Abstract
This quantitative cross-sectional study validated the construct validity of the 18-item empowering leadership scale (ELS) in South Africa using principles from both classical test theory and item response theory. 221 employees participated in the research by completing an online Google survey that was hosted on Prolific. Exploratory and confirmatory factor analysis were employed to explore and confirm the two-factor structure of the ELS. Correlation analysis was performed to investigate the convergent validity of the ELS to the psychological empowerment scale. Moreover, Rasch analysis was used to inspect the fit of the ELS items to the Rasch measurement model and to determine the hierarchical ordering of the items. To determine the functionality of the original 7-point response format of the ELS, the category probability curves were inspected. Generally, the results suggest that a two-factor model is superior to a one-factor model; a two-factor modified model with some item revisions yielded the best model fit. The ELS was positively related to psychological empowerment, demonstrating good convergent validity. The ELS conformed to Rasch model expectations, apart from one item that showed misfit. Rating scale analysis provided evidence for collapsing the 7-point scale to a 4-point response format. The ELS displays acceptable construct validity within the South African context although certain provisions should be made prior to its administration.