Personality Prediction from Moments to Lifetimes

Personality is robustly linked to long-term outcomes, yet predicting momentary behavior (i.e., who does what, when, and where) remains challenging. We argue that personality and its outcomes can be understood at multiple levels within an emergent hierarchy, and that prediction requires identifying what level(s) is most informative for a given target. Broad traits reliably predict distal outcomes such as cognitive aging and mortality but reveal little about underlying processes. Within-person patterns capture shorter-term regularities, though heterogeneity is often overlooked. Momentary processes unfold as complex, multivariate systems that support personalized prediction but challenge generalization. We therefore propose a bottom-up integrative framework that derives comparable features from person-specific dynamics and tests whether they predict outcomes at other levels. Emerging evidence suggests that such features can bridge levels and support more integrative, explainable prediction.