Abstract
CB-SEM multicollinearity occurs when a statistical model’s exogenous constructs are highly correlated. Multicollinearity typically alters regression path coefficients and potentially leads towards higher standardized errors, making standardized path coefficient estimates less precise, less reliable, and so more difficult to precisely interpret. Multicollinearity detrimentally affects overall fit of the CB-SEM model and typically delivers weaker model fit indices. Thus, particularly in complex CBSEM models, hypothesized framework model relationships can become more difficult to interpret, as multicollinearity typically reduces the impact of each relevant measurement item(s) on its attached exogenous construct(s). Overall, recommendations to deal with multicollinearity are provided.
Recommended Citation
Hamilton, John R. and Tee, Singwhat, "AMOS CB-SEM Multicollinearity" (2025). ICEB 2025 Proceedings (Hanoi, Vietnam). 5.
https://aisel.aisnet.org/iceb2025/5