Software projects: Improving the odds
Projects in global software development (GSD) are believed to be risky and have a high rate of failure. Hence, it may benefit software developers to learn from failure experts. Such was the philosophy guiding the EU-funded project 'Evidence-based risk management in global software development projects' (E-RISK). One goal was to help project managers identify and mitigate risks in outsourced software development. Additionally, the group set targets for training and knowledge transfer, and assessed whether expertise in software engineering could be combined with the SLR technique. The two-year undertaking concluded in May 2013. A major part of the project's first phase involved training of a research fellow in SLR techniques through a pilot study. The work led to several technical reports and one conference paper. The latter was fleshed out during the project's second phase to become a journal paper. During the pilot study, project members found that Brazilian researchers were working on a similar topic. The resulting collaboration produced a research paper, and showed that literature searches for a mapping study on this topic could be conducted reliably using just two search engines. Since poor quality research may not yield reliable results, suitable quality assessment criteria were needed in order to identify, and possibly exclude, poor quality primary studies. Given that suitable criteria were not available from previous research, the project developed its own to classify the quality of papers included in the main SLR. The initial search yielded over 1 500 candidate studies, which was too many to complete with available resources. Hence the criteria were narrowed to include only GSD projects having a client company that outsourced development to a different legal entity. Thereby 77 papers were accepted for analysis. Two researchers worked independently to extract data. The E-RISK project has benefited the software development community by identifying both risks and management strategies. Applying the results of the project to European GSD project management could lead to fewer failures.
Keywords
Software, systematic literature review, global software development, software developers, risk management, outsourced software development, software engineering
