Emotional intelligence: psychometric status and developmental characteristics--comment on Roberts, Zeidner, and Matthews (2001).

There is now sufficient work in the literature on emotional intelligence to suggest that this construct or series of constructs deserves serious attention, but several questions remain as to adequate construct validation as well as to the emergence and development of these constructs. There is a need to conduct convergent and divergent validity studies on a midlife sample that is likely to show the optimal level of differentiation of the new constructs. The reference domain of cognitive intelligence should be constructed in a multiple-construct manner, and the validation procedure should use confirmatory factor analysis and P. S. Dwyer's (1937) extension method. Once properly validated, there is a need to study the emergence, age differences, and age changes in the level and structure of emotional intelligence. A paradigm that investigates the invariance of factor structure across age and uses the model of differentiation-dedifferentiation would be useful for this purpose.