Comparison of the Artificial Neural Network Method and Structural Equation Method for Designing Employees' Justice Perception Model

Comparing the validity and the fitness power of two methods of structural equations (SEM), and artificial neural network (ANN) in the field of human resource studies represents the main goal of this article. A conceptual model was examined by both SEM and ANN, including all functions of HRM, individual culture, organizational culture and characteristics of accountability system as independent variables and the employees' justice perception as dependent variable. This survey research was implemented within three Iranian Banks (Mellat, Tejarat and keshvarzi) by random sampling of 325 employees. The researchers used the R² value as the metric of comparison that comes up with these findings: 1- There was not significant difference between the performance of SEM and ANN when we had a few number of independent variables. When the number of independent variables increased, we found strong support for ANN having better result than SEM, with regards to R². The outcomes of this research were partially similar to the same studies and provide useful insight into capabilities of ANN and SEM used in HRM researches.