Social selection and psychological distress in a sober active community: a longitudinal social network analysis

Purpose This study aims to investigate the coevolution of supportive social ties and psychological distress within The Phoenix, a sober active community. It also aims to disentangle two concurrent processes: whether psychological distress influences the formation of social ties (social selection) or whether peer relationships influence individual distress levels (social influence) over time. Design/methodology/approach Longitudinal social network data were collected from members of The Phoenix in Denver and Wichita (n = 23) across two waves separated by one year. This study used the Simulation Investigation for Empirical Network Analysis (SIENA) framework to simultaneously model structural network dynamics and changes in psychological distress. Findings Network ties were characterized by reciprocity and clustering. A significant negative selection effect was identified (PE = −0.37, p < 0.05): members with lower psychological distress were more likely to be nominated as supportive partners, while those with higher distress were less likely to be selected. No evidence of peer influence on psychological distress was observed over the one-year period. Research limitations/implications Limitations include small network sizes and a two-wave design, which restricted model complexity. Future research should incorporate additional waves, larger networks and mixed-method approaches to capture nuanced changes in mental health and social integration. Practical implications Findings suggest that members with higher psychological distress are less likely to be selected as supportive ties, potentially isolating those most in need of connection. “Structured socialization” strategies, such as assigned mentorships or care circles, could be implemented to proactively integrate members with higher distress levels who might otherwise be filtered out by the community’s selection mechanisms. Social implications While recovery communities provide vital protection, they can also function as exclusionary structures. To ensure equitable access to recovery capital, communities must adopt intentional strategies that mitigate social stratification based on mental health status. Originality/value This study addresses a gap in recovery literature by using longitudinal network analysis to separate selection from influence. It provides empirical evidence that mental health status drives social connectivity in recovery settings, rather than connectedness simply improving mental health.