Recognizing Communication Patterns in Chronic Care Innovation Networks

In this paper we describe a methodology to measure communication behaviors in innovation teams. We collected data over two years by mining email communication archives and applied the methodology to study innovation teams working to re-design the delivery of care for chronically ill children. We used social network analysis and sentiment analysis to calculate network metrics such as group density, group betweenness centrality, actor betweenness centrality, average weighted variance in contribution index, and emotionality ratio. We administered surveys to senior leaders to assess their perception of teams’ performance. We conducted “virtual mirroring” sessions with teams to share preliminary results. This paper explores the associations between performance, network density and average weighted variance in contribution index. Our preliminary findings show that high-performing teams tend to interact in less close-knit networks characterized by lower network density and a more balanced exchange of information among team members.