Title Details

Silvia Salini, Department of Economics, Business and Statistics ,University of Milan, Italy

Preface xix

Contributors xxiii

PART I BASIC ASPECTS OF CUSTOMER SATISFACTION

SURVEY DATA ANALYSIS

1 Standards and classical techniques in data analysis of customer satisfaction surveys 3
Silvia Salini and Ron S. Kenett

1.1 Literature on customer satisfaction surveys 4

1.2 Customer satisfaction surveys and the business cycle 4

1.3 Standards used in the analysis of survey data 7

1.4 Measures and models of customer satisfaction 12

1.4.1 The conceptual construct 12

1.4.2 The measurement process 13

1.5 Organization of the book 15

1.6 Summary 17

References 17

2 The ABC annual customer satisfaction survey 19
Ron S. Kenett and Silvia Salini

2.1 The ABC company 19

2.2 ABC 2010 ACSS: Demographics of respondents 20

2.3 ABC 2010 ACSS: Overall satisfaction 22

2.4 ABC 2010 ACSS: Analysis of topics 24

2.5 ABC 2010 ACSS: Strengths and weaknesses and decision drivers 27

2.6 Summary 28

References 28

Appendix 29

3 Census and sample surveys 37
Giovanna Nicolini and Luciana Dalla Valle

3.1 Introduction 37

3.2 Types of surveys 39

3.2.1 Census and sample surveys 39

3.2.2 Sampling design 40

3.2.3 Managing a survey 40

3.2.4 Frequency of surveys 41

3.3 Non-sampling errors 41

3.3.1 Measurement error 42

3.3.2 Coverage error 42

3.3.3 Unit non-response and non-self-selection errors 43

3.3.4 Item non-response and non-self-selection error 44

3.4 Data collection methods 44

3.5 Methods to correct non-sampling errors 46

3.5.1 Methods to correct unit non-response errors 46

3.5.2 Methods to correct item non-response 49

3.6 Summary 51

References 52

4 Measurement scales 55
Andrea Bonanomi and Gabriele Cantaluppi

4.1 Scale construction 55

4.1.1 Nominal scale 56

4.1.2 Ordinal scale 57

4.1.3 Interval scale 58

4.1.4 Ratio scale 59

4.2 Scale transformations 60

4.2.1 Scale transformations referred to single items 61

4.2.2 Scale transformations to obtain scores on a unique interval scale 66

Acknowledgements 69

References 69

5 Integrated analysis 71
Silvia Biffignandi

5.1 Introduction 71

5.2 Information sources and related problems 73

5.2.1 Types of data sources 73

5.2.2 Advantages of using secondary source data 73

5.2.3 Problems with secondary source data 74

5.2.4 Internal sources of secondary information 75

5.3 Root cause analysis 78

5.3.1 General concepts 78

5.3.2 Methods and tools in RCA 81

5.3.3 Root cause analysis and customer satisfaction 85

5.4 Summary 87

Acknowledgement 87

References 87

6 Web surveys 89
Roberto Furlan and Diego Martone

6.1 Introduction 89

6.2 Main types of web surveys 90

6.3 Economic benefits of web survey research 91

6.3.1 Fixed and variable costs 92

6.4 Non-economic benefits of web survey research 94

6.5 Main drawbacks of web survey research 96

6.6 Web surveys for customer and employee satisfaction projects 100

6.7 Summary 102

References 102

7 The concept and assessment of customer satisfaction 107
Irena Ograjenˇsek and Iddo Gal

7.1 Introduction 107

7.2 The quality–satisfaction–loyalty chain 108

7.2.1 Rationale 108

7.2.2 Definitions of customer satisfaction 108

7.2.3 From general conceptions to a measurement model of customer satisfaction 110

7.2.4 Going beyond SERVQUAL: Other dimensions of relevance to the B2B context 112

7.2.5 From customer satisfaction to customer loyalty 113

7.3 Customer satisfaction assessment: Some methodological considerations 115

7.3.1 Rationale 115

7.3.2 Think big: An assessment programme 115

7.3.3 Back to basics: Questionnaire design 116

7.3.4 Impact of questionnaire design on interpretation 118

7.3.5 Additional concerns in the B2B setting 119

7.4 The ABC ACSS questionnaire: An evaluation 119

7.4.1 Rationale 119

7.4.2 Conceptual issues 119

7.4.3 Methodological issues 120

