Telfer Research Seminar Series - Josianne Marsan

Free/libre and open source software (FLOSS) has evolved significantly over the past 20 years and estimates suggest that it accounts for 80-90% of any given piece of modern software. A consequence of this evolution is that many of the assumptions made by FLOSS researchers may be obsolete. This would have major negative implications for research validity and hampers theory generation on FLOSS. This study sought to identify significant obsolete assumptions that persist in FLOSS research. Using Delphi research design with a panel of 20 expert researchers, 21 obsolete assumptions about FLOSS were identified and ranked. A thematic analysis was performed that resulted in these obsolete assumptions being grouped into six themes: Sampling, Project/Community, Product, Contributor, Evaluation, and Development Process. The Sampling theme was ranked as having the most significant obsolete assumptions although only two assumptions were associated with this theme. The Project/Community theme contained six obsolete assumptions – the most of any theme. This study provides specific elements for FLOSS researchers to challenge and be aware of in future studies, helping to mitigate potential negative impacts on FLOSS research validity and legitimacy. It bridges the FLOSS research-practice gap by tackling obsolete assumptions that threaten real-world applicability of FLOSS research. More accurate and relevant research is expected to encourage greater participation and collaboration from FLOSS practitioners in research projects, thus ultimately contributing to the advancement of knowledge in the field. This presentation is based on a paper presented at the European Conference on Information Systems (ECIS) 2022 and written by Josianne Marsan with PhD candidate Patrick Marois (FSA ULaval) and colleagues Kevin Carillo (Toulouse BS), Klaas-Jan Stol (UC Cork) and Brian Fitzgerald (U of Limerick). The authors are currently working on a version to be submitted to an academic journal and including data from a panel of FLOSS practitioners to compare their view with that of FLOSS researchers.