Abstract
Background: Multi-center research initiatives offer opportunities to develop and strengthen connections among researchers. These initiatives often have goals of increased scientific collaboration which can be examined using social network analysis.Methods: The National Cancer Institute (NCI)-funded Implementation Science Centers in Cancer Control (ISC3) initiative conducted an online social network survey in its first year of funding (2020) to examine early scientific linkages among members (faculty, staff, trainees) and recognize areas for network growth. Members of the seven funded centers and NCI program staff identified collaborations in: planning/conducting research, capacity building, product development, scientific dissemination, practice/policy dissemination.Results: Of the 192 invitees, 182 network members completed the survey (95%). The most prevalent roles were faculty (60%) and research staff (24%). Almost one-quarter (23%) of members reported advanced expertise in implementation science (IS), 42% intermediate, and 35% beginner. Most members were female (69%) and white (79%). Across all collaboration activities, the network had a density of 14%, suggesting high cohesion for its first year. One-third (33%) of collaboration ties were between members from different centers. Degree centralization (0.33) and betweenness centralization (0.07) measures suggest a fairly saturated network (no one or few central member(s) holding all connections). The most prevalent and densely connected collaboration network was for planning/conducting research (1470 ties; 8% density). Practice/policy dissemination had the fewest collaboration ties (284), lowest density (3%), and largest number of non-connected members (n=43). Median degree (number of collaborations) varied across member characteristics and collaboration activities. Members with advanced IS expertise were more connected than intermediate/beginner groups for most activities (e.g., advanced IS members had a median of 24 capacity building collaborations (range: 4-58) vs. intermediate (median 9; range 2-53) and beginner (median 7; range 1-49) members. The number of practice/policy dissemination collaborations were similarly low across IS expertise levels (median degree 3 for advanced, 2 intermediate, 2 beginner). Conclusions: Results provide important directions for interventions within the ISC3 network to increase scientific collaboration and capacity, with a focus on growing cross-center collaborations and increasing engagement of under-represented groups. Findings will be used to capture infrastructure development as part of the initiative’s evaluation.
International Scientific Collaboration in Research of Depression based on Social Network Analysis