RESUMO
Introduction: The Multifactor Leadership Questionnaire (MLQ-5X) has been used frequently to assess leadership in different settings. Despite its popularity, there are many critiques of the MLQ-5X such as its questionable multidimensional structure, lack of connection to the theory and the different factor structures of the measurement tool. The purpose of this study was to re-examine the psychometric properties of the MLQ-5X in the Singapore educational context using two datasets. Methods: A total of 872 teachers (40.1% male and 59.9% female) from 20 secondary schools in Singapore completed two sets of MLQ-5X, one set for their immediate reporting officer and one set for their school leaders. Results: Congeneric Confirmatory Factor Analysis, Rho's coefficients, and AVE were used to analyze MLQ-5X's convergent validity and internal consistency. After five items were deleted, the MLQ-5X showed acceptable internal consistency and convergent validity. Eight measurement models were tested with the original 36 items and the reduced items MLQ-5X. Latent factor correlation matrix with confidence intervals was used to assess the discriminant validity of the MLQ-5X. The results provided support for a nine first-order factors and three second-order factors model (transformation [IIA, IIB, IM, IS, IC, CR], transactional (MBEA), and non-leadership (MBEP and LF). Discussion: The discriminant validity of the hierarchical measurement model of MLQ-5X is supported using dataset 2.
