AbstractAcknowledging that the composition and structure of personal networks is affected by meeting opportunities, social distance, and national origin similarity, we aim to disentangle their association with triadic closure in the core of personal networks. We use data (collected 2009) on the core networks of three groups of Swedes (all born in 1990): native Swedes, and first- and second-generation immigrants from Iran and former Yugoslavia, where the respondent (ego) mentions up to five core network members (alters) and whether each pair of alters (dyad) know each other (triadic closure). A three-level multiple membership logistic regression model is performed, which allows the testing of dyadic alter-alter effects, ego effects, and their interaction (i.e., ‘triadic’ effects) on triadic closure. We show that social distance, national origin similarity, and the sharing of social contexts are all associated with triadic closure in the expected direction, and that the effects of social distance and national origin similarity become smaller if shared social contexts are taken into account. The effects of the sharing of social contexts are the largest and are robust, indicating that shared social contexts are a dominant and more important condition for triadic closure than are similarity on relevant socio-demographic characteristics.
Social distance in France: Evolution of homogeneity within personal networks from 2001 to 2017
