Emotional intelligence: The MSCEIT from the perspective of generalizability theory

The Mayer, Salovey, & Caruso Emotional Intelligence Test (MSCEIT) has been reported to provide reliable scores for the four-branch ability model of emotional intelligence [Mayer, J. D., Salovey, P., & Caruso, D. R. (2002). Mayer–Salovey–Caruso Emotional Intelligence Test (MSCEIT). User's manual. Toronto, Canada: Multi-Health Systems]. However, no studies have yet been reported that have carried out a comprehensive analysis of reliability of scores from MSCEIT, taking into account the different conceptual features of the multifacet measurement design. Results from generalizability analyses of scores from 111 Norwegian executives' responses on the Norwegian version of MSCEIT show that scores reflect considerable amounts of measurement error. Ability scores from Perceiving Emotions are multidimensional, reflecting different types of emotion and the presence or absence of rated emotions in the stimuli. Generalizability (reliability) coefficients for scores from Perceiving Emotions, Facilitating Thought, Understanding Emotion, and Managing Emotions were estimated to .71, .37, .50, and .46, respectively, which is substantially lower than reported in previous studies. The low estimated generalizability coefficients suggest that the scores may not generalize well to intended domains, and the validity of some of the scores may be questioned.