Towards Growing a COIN in a Medical Research Community

Abstract

In this paper we describe a longitudinal analysis of the evolution of a team of medical researchers who are trying to establish a COIN consisting of patients of a chronic disease, family members of patients, doctors, and researchers. The data analysis is based on an e-mail archive consisting of the full mailboxes of the six project leaders covering the whole period of COIN formation. We conducted a dynamic analysis of the e-mail network looking at social network analysis metrics such as betweenness centrality and density over time, as well as contribution index and core/periphery structure. We also did a semantic analysis of the e-mail subject headings. A netnographic analysis of patients was done through analyzing their Facebook friendship and discussion networks. External perception of team effectiveness was done through a Web buzz analysis. Preliminary results indicate that after the first physical meeting of the full team the betweenness centrality of the official leaders went down noticeably, as tasks were delegated to other team members, whose centrality increased as the leaders’ centrality fell. The patients, who are the main constituency of this project and could contribute invaluable advise, are still surprisingly peripheral in the entire project team and have a low contribution index. We also found that a peripheral position was held by some very senior team members who seem focused almost exclusively on external projects, and are (not yet) well connected to the core of the project team. The team seems to be inwardly focused, engaged in internal coordination activities and scarcely involved in managing across boundaries. To increase the success of COIN formation, the team might need to extend their ties to the peripheral patients and senior members, whose specialized expertise is essential to disseminate knowledge outside of the team boundaries.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 9 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.