Evidence-Based Software Engineering

Software development is critical for the modern society. Still, many software projects are completed late and have substantial cost overruns. Recently, agile software development has emerged as a radical new way of organizing the development process, stro ngly emphasizing change tolerance, evolutionary delivery and active end-user involvement. However, there are many agile development methods, few evidence-based guidelines and little scientific knowledge of the principal factors that enable the reported im provements. This is a barrier to the successful adoption of agile methods in industrial projects. EvidenSE aims at overcoming this barrier by developing theories and evidence-based methods for the trade-off between agility and discipline in effective so ftware development with a strong focus on project management, collaborative processes, and team learning. The theory building activity includes the cumulation of scientific knowledge related to agile project management, collaborative learning processes i n agile teams, and a framework for describing software engineering theories. The evidence-based methods activity includes developing guidelines for conducting, evaluating and reporting from individual empirical software engineering studies, synthesizing evidence from disparate individual studies, and developing guidelines for evidence-based professional software engineering practice. EvidenSE will be tightly integrated with applied research projects with the Norwegian software industry, and will be carr ied out in close cooperation with top international researchers in relevant fields. The project will include two PhD candidates.