Partial Least Squares Path Modeling: Basic Concepts, Methodological Issues and Applications

Foreword
Preface
Abbreviations
Contents
Editors and Contributors
	About the Editors
	Contributors
Part I Basic Concepts and Extensions
1 Introduction to the Partial Least Squares Path Modeling: Basic Concepts and Recent Methodological Enhancements
	1.1 Introduction
	1.2 Overview of Three Primary SEM Methods
	1.3 Recent Developments in the PLS-PM/PLS-SEM Method
	1.4 Essential Emerging PLS-SEM Tools for Social Sciences Scholars
		1.4.1 Mediation
		1.4.2 Moderation
		1.4.3 Moderated Mediation
		1.4.4 Non-linear SEM Solutions
		1.4.5 Out-of-Sample Prediction
	1.5 Observations and Conclusions
	References
2 Quantile Composite-Based Path Modeling with R: A Hands-on Guide
	2.1 Introduction
	2.2 Quantile Composite-Based Path Modeling in a Nutshell
	2.3 Data Description
	2.4 Running QC–PM with R
		2.4.1 Loading and Pre-processing of the Data
		2.4.2 Model Specification, Estimation, and Results
		2.4.3 Model Assessment and Validation
		2.4.4 Post-processing: Graphs and Result Exporting
	2.5 Concluding Remarks
	References
3 Use of Partial Least Squares Path Modeling Within and Across Business Disciplines
	3.1 Introduction
	3.2 Frequency of PLS-PM Use in Financial Times Journals
	3.3 Rationale for PLS-PM Use in Financial Times Journals
		3.3.1 Problematic Rationale: Small Sample Size
		3.3.2 Problematic Rationale: Data Normality
		3.3.3 Questionable Rationale: Model Complexity
		3.3.4 Appropriate Rationale: Model Assessment
	3.4 The Future of PLS-PM Use in Business Disciplines
	3.5 Conclusions
	References
4 Statistical and Psychometric Properties of Three Weighting Schemes of the PLS-SEM Methodology
	4.1 Introduction
	4.2 Two Distinctive Features of PLS-SEM and the Environmental Variable
	4.3 PLS-SEM Modes A and B
	4.4 PLS-SEM Mode normal upper B Subscript normal upper ABA
	4.5 Scale Invariance and Scale-Inverse Equivariance
		4.5.1 Analytical Results
		4.5.2 Numerical Results
		4.5.3 Sample Results
	4.6 Sensitivity of Weights to Misspecified Models
		4.6.1 PLS-SEM Mode A
		4.6.2 PLS-SEM Mode B
		4.6.3 PLS-SEM Mode normal upper B Subscript normal upper ABA
	4.7 Two Real Data Examples
	4.8 Conclusion and Discussion
	References
5 Software Packages for Partial Least Squares Structural Equation Modeling: An Updated Review
	5.1 Introduction
	5.2 Software for PLS-SEM
		5.2.1 ADANCO
		5.2.2 SmartPLS
		5.2.3 WarpPLS
		5.2.4 XLSTAT-PLSPM
		5.2.5 plssem
		5.2.6 cSEM
		5.2.7 SEMinR
		5.2.8 Summary of Software Features
	5.3 Conclusion
	References
Part II Methodological Issues
6 Revisiting and Extending PLS for Ordinal Measurement and Prediction
	6.1 Introduction
	6.2 Ordinal (Consistent) Partial Least Squares Path Modeling
		6.2.1 Calculating Polychoric/Polyserial Correlations
		6.2.2 Performing the PLS Algorithm
		6.2.3 Correcting for Attenuation if Constructs Are Modeled as Latent Variables
		6.2.4 Estimating Path Coefficients by OLS/2SLS
	6.3 Model-Based Predictions Using PLS and PLSc (PLSpredict and PLScpredict)
	6.4 Model-Based Predictions Using OrdPLS and OrdPLSc
		6.4.1 Relationship Between the Ordinal Indicators and Their Underlying Latent Variables in the Test Dataset
		6.4.2 OrdPLSpredict and OrdPLScpredict
		6.4.3 Evaluating the Predictive Performance of OrdPLSpredict and OrdPLScpredict
	6.5 Monte Carlo Simulation
		6.5.1 Simulation Design
		6.5.2 Simulation Results
		6.5.3 Simulation Insights
	6.6 Guidelines on Performing Predictions Using the R Package cSEM
	6.7 Discussion
	References
7 Multicollinearity: An Overview and Introduction of Ridge PLS-SEM Estimation
	7.1 Introduction
	7.2 Multicollinearity: An Overview
		7.2.1 Multicollinearity Can Stem from Different Sources
		7.2.2 Canonical Structural Multicollinearity
		7.2.3 Numerical Multicollinearity
		7.2.4 Common-Factor Structural Multicollinearity
	7.3 Ridge PLS-SEM
		7.3.1 The Building Blocks of Ridge PLS-SEM
		7.3.2 Ridge PLS-SEM
	7.4 Concluding Remarks
		7.4.1 Conclusion and Contributions
		7.4.2 Limitations and Suggestions for Further Research
	Appendix
		Model
		Regression Coefficients
		Variance Regression Coefficients
	References
8 Demystifying Prediction in Mediation Research and the Use of Specific Indirect Effects and Indirect Effect Sizes
	8.1 Introduction
	8.2 Prediction in Mediation
		8.2.1 Conceptual Time Ordering and Types of Prediction in Mediation Analysis
		8.2.2 Types of Prediction and the PLS-SEM Technique
	8.3 Hypothesizing and Testing Specific Indirect Effects
		8.3.1 Operationalization of the Mediation Model
		8.3.2 Specific Indirect Effects Analysis in SmartPLS
	8.4 Indirect Effect Sizes
		8.4.1 Some Considerations for Effect Sizes in General
