Partial Least Squares

A 2-Day Seminar on Partial Least Squares Structural Equation Modeling 
Taught by Joe Hair, Ph.D. and Marko Sarstedt, Ph.D.

Structural equation modeling (SEM) extends traditional analysis methods by facilitating the estimation and explanation of relationships among both observed variables and latent variables (constructs). SEM has traditionally been done by maximum likelihood analysis of covariance structures. But there is an alternative approach, known as partial least squares SEM (PLS-SEM) that has several attractions in many situations commonly encountered in empirical research.

For example, PLS-SEM performs well when sample sizes are small, when the data are not normally distributed, when “formative” measures are used, or when complex models with many indicators and relationships are estimated. PLS is also particularly well suited to applied settings where the emphasis is on predictive modeling.

This seminar will provide an in-depth introduction to the PLS-SEM approach, including the nature of causal modeling, analytical objectives, the evaluation of results, and an introduction to complementary analytical techniques.

In addition to the fundamentals of PLS-SEM, the seminar will provide a comprehensive introduction to the freeware package SmartPLS. With its user-friendly interface and extraordinary analytical and graphical abilities, SmartPLS is the preferred PLS-SEM software for data analysts in many disciplines. 

SmartPLS empowers users to engage data in a hands-on and interactive manner, without the need to devote extensive time and effort to programming. There is also a global online network of more than 35,000 SmartPLS users providing a rich community of support.

Participants will develop facility with SmartPLS data manipulation commands, its wide array of tools for exploring data, and the implementation of frequently used statistical tests and models in the PLS-SEM environment. They will also learn how to apply and interpret numerical and graphical output, and how to incorporate these results directly into research reports. Upon completion of the seminar, participants will have gained the ability to confidently use SmartPLS to examine applied and theoretical models using PLS-SEM.


This is a hands-on course that includes instruction and practice exercises using SmartPLS. To optimally benefit, you should bring your own laptop computer with SmartPLS already installed. You can download the software at Power outlets will be provided at each seat. 

Who should attend? 

This seminar will benefit researchers, analysts, and students interested in identifying and mastering a statistical software package for analysis of complex data. Data analysts interested in extending their toolbox of statistical methods and software to include PLS-SEM and SmartPLS software will find it particularly useful. The seminar is designed to appeal to those who seek an accelerated, in-depth introduction to an emerging multivariate data analysis method as well as a powerful, popular, and flexible statistical software package.

The course will focus on understanding PLS-SEM, when and why it should be used, and how it should be interpreted. Familiarity with basic statistical concepts including topics such as basic probability distributions, hypothesis testing, and ordinary least squares regression will be assumed. While prior exposure to statistical software may be useful, no prior experience with SmartPLS or any other statistical package is required.


Participants receive a bound manual containing detailed lecture notes, examples of computer printout, and many other useful features. This book frees participants from the distracting task of note taking.

Registration and lodging

The fee of $895.00 includes all seminar materials. 

Lodging Reservation Instructions

A block of rooms has been reserved at the Boston Common Hotel and Conference Center, 40 Trinity Place, Boston MA at a rate of $169 per night. To register, please call (617) 933-7700 and identify yourself with Statistical Horizons. For guaranteed rate and availability, you must make your reservation before the cut-off date of May 19, 2013. 

Seminar outline

First Day:
Structural Equation Modeling Overview

  • What is SEM
  • CB-SEM vs. PLS-SEM
  • Specifying Path Models
  • Types of Measurement Models
  • SmartPLS Software
  • Simple Illustration – Reputation Model 

Path Model Estimation

  • PLS Algorithm
  • Statistical Properties
  • Bootstrapping
  • SmartPLS Software
  • Simple Illustration – Reputation Model

Evaluation of Reflective Measurement Models

  • Convergent Validity
  • Discriminant Validity
  • Internal Consistency Reliability
  • Hands-On Example 

Evaluation of Formative Measurement Models

  • Convergent Validity
  • Collinearity Issues
  • Significance & Relevance
  • Hands-On Example

Second Day:
Evaluating the Structural Model

  • Collinearity Assessment
  • Path Coefficients
  • Coefficient of Determination (R2)
  • Effect Size f 2
  • Predictive Relevance/Blindfolding (Q2 & q2)
  • Advanced Illustration – Reputation Model

Higher-Order Models

Mediation Analysis

Moderation Analysis 

Hands-On Exercise – Applied Example