Structural Equation Modeling (SEM) Using SmartPLS Software: Analyzing Path Models Using Partial Least Squares (PLS) Based SEM

摘要

This 'hands-on' workshop will demonstrate how to obtain and use sophisticated, but (currently) free, visual software for assessing the measurement and structural model characteristics of path-based SEM quantitative models using a partial least squares (PLS) approach. The workshop will be presented as a teaching aid for faculty, doctoral students, and for other researchers who are engaged in SEM modeling. The workshop will demonstrate how to use the features of SmartPLS software. SmartPLS, developed at the University of Hamburg (Germany), is (currently) a freeware application for (graphical) path modeling with latent variables using the PLS approach. Currently, there are more than 10,000 registered SmartPLS users worldwide. As opposed to covariance-based SEM approaches (exemplified by software such as LISREL or EQS), the PLS method has been documented as possibly more appropriate for a majority of SEM studies and data sets typically used by researchers. PLS has different requirements for sample size, missing data and data distribution characteristics (e.g. normality, skewness, kurtosis, etc.). Numerous research articles reporting PLS-based SEM analyses have been published in leading MIS journals, including MIS Quarterly, Information Systems Research, and Journal of Management Information Systems, among others.