		8.4.2 Local Effects Sizes
		8.4.3 Indirect Effect Sizes
		8.4.4 Conducting Indirect Effect Size Analysis in SmartPLS
	8.5 Concluding Remarks and Future Research
	References
9 Alternative Approaches to Higher Order PLS Path Modeling: A Discussion on Methodological Issues and Applications
	9.1 Introduction
	9.2 Reflective and Composite Constructs
	9.3 Relevance of HOC Path Modeling
		9.3.1 HOC PLS-PM Theoretical Framework  and Its Different Types
		9.3.2 Alternative Approaches for Higher-Order PLS-PM Estimation
		9.3.3 Some Guidelines for Choosing the Best HOC Approach
	9.4 Application to Real Data: Comparison of the HOC PLS-PM Approaches
		9.4.1 A Model-Based Multidimensional Poverty Index (MPI)
		9.4.2 Analysis
	9.5 Some Key Findings
	References
10 How to Apply Necessary Condition Analysis in PLS-SEM
	10.1 Introduction
	10.2 Fundamentals of Necessity Logic and NCA
	10.3 Research Application
		10.3.1 The Conceptual Model, Data, PLS-SEM Analyses, and Evaluation (Steps 1–4)
		10.3.2 Extraction of Scores (Step 5)
		10.3.3 Run the Necessary Condition Analysis (Step 6)
		10.3.4 Evaluate the Structural Model Relationships (Step 7)
		10.3.5 Interpret the Findings (Step 8)
	10.4 Discussion and Conclusion
	Appendix
		A1: Data Description
		A2: Results of the Reflective Measurement Models
		A3: Results from PLSpredict
		A4: Results of the Structural Model
		A5: Results of the NCA
		A6: Results Overview: PLS-SEM and NCA
		A7: Guidelines for the Combined Use of PLS-SEM and NCA
	References
Part III Applications
11 New Insights for Public Diplomacy Using PLS-SEM to Analyze the Polyphony of Voices: Value Drivers of the Country Image in Western European and BRICS Countries
	11.1 Introduction
	11.2 Literature Review: The Formation of the Country Image
		11.2.1 A Multidisciplinary Research Field
		11.2.2 Developing a Country Image Measurement Based on Identity, Attitude, and Reputation Theory
		11.2.3 From 4D to 5D Model of Country Image
		11.2.4 The Weakening Influence of the Functional Dimension
		11.2.5 The Role of Stereotypes and Prejudices
		11.2.6 The Role of Proximity
	11.3 Methodology
		11.3.1 Data
		11.3.2 Measures
		11.3.3 Data Analysis
	11.4 Results
		11.4.1 All Countries
		11.4.2 Cluster Analysis
		11.4.3 PLS-SEM Model Evaluation (Per Cluster)
		11.4.4 Comparisons Between BRICS and Western Europe
	11.5 Discussion
	11.6 Conclusion
	Appendix
	References
12 To Survive or not to Survive: Findings from PLS-SEM on the Relationship Between Organizational Resources and Startups’ Survival
	12.1 Introduction
	12.2 Theoretical Framework
	12.3 The Human Capital and Startup Survival
	12.4 The Social Capital and Startup Survival
	12.5 The Entrepreneurial Capital and Startup Survival
	12.6 The Organizational Capital and Startup Survival
	12.7 The Role of Incubation in the Startup Survival
	12.8 The Startup’s Survival
	12.9 Hypothesis and Startup Survival Structural Model
	12.10 Method
		12.10.1 Sample and Data Collection
		12.10.2 Procedure
	12.11 Results
		12.11.1 Evaluation of Measurement Models
		12.11.2 Evaluation of the Structural Model
		12.11.3 Validation of Results with Ord-PLS
		12.11.4 Evaluation of the Importance and Performance of the Variables (IPMA)
	12.12 Discussion
	12.13 Conclusion
	Appendix 12.1: Summary of Theoretical Frameworks
	Appendix 12.2: Variables Linked to the Model
	References
13 Influence of Earnings Quality Dimensions on the Perception of Earnings Quality: An Empirical Application of Composite PLS Using Archival Data
	13.1 Introduction
	13.2 Theoretical Framework
		13.2.1 Definition of Earnings Quality
	13.3 Research Design
		13.3.1 Sample Selection
		13.3.2 Measure and Scale
		13.3.3 Data Analysis
	13.4 Results
		13.4.1 Descriptive Statistics
		13.4.2 Measurement Validity Assessment
		13.4.3 Structural Model Valuation
		13.4.4 Additional Analyses
		13.4.5 Discussion of the Results
	13.5 Conclusions, Limitations, and Future Research Lines
	Appendix 1: Estimation Models
	References
14 Importance-Performance Map Analysis of Capital Structure Using PLS-SEM: Evidence from Non-financial Sector
	14.1 Introduction
	14.2 Theoretical Context and Literature Review
		14.2.1 Firm Dynamics
		14.2.2 Industry Dynamics
		14.2.3 Macroeconomic Dynamics
	14.3 Research Design
		14.3.1 Trial Sample
		14.3.2 Measures
		14.3.3 Data Analysis
	14.4 Results
		14.4.1 Measurement Model Assessment
		14.4.2 Structural Model Assessment
		14.4.3 Testing of Hypotheses
		14.4.4 Importance-Performance Map Analysis
		14.4.5 Discussion of Results
	14.5 Conclusion
		14.5.1 Practical Implications
		14.5.2 Limitation and Future Research Recommendations
	Appendix
	References
Author Index
Subject Index