Idiographic Time-Varying Networks of Personality Facets

Modern theories conceptualize personality as a dynamic system of interacting psychological states, a perspective that echoes early accounts. These early theorists also emphasized the continual differentiation of personality, viewing ongoing change as intrinsic to the person’s dynamic organization. Yet most empirical models treat the relations among personality components as static over time. To test this assumption, the present study applied kernel-based vector autoregressive models to estimate person-specific contemporaneous and lagged time-varying networks of Big Five facet states using intensive longitudinal data from three undergraduate samples and one community sample (N = 392; 31,564 total surveys). Across samples, 76.8% of contemporaneous and 74.5% of lagged networks exhibited non-stationary dynamics, meaning that the structural relations among facets changed over the two- to three-week sampling period. Network non-stationarity was modestly associated with variability in perceived situational valence (pOsitivity and Negativity) but not with traits, demographics, or well-being measures. Time-varying networks were more predictive than corresponding stationary networks, though only contemporaneous models demonstrated acceptable reliability. These findings support dynamic systems perspectives of personality as an evolving, temporally embedded process, suggesting that short-term differentiation in personality structure represents a meaningful and underrecognized source of psychological variability.