Replication of FLOSS research as eResearch
Andrea Wiggins, James Howison, Kevin Crowston · 2008
We are working to introduce the ideas of eResearch to a multi-disciplinary research domain: those researchers examining Free/Libre and Open Source Software (FLOSS) and its development (Howison, Wiggins, & Crowston, 2008). The first phase of this work focused on building a repository for data on FLOSS teams, FLOSSmole (Howison, Conklin, & Crowston, 2006), and collaborating with other nascent data repositories in the field. Recently we have begun a second phase, which is to introduce another established principle of eResearch, that of broader collaboration through shared workflows accessing these data repositories. To provide an example of the potential value of this principle, we are replicating seminal FLOSS papers using eResearch approaches. This paper describes research outcomes and lessons learned from translating published literature into eResearch workflows. The replication involves the reconstruction of workflow, data, and analyses of several published papers in the field of empirical FLOSS software engineering research. The effort has three purposes: to provide examples of workflows and workflow elements for FLOSS research, to determine how replicable current research is, and to provide a