Content validation and item-score reliability of a trait emotional intelligence scale in teacher leadership

Abstract

In a two-stage process, we validated and established the item quality of an initial 15-item trait emotional intelligence (trait EI) scale in teacher leadership in higher education. First, we compute the interrater agreement (κ = 0.87) based on the trait EI domain using data obtained from a panel of nine experts. Second, item-rest correlation and item factor loading for the 15-item trait EI scale were determined based on data collected from a pilot sample of 61 teacher-leaders working with higher educational institutions in Northeast Nigeria. The results of the item-rest correlations (in two rounds) reduced the scale to nine items with excellent psychometrics (Cronbach alpha = 0.928; average interitem correlation = 0.589). This result was confirmed through the item factor loadings (range: 0.626-0.855) computed using JASP software. The scale showed good to reasonable model fit (χ 2 = 127.105; df = 90; χ 2 /df = 1.412; p >0.006; RMSEA = 0.099; and TLI 0.870). Overall, this study produced a valid and reliable 9-item trait EI scale in teacher leadership that can be used in assessing the trait EI of teacher-leaders in higher education. Also, the study contributes to the refinement of the sampling domain of the trait EI construct.