A Co-authorship network analysis of CNPq’s productivity research fellows in the probability and statistic area

Abstract

Abstract In this paper, we analyzed the co-authorship network between all CNPq’s productivity research fellows in the Probability and Statistics area in Brazil. Our aim was to describe and to understand how network measures influence researchers’ productivity. The data was gathered from the CNPq’s Lattes Platform using the software scriptLattes, and a link between two fellows represents the fact that they wrote an article together from 2009 to 2013. The network is disconnected and has only 4.7% of its possible connections. Through a regression analysis, we were able to infer that the centrality positions of an author matters to his/her productivity. As expected, closeness centrality had a negative effect on fellows’ productivity, while the degree centrality had a positive effect.

Keywords

• social network analysis
• scientific performance
• co-authorship
• probability and statistic