The purpose of this project is to evaluate the growth and productivity of NIMBioS working group teams using social network analysis. Evaluators of collaborative research programs face a daunting challenge of measuring not only the productivity of group members in terms of scholarly products, but also the types of collaborative activities that lead to those products. This presentation will discuss the use of network analysis methods to evaluate the growth and productivity of several interdisciplinary research groups over a three-year period. It will focus on (1) how to look at the patterns of change in group composition of time, (2) what the patterns of connectedness look like across disciplinary and geographic boundaries, and over time, and (3) to what extent network characteristics of productive research group members correlate with productivity. The presentation will highlight the lessons learned from this evaluation, and how the process fits into the science institute’s overall evaluation program.
The National Institute for Mathematical and Biological Synthesis (NIMBioS) is funded through a National Science Foundation award and located on the University of Tennessee campus. The institute draws a diverse cadre of researchers from around the world to take part in interdisciplinary working groups, workshops and conferences to find creative solutions to pressing problems from animal disease to wildfire control. A need exists to evaluate the mechanisms through which researchers at the center are interacting to produce scholarly research in order to fully understand the reasons behind group productivity. This paper attempts to evaluate these mechanisms using network methods where research group participants indicated the nature of their collaborative relationships with other group members before and after each meeting with their groups over a period of about two years.
Several network measures were applied to evaluate growth over time, including size, density, complexity, and centralization of the full network for each group at the each administration of the survey. Average numbers of cliques that have developed, as well as average effective network size were also used. Cross-disciplinary and cross-geographic collaboration were evaluated by examining densities of interactions within and among self-reported disciplinary areas. The relationship between network position (degree centrality and betweeness centrality) and productivity was examined to determine the network characteristics of productive center members and research groups as well.
Preliminary Data Results of Within-Group Collaboration Over Time
The E-I (external – internal) index takes the number of ties of group members to outsiders, subtracts the number of ties to other group members, and divides by the total number of ties. Shown here is collaboration across gender, professional status, primary field of study (FOS), and country where participants reside.