“Emotional Intelligence and Transformational and Transactional Leadership: A Meta-Analysis”, by P.D. Harms of the University of Nebraska, Lincoln, Nebraska, and Marcus Credé of State University of New York at Albany, was the most frequently read article in the Journal of Leadership and Organizational Studies in 2010. P.D. Harms has provided a brief perspective on the article.
Originally, we had no intention of doing research on emotional intelligence (EI) and leadership. I (Harms) was aware that there was a great deal of scepticism surrounding the validity of EI measures, but was not that interested in it myself. However, as I was doing a literature review of the antecedents of transformational leadership, I couldn’t help but notice that every few months someone would publish an article or write a dissertation using EI as a predictor of leadership and stating that there was a great deal of controversy regarding the link between EI and leadership. Another curious thing was that very few researchers seemed aware of the extent of the prior literature on this topic. This seemed like a great opening for meta-analytic research. A controversial topic that was badly in need a comprehensive review. So I phoned one of the most gifted meta-analysts I knew, Marcus Credé , and enlisted his help in gathering and coding the articles. We had a feeling that this paper would make a big splash so we raced to complete the project in a matter of weeks.
The results were stunning. Although we found a robust relationship between EI and transformational leadership, this relationship seemed largely driven by poorly designed studies. In particular, same-source method effects appeared to be responsible. When we looked at the studies that employed better designs, the relationship between EI and leadership nearly disappeared. This was true regardless of what EI measure we tested. Our subsequent analyses showed that EI measures contributed 0% incremental validity above and beyond cognitive abilities and Big Five personality traits. That is pretty conclusive evidence that either there is no relationship between EI and effective leadership or that EI measures are poorly designed. Both explanations should give pause to anyone thinking about pursuing further research on this topic.
My hope for this article is that it is read by both researchers and practitioners. Researchers investigating EI need to be aware of the severe limitations of the EI measures currently available. They also need to consider how poorly designed studies (whether they deal with EI or not) can give rise to spurious effects. As for practitioners, my hope is that they will see the need to demand that test publishers conduct proper validation studies before they sell assessment tools or make outrageous claims about the effectiveness of their products.