Important relationships in a multilevel world: The role of network structure in explaining closeness of relationships and access to resources in later life

We examine why some relationships are more important than others, using a multilevel statistical model and data on personal networks of Australians 50 years and older, collected via a purpose-built Facebook application. While the network data were collected automatically, participants in our study provided data on the importance of their relationships, measured by perceived closeness and access to resources. We find that the information on how network members are connected with each other (network structure) provides powerful insights into what makes a relationship important. When importance is measured by closeness of relationship, important alters are kin, and are alters who are highly connected with others or act as bridges between different groups. When importance is measured by access to resources, important alters are those who act as bridges and are in more densely-knit networks. We discuss our findings in a broader context of research into important relationships in later life, and collecting personal network data via online social networks.