Preference Clusters and Academic Outcomes

We measured time, risk, social, and competitive preferences in a sample of 1,035 secondary school students in Hungary. We examine whether meaningful groups of adolescents can be identified based on distinct constellations of these preferences, which we refer to as preference clusters. We also explore whether these clusters are associated with academic outcomes. Using cluster analysis, we consistently identify a group of students who are relatively more patient, less time-inconsistent, more risk-tolerant, more prosocial, and more cooperative. This preference cluster is positively associated with higher scores on standardized math and reading tests. Comparing cluster-based specifications to models with individual preferences entered separately, we find little loss in explanatory power, while preference clusters provide a more compact and interpretable description of how preferences are jointly organized within students.