engleski

Examining validity of the three ability-based emotional intelligence measures within CFA-MTMM framework

The paper presents the data on empirical validation of the three ability-based measures of emotional intelligence (EI) intended for early adolescents. The data derive from the longitudinal study of EI development conducted in three time points on the sample of 517 primary school students aged 10 to 15 years. Performance based EI measures administered in the study included: Perception of affective content in art test (TAES ; Takšić et al., 2004), adapted version of the Emotional analysis test (TAE: Kulenović, 2003) and Emotion management test (TUE, Babić & Buško, 2013). Along with the objective measures of EI abilities, peer ratings and teacher ratings of the emotion-related reasoning abilities pertaining to the same branches of the Mayer and Salovey’s (1997) model, that is, perception, understanding, and management of emotions, were also collected. The EI data from all three sources were analysed within the SEM methodology using the multitrait- multimethod framework. Several nested CFA-MTMM models with three latent EI dimensions and method factors were specified and compared. The parameter estimates of the best fitting models suggest a moderate level of convergence of different indicators of the specified latent dimensions with high correlations obtained among the latent factors. The estimated factor loadings were expectedly higher for ratings than for performance based measures. Results point to low discriminant validity of EI assessments, which particularly applies to teacher ratings, and to a somewhat lower extent to peer ratings of EI abilities, with strong effects of different measurement methods on observed variability in EI measures. The obtained latent structure is interesting as from the measurement and methodological standpoint so from the theoretical and interpretational view.

emotional intelligence ; ability-based measures ; validation

nije evidentirano