7.4.4 Overall ABC ACSS questionnaire asssessment 121

7.5 Summary 121

References 122

Appendix 126

8 Missing data and imputation methods 129
Alessandra Mattei, Fabrizia Mealli and Donald B. Rubin

8.1 Introduction 129

8.2 Missing-data patterns and missing-data mechanisms 131

8.2.1 Missing-data patterns 131

8.2.2 Missing-data mechanisms and ignorability 132

8.3 Simple approaches to the missing-data problem 134

8.3.1 Complete-case analysis 134

8.3.2 Available-case analysis 135

8.3.3 Weighting adjustment for unit nonresponse 135

8.4 Single imputation 136

8.5 Multiple imputation 138

8.5.1 Multiple-imputation inference for a scalar estimand 138

8.5.2 Proper multiple imputation 139

8.5.3 Appropriately drawing imputations with monotone missing-data patterns 140

8.5.4 Appropriately drawing imputations with nonmonotone missing-data patterns 141

8.5.5 Multiple imputation in practice 142

8.5.6 Software for multiple imputation 143

8.6 Model-based approaches to the analysis of missing data 144

8.7 Addressing missing data in the ABC annual customer satisfaction survey: An example 145

8.8 Summary 149

Acknowledgements 150

References 150

9 Outliers and robustness for ordinal data 155
Marco Riani, Francesca Torti and Sergio Zani

9.1 An overview of outlier detection methods 155

9.2 An example of masking 157

9.3 Detection of outliers in ordinal variables 159

9.4 Detection of bivariate ordinal outliers 160

9.5 Detection of multivariate outliers in ordinal regression 161

9.5.1 Theory 161

9.5.2 Results from the application 163

9.6 Summary 168

References 168

PART II MODERN TECHNIQUES IN CUSTOMER SATISFACTION SURVEY DATA ANALYSIS

10 Statistical inference for causal effects 173
Fabrizia Mealli, Barbara Pacini and Donald B. Rubin

10.1 Introduction to the potential outcome approach to causal inference 173

10.1.1 Causal inference primitives: Units, treatments, and potential outcomes 175

10.1.2 Learning about causal effects: Multiple units and the stable unit treatment value assumption 176

10.1.3 Defining causal estimands 177

10.2 Assignment mechanisms 179

10.2.1 The criticality of the assignment mechanism 179

10.2.2 Unconfounded and strongly ignorable assignment mechanisms 180

10.2.3 Confounded and ignorable assignment mechanisms 181

10.2.4 Randomized and observational studies 181

10.3 Inference in classical randomized experiments 182

10.3.1 Fisher’s approach and extensions 183

10.3.2 Neyman’s approach to randomization-based inference 183

10.3.3 Covariates, regression models, and Bayesian model-based inference 184

10.4 Inference in observational studies 185

10.4.1 Inference in regular designs 186

10.4.2 Designing observational studies: The role of the propensity score 186

10.4.3 Estimation methods 188

10.4.4 Inference in irregular designs 188

10.4.5 Sensitivity and bounds 189

10.4.6 Broken randomized experiments as templates for the analysis of some irregular designs 189

References 190

11 Bayesian networks applied to customer surveys 193
Ron S. Kenett, Giovanni Perruca and Silvia Salini

11.1 Introduction to Bayesian networks 193

11.2 The Bayesian network model in practice 197

11.2.1 Bayesian network analysis of the ABC 2010 ACSS 197

11.2.2 Transport data analysis 201

11.2.3 R packages and other software programs used for studying BNs 210

11.3 Prediction and explanation 211

11.4 Summary 213

References 213

12 Log-linear model methods 217
Stephen E. Fienberg and Daniel Manrique-Vallier

12.1 Introduction 217

12.2 Overview of log-linear models and methods 218

12.2.1 Two-way tables 218

12.2.2 Hierarchical log-linear models 220

12.2.3 Model search and selection 222

12.2.4 Sparseness in contingency tables and its implications 223

12.2.5 Computer programs for log-linear model analysis 223

12.3 Application to ABC survey data 224

