The development of comparative bias index

• Main Authors: Zainudin, Awang, Aimran, A.N, Ahmad, S, Afthanorhan, A
• Format: Conference or Workshop Item
• Language: English
• Published: 2017
• Subjects: HA Statistics; QA Mathematics
• Online Access: http://eprints.unisza.edu.my/1529/1/FH03-FESP-17-09949.jpg; http://eprints.unisza.edu.my/1529/
• Institution: Universiti Sultan Zainal Abidin

Structural Equation Modeling (SEM) is a second generation statistical analysis techniques developed for analyzing the inter-relationships among multiple variables in a model simultaneously. There are two most common used methods in SEM namely Covariance-Based Structural Equation Modeling (CB-SEM) and Partial Least Square Path Modeling (PLS-PM). There have been continuous debates among researchers in the use of PLS-PM over CB-SEM. While there is few studies were conducted to test the performance of CB-SEM and PLS-PM bias in estimating simulation data. This study intends to patch this problem by a) developing the Comparative Bias Index and b) testing the performance of CB-SEM and PLS-PM using developed index. Based on balanced experimental design, two multivariate normal simulation data with of distinct specifications of size 50, 100, 200 and 500 are generated and analyzed using CB-SEM and PLS-PM.