I don't think that I've ever led or been involved in social network data-collection that led to anything but one type of relational structure. Whether sorority in Ohio, elementary or middle school in Colorado, scientist community in Japan, rural primary care providers in the Rocky Mountains, biomedical researchers in Michigan, healthcare managers in Oregon, physicians in a big healthcare system, the directors of substance abuse treatment clinics in New York or Washington, judges in Pennsylvania, or STEM educators in community colleges, it's always a radial, hub & spokes, center-periphery structure of advice-seeking and advice-giving. Influence structures consistently exhibit considerable variance in individual-level and organizational-level informal influence. A small proportion of nodes have a lot of influence, the vast majority have very little. Sure I've had some results of other types, but in every case that has been due to sampling inadequacies. If you don't collect data from enough respondents who comprise a network, you may not see the connections that are there.
Over the course of a program of study or a career, general tendencies are a beautiful thing. They are rare compared to all the other times when data analysis produces contradiction or ambiguous results, or the common experience of multiple studies of a type just not producing consistent evidence. Patterns are beauty in communication science.
- Jim DearingShare via these networks: