Abstract
Twenty-six complex CB-SEM models are generated against one quantitative management consulting Likert scale dataset. Two complex alternate models contrast exogenous effect latent variable (LV) approaches. Model interaction effects successively reduce from structural (full parameter or optimal parameter or minimal three factor measurement item LV models) to path hybrid models to single indicator latent composite (SILC) path models. Subject to many conditions, SILC models can occasionally be used in place of full parameter model counterparts models to expose paths otherwise hidden by complex CB-SEM structural model interaction effects, but this approach should be discussed with caution against path and measurement considerations possibilities. This article expands understanding of when researchers may use choose either an appropriate path or a structural CB-SEM approach – particularly under complex multilevel CB-SEM modelling situations.
Recommended Citation
Hamilton, John R. and Sukresna, IMade, "Complex Structural Equation Modelling: Comparing Single Indicator Latent Composite Variants to Optimal, or Full-Parameter Models" (2026). ICEB 2026 Proceedings (Beijing, China). 24.
https://aisel.aisnet.org/iceb2026/24