Abstract
As is true for most psychological constructs, the statistical methods commonly used to conceptualize alexithymia are rather restrictive. In the present talk, I showcase three uncommon statistical approaches to further explore the construct of alexithymia and to compare different existing measures. First, the analysis of a construct’s measurement invariance across sociodemographic groups (e.g., age) is increasingly well known. Less popular is the idea that such invariance can also be tested across levels of other theoretically relevant psychological constructs (e.g., emotion beliefs or empathy). Second, hierarchical factor analysis can reveal conceptual links between different alexithymia factors and even individual items in one scale or across several scales. This analysis can, for instance, inform researchers about when different scales overlap conceptually and when they measure clearly distinct aspects of alexithymia. Third, we can use additional statistical features to further distinguish between multiple factors of alexithymia. In addition to the prominent distinction between factors through clustering items by similarity in their response patterns, factors can also be distinguished conceptually by their differing degrees of interindividual variance or differences in their average factor score in a population. The three introduced approaches are accompanied by applied examples of alexithymia measurement.