Track: Data Analytics

Abstract

 This research explores the utilization of Partial Least Squares Structural Equation Modeling (PLS-SEM) in empirical archival research. PLS-SEM is a variant of SEM that has gained attention as an under-utilized method, particularly in the context of big data and secondary data analysis. The study highlights key considerations and implications when employing PLS-SEM in archival research. It emphasizes the importance of justifying the choice of PLS-SEM based on data characteristics and research goals, particularly when dealing with non-normal data distributions and limited theoretical foundations. The article provides practical guidance on model specification, data preparation, estimation, model evaluation, and reporting in PLS-SEM analysis. Additionally, the paper discusses the significance of reporting model fit indices, variable selection, and sample size estimation. The conclusion highlights the strengths and future research directions of PLS-SEM in archival research, including methodological advancements, integration with other statistical techniques, comparative studies, application in diverse fields, and the establishment of reporting guidelines. Overall, this research aims to enhance the understanding and application of PLS-SEM in the analysis of archival data, facilitating its wider adoption and contributing to empirical knowledge in various domains.