Social Network Types, Social Support and SES: Conditional Indirect Effect on Psychological Distress

By pursuing a person-centered approach, multidimensional view, and Berkman’s conceptual model, this thesis sheds light on the mechanisms linking one’s social world and psychological distress. This dissertation continuous the growing tradition of bridging the gap between the egocentric social network analysis and health-oriented investigations. It uses fuzzy-oriented paradigm (fuzzy c-means clustering procedure) to determine social network types and implements a series of mediation and moderated mediation models (based on Hayes’ statistical tool) to ascertain the mediating role of social support and moderating role of SES in the association between network types and distress. The paper employs extensive egocentric data of the UC Berkeley Social Networks Survey (UCNets) collected in 2015. The results suggest that members of the four community-based networks tend to benefit from the provision of social support, whereas family-based (Small – Dense – Negative – Close family/Partner) web of personal relationships deteriorates ones’ psychological state. The more scrutinized moderated mediation analysis showed that these effects apply only to high- or low-income groups, while moderately wealthy population does not seem to gain health-related (dis-)advantages from social assistance. In contrast to income, there was no evidence that education as an element of socioeconomic status (SES) operates as a moderator in the proposed relationships.