Trainings & Webinars

Attend the workshop on PLS Path Modeling taking place prior to the PLS17 conference in Macau, China. Special price for academics: $495.

2-Day Workshop on PLS Path Modeling

After a return to the origins of structural equation modeling (SEM), the PLSPM algorithm will be presented. A methodology to interpret results will then be suggested on the basis of real life cases. The training session will be illustrated by applications using XLSTAT.

DAY 1: Algorithm, Estimation and Practice

• Introduction to Structural Equation Modeling
  • Covariance-based and Component-based apporaches
• PLS Path Modeling Algorithm
  • PLS Path Modeling algorithm for model estimation:
    • Measurement (Outer) Model Specification and Estimation Modes (Reflective vs. Formative – Mode A and Mode B)
    • Structural (Inner) Model Specification and Estimation Schemes (Centroid – Factorial – Path Weighting Schemes)
    • PLS algorithm for computing Latent Variable Scores
  • PLS Algorithm for the case of one and two blocks:
    • Principal Component Analysis, Tucker’s Inter-battery Analysis, Canonical Correlation Analysis, PLS Regression, Redundancy Analysis
  • Hierarchical PLS-PM and the super-block option

• Introductory Tutorial on XLSTAT-PLSPM with Case Studies
  • Model Specification: exploring graphical interface features
  • Model Estimation: measurement and structural options
  • Scaling Latent Variable Scores: standardized vs. normalized
  • Output Retrieval (graphical and tabular) and Interpretation:
    • Outer weights, normalized weights, standardized loadings
    • Path coefficients (direct, indirect and total effects), R², standardized path coefficients, contribution to R², simple and partial correlations
  • Latent variables scores (casewise values, summary statistics)

DAY 2: Model Assessment, Improvement and Advances in PLS-PM

• Model Assessment and Improvement: Diagnostics and Solutions
  • Convergent validity: composite reliability, eigenvalues, condition number, critical value, weights and loadings, average variance extracted (AVE), communality
  • Discriminant validity: cross-loadings vs. loadings, latent variables correlations vs. AVE
  • Predictive relevance: Redundancy, R², Absolute and Relative Goodness of Fit (GoF), Effect Size f²
  • Statistical significance:  Bootstrapping, Jackknifing, t-test, F-test, critical ratios
  • Cross-validation: Blindfolding, CV-Communality, CV-Redundancy
  • Handling Missing Data:  Lohmöller’s option, Impact on Latent Variable Scores

• Continuous Moderating Effects
  • Why & How to Investigate Moderating Effects?
  • Discrete (categorical) vs. continuous moderator variable
  • Methods for Assessing Interaction Effects: Product-Indicator, Two-Stage, Hybrid, Orthogonalizing
  • Interaction with Formative Indicators
  • Centering or Standardizing the Indicators
  • Choosing the appropriate method
  • Additional Methods for Non Linear Relations: Measurement and Structural Level

• Discrete Moderating Effects: Multi-Group Comparison
  • Bootstrap parametric approaches: t-test, empirical confidence intervals
  • Permutation-based comparisons

• Mediating Effects
  • Mediator vs. Confounder
  • Causal Steps for Testing Mediation
  • Methods for Assessing Mediating Effects: Sobel, resampling
  • Mediator versus Moderator
  • Moderated Mediation

• Handling Multidimensionality
  • Detection of Block Multidimensionality
  • Mode PLS for the Measurement Model: a continuum from Mode A to Mode B
  • PLS Regression to cope with multicollinearity in the Structural Model

• Uncovering Segments
  • Definition of Unobserved Heterogeneity
  • REBUS-PLSPM

Vincenzo Esposito Vinzi

Vincenzo ESPOSITO VINZI has a Ph.D. in Computational Statistics and is now Full Professor of Statistics at ESSEC Business School of Paris. Vincenzo is also the President-Elect of the International Society for Business and Industrial Statistics and the Chairman-Elect of the European Board of Directors of the International Association for Statistical Computing. His research includes among others multivariate statistics, structural equation modelling, PLS regression and path modelling, with business oriented applications. Vincenzo has delivered many invited lectures and organized sessions on PLS and related methods during international events. He has also been chairing International conferences and co-editing several conference proceedings and special issues of international journals on PLS methods. Vincenzo is an Associate Editor of Computational Statistics and Data Analysis, Advances in Data Analysis and Classification, Statistical Methods and Applications and Computational Statistics. He is the Editor-in-Chief of the "Handbook of Partial Least Squares: Concepts, Methods and Applications" published in 2010.

Wynne Chin

Wynne W. CHIN is Professor in the department of Decision and Information Sciences in the C.T. Bauer College of Business at the University of Houston. He received his A.B. in Biophysics from U.C. Berkeley, MS in Biomedical/Chemical Engineering from Northwestern University, and an MBA and Ph.D. in Computers and Information Systems from the University of Michigan. Wynne has also taught previously at the University of Calgary, Wayne State University, and the University of Michigan and has been a visiting fellow at the University of Canterbury, Queens University, City University of Hong Kong, and the University of New South Wales. Wynne's research includes sales force automation, IT adoption, outsourcing, acceptance, satisfaction, group cohesion and negotiation, and psychometric modeling issues. Wynne is on the editorial board of Structural Equation Modeling journal, Journal of AIS, Journal of Information Technology, IEEE Transaction of Management. He is also the developer of PLS-Graph, the first graphical based software dating back to 1990 to perform Partial Least Squares analysis.

The quality certification has been delivered for the following category of action:

Training actions