12.4 Summary 227

References 228

13 CUB models: Statistical methods and empirical evidence 231
Maria Iannario and Domenico Piccolo

13.1 Introduction 231

13.2 Logical foundations and psychological motivations 233

13.3 A class of models for ordinal data 233

13.4 Main inferential issues 236

13.5 Specification of CUB models with subjects’ covariates 238

13.6 Interpreting the role of covariates 240

13.7 A more general sampling framework 241

13.7.1 Objects’ covariates 241

13.7.2 Contextual covariates 243

13.8 Applications of CUB models 244

13.8.1 Models for the ABC annual customer satisfaction survey 245

13.8.2 Students’ satisfaction with a university orientation service 246

13.9 Further generalizations 248

13.10 Concluding remarks 251

Acknowledgements 251

References 251

Appendix 255

A program in R for CUB models 255

A.1 Main structure of the program 255

A.2 Inference on CUB models 255

A.3 Output of CUB models estimation program 256

A.4 Visualization of several CUB models in the parameter space 257

A.5 Inference on CUB models in a multi-object framework 257

A.6 Advanced software support for CUB models 258

14 The Rasch model 259
Francesca De Battisti, Giovanna Nicolini and Silvia Salini

14.1 An overview of the Rasch model 259

14.1.1 The origins and the properties of the model 259

14.1.2 Rasch model for hierarchical and longitudinal data 263

14.1.3 Rasch model applications in customer satisfaction surveys 265

14.2 The Rasch model in practice 267

14.2.1 Single model 267

14.2.2 Overall model 268

14.2.3 Dimension model 272

14.3 Rasch model software 277

14.4 Summary 278

References 279

15 Tree-based methods and decision trees 283
Giuliano Galimberti and Gabriele Soffritti

15.1 An overview of tree-based methods and decision trees 283

15.1.1 The origins of tree-based methods 283

15.1.2 Tree graphs, tree-based methods and decision trees 284

15.1.3 CART 287

15.1.4 CHAID 293

15.1.5 PARTY 295

15.1.6 A comparison of CART, CHAID and PARTY 297

15.1.7 Missing values 297

15.1.8 Tree-based methods for applications in customer satisfaction surveys 298

15.2 Tree-based methods and decision trees in practice 300

15.2.1 ABC ACSS data analysis with tree-based methods 300

15.2.2 Packages and software implementing tree-based methods 303

15.3 Further developments 304

References 304

16 PLS models 309
Giuseppe Boari and Gabriele Cantaluppi

16.1 Introduction 309

16.2 The general formulation of a structural equation model 310

16.2.1 The inner model 310

16.2.2 The outer model 312

16.3 The PLS algorithm 313

16.4 Statistical interpretation of PLS 319

16.5 Geometrical interpretation of PLS 320

16.6 Comparison of the properties of PLS and LISREL procedures 321

16.7 Available software for PLS estimation 323

16.8 Application to real data: Customer satisfaction analysis 323

References 329

17 Nonlinear principal component analysis 333
Pier Alda Ferrari and Alessandro Barbiero

17.1 Introduction 333

17.2 Homogeneity analysis and nonlinear principal component analysis 334

17.2.1 Homogeneity analysis 334

17.2.2 Nonlinear principal component analysis 336

17.3 Analysis of customer satisfaction 338

17.3.1 The setting up of indicator 338

17.3.2 Additional analysis 340

17.4 Dealing with missing data 340

17.5 Nonlinear principal component analysis versus two competitors 343

17.6 Application to the ABC ACSS data 344

17.6.1 Data preparation 344

17.6.2 The homals package 345

17.6.3 Analysis on the ‘complete subset’ 346

17.6.4 Comparison of NLPCA with PCA and Rasch analysis 350

17.6.5 Analysis of ‘entire data set’ for the comparison of missing data treatments 352

17.7 Summary 355

References 355

18 Multidimensional scaling 357
Nadia Solaro

18.1 An overview of multidimensional scaling techniques 357

18.1.1 The origins of MDS models 358

