Does It Pay to Measure Emotional Intelligence During Selection?

Topic: Assessment, Emotional Intelligence, Staffing
Publication: International Journal of Selection and Assessment (MAR 2010)
Article: Emotional intelligence in selection contexts: Measurement method, criterion-related validity, and vulnerability to response distortion
Authors: N.D. Christiansen, J.E. Janovics, and B.P. Siers
Reviewed By: Benjamin Granger

Emotional intelligence (EI) is a hot topic in both the personnel selection literature and the popular business press.  While there are many available measures of EI, approaches to its measurement can be organized into two general categories: (1) self-report questionnaires and (2) performance-based measures.  Self-report EI questionnaires are similar to personality measures in that they treat EI as non-cognitive traits and temperaments. Performance- or ability-based EI measures, on the other hand, treat EI as a largely ability-based trait that reflects how people process information related to their emotions and the emotions of others.

These two general approaches are both intended to measure EI.  In a recent study, Christiansen, Janovics, and Siers (2010) compared two popular self-report measures of EI (TMMS and SREIT) to a performance-based EI measure (MSCEIT). They found evidence that the performance-based and self-report approaches operate quite differently and actually don’t appear to measure the same thing. For instance, the self-report measures were very strongly related to measures of personality, whereas the performance-based measure was not.

Conversely, the performance-based measures of EI were much more strongly related to cognitive ability than were the self-report measures.  Christiansen et al. even note that it may not be appropriate to label self-report measures of EI as measures of “intelligence” at all! Another difference is that supervisory ratings of employee performance were predicted by scores on the performance-based EI measure but not by scores on the self-report measures. Finally, self-report measures of EI tend to be more easily “fakeable” than performance-based measures.  That is, with the self-report measures, it is quite easy for job applicants to choose the “best” answer and endorse it – even if it is not representative of their typical behavior.

Overall, Christiansen et al.’s findings suggest that measures of EI – even performance-based measures – offer very little additional information about job applicants when measures of cognitive ability and personality are already used in the selection process.  Self-report measures of EI are largely redundant when measures of personality are already present, and performance-based EI measures add little predictive power beyond cognitive ability and conscientiousness. This calls into question some claims in the popular press about the power of EI for predicting job performance and career success (e.g., Emotional Intelligence is even more important than cognitive ability).

It may also pose a dilemma for organizations currently using EI for employee selection.

Christiansen, N.D., Janovics, J. E., & Siers, B. P. (2010). Emotional intelligence in selection contexts: measurement method, criterion-related validity, and vulnerability to response distortion. International Journal of Selection and Assessment, 18, 87-101.