Second-Order Growth Mixture Modeling in Organizational Psychology: An Application in the Study of Job Performance Using the Cusp Catastrophe Model

In recent years, research in organizational psychology has witnessed a shift inattention from a mostly variable-focused approach, to a mostly person-focused approach. Indeed, it has been widely recognized that the study of a sample’s heterogeneity is a meaningful and necessary task of researchers dealing with human behavior in organizational contexts. As a consequence, there has been growing interest in the application of statistical analyses able to uncover latent sub-groups of individuals. The present contribution was conceived as a tutorial for the application of one of these statistical analyses, namely second-order growth mixture modeling, and to illustrate its inner links with concepts from non-linear dynamic models. Throughout the paper, we provided (a) a discussion on the relationships between growth mixture modeling and the cusp catastrophe model; (b) Mplus syntaxes and output excerpts of a longitudinal analysis conducted on job performance (N = 420 employees rated once a year for four consecutive years); (c) an overview of two important topics regarding the correct implementation of growth mixture modeling (i.e., optimal number of classes and local maxima).