تحليل الوساطة في نمذجة المعادلة البنائية النموذج الواحد والنماذج منفصلة(دراسة نظرية)

Abstract

This study aimed to analyze the concept of mediation within the framework of Structural Equation Modeling (SEM), and to determine whether it is more appropriate to test a mediation model as a single integrated model or through separate models. The study adopted a critical analytical approach, reviewing a comprehensive body of recent literature and research in this field. The findings revealed that employing a single integrated model statistically outperforms separate models in terms of accuracy and inferential power. It reduces the risk of Type I error and allows for simultaneous estimation of all pathways—both direct and indirect—within the mediation model, including indirect effects, thereby enhancing the robustness of research conclusions. Furthermore, the unified model provides overall model fit indices such as GFI, CFI, and RMSEA, which are essential tools for assessing model validity. In addition, the integrated approach demonstrates superior capability in handling complex models involving latent variables or multiple mediation paths—something that traditional methods, such as the Baron and Kenny approach, struggle to address effectively. The study also highlighted the advantages of modern techniques like Bootstrapping and PLS-SEM in improving estimation accuracy, making them more suitable for practical applications. Based on these findings, researchers are recommended to adopt the unified modeling approach when conducting mediation analysis within SEM frameworks, due to its precision, comprehensiveness, and flexibility in dealing with latent variables and complex structural models.