Abstract
<p>Based on analyses with the necessary condition analysis (NCA) and mediation analyses, Zhao et al. concluded that emotional intelligence and teacher professional identity were necessary conditions for critical thinking disposition and, moreover, that emotional intelligence had an effect on critical thinking disposition partly mediated by teacher professional identity. However, we show that the necessity effects could be accounted for by correlations between the variables, meaning that they did not require true necessity and could, therefore, have been spurious. Moreover, both the necessity effects and the partly mediated effect of emotional intelligence on critical thinking disposition were compatible with a parsimonious model where the three manifest variables were defined as indicators of a latent positive self-evaluation factor. It is important for researchers to bear in mind that correlations, including effects in the NCA and mediated regression effects, in observational (i.e., non-experimental) data may be spurious. We recommend fitting alternative models to data for a more comprehensive evaluation of what conclusions the data at hand may support.</p>