Dynamic networks of social contact, social desire, and affect across time scales

Social relationships are central to well-being because they fulfill social affiliation needs. To explain how social needs are regulated, theories describe daily-life processes among social desire, social contact, and affect. Still, these processes remain empirically underexplored because of their complexity. In this study, we estimated multivariate associations of social desire and affect with social contact across different modalities (in-person, digital), time scales (hourly, daily), and levels of analysis (between-person, contemporaneous, temporally lagged). Participants from two age-heterogeneous samples answered experience sampling questions and contributed data through unobtrusive smartphone sensing, with roughly hourly assessments across 2 days (N = 303) and daily assessments across 14 days (N = 377). Multilevel vector autoregressive network models revealed associations between social contact, social desire, and affect across levels of analysis. Results were highly specific to the examined time scale. When measured at an hourly timescale, people desired more social contact than usual when they engaged in more in-person contact, and higher social desire predicted more future social contact in both experience sampling and smartphone sensing. In contrast, at a daily timescale, social desire did not predict future contact. Bidirectional linkages of affect and social contact were also much denser hourly (vs. daily). Compared with in-person contact, calls and communication app usage generally showed distinct associations with affect. We discuss theoretical implications for the dynamic regulation of social needs, especially regarding homeostatic temporal processes and the role of positive affect in predicting social contact. Finally, we delineate future directions of multimethod research into daily-life social dynamics. (PsycInfo Database Record (c) 2026 APA, all rights reserved).