assessing Measurement Model Validity719

HBAT CFA Summary727

Part 2:What Is a Structural Model?727

A Simple Example of a Structural Model728

An Overview of Theory Testing with SEM729

Stages in Testing Structural Theory730

One-Step Versus Two-Step Approaches730

Stage 5:Specifying the Structural Model731

Unit of Analysis731

Model Specification Using a Path Diagram731

Designing the Study735

Stage 6:Assessing the Structural Model Validity737

Understanding Structural Model Fit from CFA Fit737

Examine the Model Diagnostics739

SEM Illustration740

Stage 5:Specifying the Structural Model740

Stage 6:Assessing the Structural Model Validity742

Part 3:Extensions and Applications of SEM749

Reflective Versus Formative Measures749

Reflective Versus Formative Measurement Theory749

Operationalizing a Formative Construct750

Distinguishing Reflective from Formative Constructs751

Which to Use—Reflective or Formative?753

Higher-Order Factor Analysis754

Empirical Concerns754

Theoretical Concerns756

Using Second-Order Measurement Theories756

When to Use Higher-Order Factor Analysis757

Multiple Groups Analysis758

Measurement Model Comparisons758

Structural Model Comparisons763

Measurement Bias764

Model Specification764

Model Interpretation765

Relationship Types:Mediation and Moderation766

Mediation766

Moderation770

Longitudinal Data773

Additional Covariance Sources:Timing773

Using Error Covariances to Represent Added Covariance774

Partial Least Squares775

Characteristics of PLS775

Advantages and Disadvantages of PLS776

Choosing PLS Versus SEM777

Summary778

Questions781

Suggested Readings781

References782

Index785