23.2 Partial Least Squared

The difference between PLS and Principal Components Regression is that Principal Components Regression focuses on variance while reducing dimensionality. PLS focuses on covariance while reducing dimensionality.

Hair Jr. et al. (2017)

  • Partial Least Squared- Structural Equation Modeling vs. Covariance-based Structural Modeling

  • PLS-SEM vs. CB-SEM

CB-SEM PLS-SEM
Base Model Common Factor Model Composite Factor Model


McIntosh, Edwards, and Antonakis (2014)

  • Reflections on Partial least Squares Path Modeling (PLS-PM)

  • There is still a debate to whether

    1. PLS-PM is a SEM method

    2. PLS-PM can reduce the impact of measurement error: yes (increase reliability)

    3. PLS-PM can validate measurement models

    4. PLS-PM provides valid inference on path coefficients

    5. PLS-PM is better than SEM at handling small sample sizes

    6. PLS-PM can be used for exploratory modeling

  • Model fit can be based on

    • global chi-square fit statistic

    • local chi-square fit statistic

    • explained variance

References

Hair Jr., Joe F., Lucy M. Matthews, Ryan L. Matthews, and Marko Sarstedt. 2017. “PLS-SEM or CB-SEM: Updated Guidelines on Which Method to Use.” International Journal of Multivariate Data Analysis 1 (2): 107. https://doi.org/10.1504/ijmda.2017.10008574.
McIntosh, Cameron N., Jeffrey R. Edwards, and John Antonakis. 2014. “Reflections on Partial Least Squares Path Modeling.” Organizational Research Methods 17 (2): 210–51. https://doi.org/10.1177/1094428114529165.