Short communication: Evolution of secondary studies in software engineering

Abstract

Other disciplines commonly employ secondary studies to address the needs of practitioners and policy-makers. Since being adopted by software engineering in 2004, many have been undertaken by researchers. To assess how the role of secondary studies in software engineering has evolved. We examined a sample of 131 secondary studies published in a set of five major software engineering journals for the years 2010, 2015 and 2020. These were categorised by their type (e.g. mapping study), their research focus (quantitative/qualitative and practice/methodological), as well as the experience of the first authors. Secondary studies are now a well-established research tool. They are predominantly qualitative and there is extensive use of mapping studies to profile research in particular areas. A significant number are clearly produced as part of postgraduate study, although experienced researchers also conduct many secondary studies. They are sometimes also used as part of a multi-method study. Existing guidelines largely focus upon quantitative systematic reviews. Based on our findings, we suggest that more guidance is needed on how to conduct, analyse, and report qualitative secondary studies. • Published secondary studies are predominantly qualitative mapping studies. • Many appear to be derived from postgraduate study. • More than half of them are led by experienced researchers. • Secondary studies are also used as part of multi-method research.