PS 26-115 - A Diversity Analysis of a Federally Funded Research Network

Abstract: Following racialized violence, and nationwide calls for social justice, we reflected on our own professional community and whether it supported a diversity of people and thought. National Science Foundation (NSF) is tasked with not only supporting scientific research, but also ensuring no discrimination in who is funded to undertake the research. We investigated the efficacy of NSF-funded research networks in creating robust collaborative communities. Specifically, we tested the hypothesis that national research networks, with the express goal of connecting researchers, grow their collaborative networks over time.

To test this hypothesis, we analyzed publication data from the NSF-funded Critical Zone Observatory (CZO) Network using biodiversity metrics to characterize authorship diversity. We linked papers to individual CZO sites by using the publication database maintained by the CZO network, which relies on self-reporting by authors. For this analysis, an author was considered a “species”, a paper was considered a sampling “site”, and each CZO site was considered a “landscape”. In this way, for each year we have data, we were able to calculate alpha diversity (# of unique authors/paper), gamma diversity (# of unique authors/CZO site in a given year), and beta diversity (here we used a Whittaker’s turnover metric). Looking across years, we then were able to see how “landscape” diversity changed over time and were thus able to determine if the collaboration network’s diversity was growing, shrinking, or stagnating.

Preliminary data from the Shale Hills CZO showed that between 2005 and 2011 all diversity metrics increased rapidly, with average alpha diversity increasing from 2 to 6 authors/paper. From 2011 to 2020, gamma diversity (unique authors on Shale Hills related papers in a given year) continued to increase, while alpha diversity showed no trend. These data indicate that over the 15 year period, this CZO site went through a colonization phase (2005 – 2011) where the authorship network rapidly increased both in terms of total authors publishing data from the site and number of authors on each publication. It then transitioned to a more mature network, where number of authors per publication did not increase, but the total number of authors publishing on Shale Hills related projects continued to increase. This robust collaboration environment is evidence for this research network site meeting the goal of connecting researchers. We see this work as the foundation for understanding research networks’ effects on creating inclusive research communities.