Building a Scale of Emotional Intelligence for Human Resource Management by using the Graded-Response Model

Abstract

This paper uses the graded-response model to construct the dimension of personal competence in the emotional intelligence scale for human resource management at Sohar University in Oman. It also aims to ensure that unidimensionality and local independence are achieved in the graded-response model for the personal competence dimension in the scale of emotional intelligence for human resource management. It also estimates the parameters of personal competence in the emotional intelligence scale of human resources management, as well as individuals’ parameters (abilities) corresponding to each possible total score on the personal competence in the emotional intelligence scale of human resources management. It also aims to identify items’ and persons’ fit in the emotional intelligence scale (personal competence) for human resource management using the graded-response model. This research aims to reveal the presence of statistically significant differences at the level of significance (α ≤ 0.05) in the level of emotional intelligence (personal competence) for human resources management using the graded-response model according to the variables of gender, job title, years of experience. The researchers adopted the descriptive approach, and data were collected from a stratified random sample of Sohar University employees. The sample consisted of 396 employees. The researchers used an emotional intelligence scale and focused on the personal competence dimension. The Multilog software program was used for statistical analysis. The results indicate that the assumptions of the graded-response model (i.e., one-dimensional and positional independence) are upheld in the present research. The item parameters were easy and medium difficulty, and the arithmetic mean  of  ability was slightly higher than average. The results show that 38 items fit the graded-response model, while 12 did not. The results also show that the model did not fit 24 individuals in the study sample. The results indicate non-significant differences in the employees’ abilities in personal competence according to gender. The results also reveal statistically significant differences in the employees’ abilities in personal competence according to the job in favor of academic employees.