Convergent and divergent validity of integrative versus mixed model measures of emotional intelligence

The construct of emotional intelligence (EI) has garnered increased attention in the popular media and scientific literature. Several competing measures of EI have been developed, including self-report and performance-based instruments. The current study replicates and expands on previous research by examining three competing EI measures (Mayer–Salovey–Caruso Emotional Intelligence Test, MSCEIT; Bar-On Emotion Quotient Inventory, EQ-i; and Self-Rated Emotional Intelligence Scale, SREIS) and their relationships with cognitive functioning (Wechsler Abbreviated Scale of Intelligence; WASI), Big Five personality traits (NEO-PI-R) and emotional well-being (Beck Depression Inventory, BDI and Positive and Negative Affect Schedule, PANAS). Results indicated that significant variability in the self-report EI measures was accounted for by personality and emotional well-being measures, whereas the MSCEIT was more strongly associated with IQ. Overall, nearly two-thirds (62%) of the variance in EQ-i scores was accounted for by Big Five personality traits, emotional well-being and full scale IQ; whereas only 14% of the variance in MSCEIT scores was accounted for by these same variables. The present findings raise questions regarding the divergent validity of self-report EI measures from existing personality and emotional well-being measures. The implication of these results and directions for future research are discussed.

► The validity of competing emotional intelligence (EI) measures is examined.
► Self-report EI shared substantial variance with personality and emotion constructs.
► The performance-based MSCEIT was discriminable from personality/emotion variables.
► Findings question the discriminant validity of commonly used EI measures.