Replication Data for: Chapter 2, Unveiling the determinants of scientific collaboration and persistence: a network analysis perspective

The chapter results are based on 10 different panel datasets: 5 for tie formation, and 5 for tie persistence (see Methodology section of Chapter 2). Each of the panel datasets uses a different community detection algorithm. The five algorithms used, as described in the data section of the chapter, are: Fastgreedy, Infomap, Multilevel, Spinglass and Walktrap.