On the reliability of mapping studies in software engineering

Abstract

BackgroundSystematic literature reviews and systematic mapping studies are becoming increasingly common in software engineering, and hence it becomes even more important to better understand the reliability of such studies. ObjectiveThis paper presents a study of two systematic mapping studies to evaluate the reliability of mapping studies and point out some challenges related to this type of study in software engineering. MethodThe research is based on an in-depth case study of two published mapping studies on software product line testing. ResultsWe found that despite the fact that the two studies are addressing the same topic, there are quite a number of differences when it comes to papers included and in terms of classification of the papers included in the two mapping studies. ConclusionsFrom this we conclude that although mapping studies are important, their reliability cannot simply be taken for granted. Based on the findings we also provide four conjectures that further research has to address to make secondary studies (systematic mapping studies and systematic literature reviews) even more valuable to both researchers and practitioners.