18.1.2 MDS input data 359

18.1.3 MDS models 362

18.1.4 Assessing the goodness of MDS solutions 369

18.1.5 Comparing two MDS solutions: Procrustes analysis 371

18.1.6 Robustness issues in the MDS framework 371

18.1.7 Handling missing values in MDS framework 373

18.1.8 MDS applications in customer satisfaction surveys 373

18.2 Multidimensional scaling in practice 374

18.2.1 Data sets analysed 375

18.2.2 MDS analyses of overall satisfaction with a set of ABC features: The complete data set 375

18.2.3 Weighting objects or items 381

18.2.4 Robustness analysis with the forward search 382

18.2.5 MDS analyses of overall satisfaction with a set of ABC

features: The incomplete data set 383

18.2.6 Package and software for MDS methods 384

18.3 Multidimensional scaling in a future perspective 386

18.4 Summary 386

References 387

19 Multilevel models for ordinal data 391
Leonardo Grilli and Carla Rampichini

19.1 Ordinal variables 391

19.2 Standard models for ordinal data 393

19.2.1 Cumulative models 394

19.2.2 Other models 395

19.3 Multilevel models for ordinal data 395

19.3.1 Representation as an underlying linear model with thresholds 396

19.3.2 Marginal versus conditional effects 397

19.3.3 Summarizing the cluster-level unobserved heterogeneity 397

19.3.4 Consequences of adding a covariate 398

19.3.5 Predicted probabilities 399

19.3.6 Cluster-level covariates and contextual effects 399

19.3.7 Estimation of model parameters 400

19.3.8 Inference on model parameters 401

19.3.9 Prediction of random effects 402

19.3.10 Software 403

19.4 Multilevel models for ordinal data in practice: An application to student ratings 404

References 408

20 Quality standards and control charts applied to customer surveys 413
Ron S. Kenett, Laura Deldossi and Diego Zappa

20.1 Quality standards and customer satisfaction 413

20.2 ISO 10004 guidelines for monitoring and measuring customer satisfaction 414

20.3 Control Charts and ISO 7870 417

20.4 Control charts and customer surveys: Standard assumptions 420

20.4.1 Introduction 420

20.4.2 Standard control charts 420

20.5 Control charts and customer surveys: Non-standard methods 426

20.5.1 Weights on counts: Another application of the c chart 426

20.5.2 The χ2 chart 427

20.5.3 Sequential probability ratio tests 428

20.5.4 Control chart over items: A non-standard application of SPC methods 429

20.5.5 Bayesian control chart for attributes: A modern application of SPC methods 432

20.5.6 Control chart for correlated Poisson counts: When things become fairly complicated 433

20.6 The M-test for assessing sample representation 433

20.7 Summary 435

References 436

21 Fuzzy Methods and Satisfaction Indices 439
Sergio Zani, Maria Adele Milioli and Isabella Morlini

21.1 Introduction 439

21.2 Basic definitions and operations 440

21.3 Fuzzy numbers 441

21.4 A criterion for fuzzy transformation of variables 443

21.5 Aggregation and weighting of variables 445

21.6 Application to the ABC customer satisfaction survey data 446

21.6.1 The input matrices 446

21.6.2 Main results 448

21.7 Summary 453

References 455

Appendix An introduction to R 457
Stefano Maria Iacus

A.1 Introduction 457

A.2 How to obtain R 457

A.3 Type rather than ‘point and click’ 458

A.3.1 The workspace 458

A.3.2 Graphics 458

A.3.3 Getting help 459

A.3.4 Installing packages 459

A.4 Objects 460

A.4.1 Assignments 460

A.4.2 Basic object types 462

A.4.3 Accessing objects and subsetting 466

A.4.4 Coercion between data types 469

A.5 S4 objects 470

A.6 Functions 472

A.7 Vectorization 473

A.8 Importing data from different sources 475

A.9 Interacting with databases 476

A.10 Simple graphics manipulation 477

A.11 Basic analysis of the ABC data 481

A.12 About this document 496

A.13 Bibliographical notes 496

References 496

Index 499