CB-SEM vs PLS-SEM Method for Traffic Characteristic at Road Intersection in Mixed Traffic Condition

Abstract

Fajaruddin Mustakim, Azlan bin Abdul Aziz, Lim Heng Siong, Yaser Bakhuraisa, Othman Che Puan and Saifulnizan Jamian

The aim of this study is to compare recently used methods of Structural Equation Modelling (SEM) which consists Covariance Based Structural Equation Modeling (CB-SEM) and Partial Least Squares based Structural Equation Modelling (PLS-SEM). The concept behind the first approach method is based on covariance, meanwhile the second approach is based on variance (partial least squares). First, the study develops CB-SEM by using eight hundred forty- one data set from right turn motor vehicle (RMV) in mixed traffic road network. Next, it further analyses the difference between PLS and Consistent PLS algorithms and finds the same result in both methods. Consequently, structural model is tested using CB-SEM and PLS-SEM. The outcomes indicate that the item loading is typically higher in CB-SEM than PLS-SEM. Model fit indices in CB-SEM have a better measurement, whereas PLS-SEM fit indices are still applying and less than CB-SEM. In PLS-SEM contribute excellent combine-based models, whereas CB-SEM are better for factor-based models. The comparison between both methods by providing visualization mapping diagram with numerical and empirical may contribute to existing literature as well as for predictive research domain.

PDF