Social network analysis applied at understanding collaboration in software projects

In the realm of software development, collaboration is important for worker satisfaction, mitigating turnover risks and better design of final products. The quality of collaboration in teams can indirectly assessed by the interdependence of team members’ designs and the extent of shared authorship within a system. Our goal is to apply Social Network Analysis to visualize collaboration on software teams. This study is situated within the framework of software projects undertaken by undergraduate Computer Science students in a 3-week programming sprint. Detailed data from the software development process was gathered through Github and further analyzed using a two-mode social network and later a regular social network. Our results indicate that these techniques help illuminate certain facets of collaboration, such as members that are isolated from the team or collaborators that tend to concentrate too much work.