Tackling missing data in PLS-SEM: Strategies and insights for business research

This study provides a practical guide for handling missing data in partial least squares structural equation modeling (PLS-SEM), a prominent multivariate technique that is widely used in business research. We compare the strengths and limitations of different missing data handling techniques, emphasizing the importance of selecting appropriate methods to enhance the accuracy and reliability of PLS-SEM analyses. Furthermore, we introduce an innovative approach for dealing with not missing at random (NMAR) data by combining imputation with subsequent weighting. By demonstrating the practical effects of various treatment strategies through empirical case studies and a comprehensive simulation study, this research offers meaningful insights and pragmatic guidelines for business researchers dealing with missing data in PLS-SEM.