Emotional intelligence as ability : assessing the construct validity of scores from the Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT)
Appears in the following Collection
- Psykologisk institutt 
AbstractThis thesis presents the results from three papers assessing the validity of the Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT; Mayer, Salovey, & Caruso, 2002). The MSCEIT is the only performance test measuring the entire four-branch ability model of EI (Mayer & Salovey, 1997; Salovey & Mayer, 1990).
Previous studies have reported low reliability coefficients for the branch scores for MSCEIT and reliability estimates vary greatly from study to study. The reported reliability coefficients may be biased or inflated, however, as the many sources of variance in the MSCEIT measurement design have not been taken into account when the coefficients have been estimated. Mixed results have also been reported regarding the construct validity of scores. More studies are therefore needed on the validity of scores from MSCEIT.
In Paper 1, Generalizability Theory (GT) is used to estimate the relative magnitude of the many sources of variance in the measurement design and generalizability (reliability) coefficients for the scores, taking the many explicit sources of variance in scores into account. Participants were 111 leaders of various businesses in Norway. The results reveal some important sources of variance in the measurement design that has neither been specified in the MSCEIT measurement design nor in the theory of EI. Only some of the branches in the MSCEIT provide scores that are generalizable. Moreover, the task Faces (from the branch Perceiving Emotions) provide scores that constitute three correlated factors, representing interactions of type of emotion expressed by faces and type of emotion to be rated. Moreover, the three scales seem to assess ability to identify emotions that are absent in faces, rather than emotions that are present in faces. These results are not in accordance with the theory of EI, suggesting that the validity of the scores derived from MSCEIT need to be assessed further. In Paper 2, the validity of scores from MSCEIT is further assessed by relating the scale scores derived in Paper 1 to subordinates ratings of empathy and leader effectiveness. The theory of EI suggest that EI should be positively related to both empathy and leader effectiveness. Previous studies using empathy as a validity criterion, however, have used self-ratings of empathy despite the validity of self-ratings of emotional abilities have been questioned. In Paper II, both self-ratings and subordinate ratings of empathy were used. Multilevel Confirmatory Factor Analysis (MCFA) with latent variables was used to analyze the scores from 104 leaders (same sample as in Paper 1) and 459 subordinates’ ratings of empathy and leader effectiveness. Scores from MSCEIT were found to be unrelated to ratings of empathy and leader effectiveness, suggesting that the validity of the scores may be questioned.
Paper 3 assesses the validity of the scores from MSCEIT further, by relating them to subordinates’ ratings of their leader’s transformational leadership behavior. It has been suggested that EI should be positively related to transformational leadership. The results from MCFA of 459 subordinates’ ratings of transformational leadership and scores from 104 leaders on MSCEIT (same scales and sample as in Paper 1 and Paper 2) revealed that scores from MSCEIT were unrelated to transformational leadership. Scores from Facilitation, however, was related to one transformational leadership factor, but this relationship was not significant after controlling for personality scores provided by the NEO PI-R.
Together, the results from the three papers question the validity of the scores from MSCEIT. These findings are important, as the MSCEIT is frequently used to measure EI in research and applied settings and is regarded as the test that provides the best available evidence for the four-branch theory of EI. The findings from this thesis thus question the validity of the most important source of evidence for the four-branch ability model of EI.
LIST OF PAPERS. The following papers constitute this thesis:
Paper I: Føllesdal, H., & Hagtvet, K. A. (2008). Emotional Intelligence as ability: The MSCEIT from the perspective of generalizability theory. Manuscript submitted for publication.
Paper II: Føllesdal, H., & Hagtvet, K. A. (2008). Assessing the validity of the Mayer, Salovey, and Caruso Emotional Intelligence Test: Predicting empathy and effectiveness among leaders. Manuscript submitted for publication.
Paper III: Føllesdal, H. & Hagtvet, K. A. (2008). Assessing the validity of the Mayer, Salovey, and Caruso Emotional Intelligence Test 2.0: Predicting transformational leadership. Manuscript submitted for publication.