Decision support material to incorporate quality requirements technologies: a systematic literature review and industrial interviews

Abstract

AbstractBackground: In an applied engineering field like requirements engineering, the final goal of the research is its adoption by industry. For technology transfer to be possible practitioners need strong evidence for a technology s applicability and possible benefits. Therefore research on a technology should be done in a realistic setting so that practitioners can use the results to build a strong case for its adoption. Quality requirements are constraints placed on the software. Empirical results show challenges in handling quality requirements, e.g. include late discovery of quality requirements. In addition, we lack technologies that can be used to incorporate them into the final software. Functional and quality requirements differ in nature; treatment given to functional requirements will not always be applicable for quality requirements. Therefore practitioners need decision support material based on empirical evidence to incorporate the suggested technologies.Objective: The purpose of the thesis is to identify technologies for quality requirements that have been empirically evaluated. Methods in relation to elicitation, specification, metrics (or measurement) and testing will be identified with the aim of providing decision support material to practitioners for incorporating quality requirements in the software. To find possible future directions of requirements engineering research, the current state of technology adoption for quality requirements will be identified. Method: The presented research is explorative and investigative in nature. A systematic literature review method was employed to identify potential technologies for adoption. An empirical study was conducted with three participants from three companies to get an insight into the state of technology adoption. Results: The systematic literature review includes 46 papers published between 2000 and 2010. All in all only four of the 46 papers offer high realism and support for technology adoption, i.e. presenting evaluations in a realistic setting, with practitioners using real world industrial applications. Another three papers were found to have potential in terms of furthering technology transfer. A general finding common for many of the papers reviewed is a lack of scientific rigour which affects the credibility of the results. Among the participants in the three companies interviewed, none of the methods presented were used. Conclusions: Technology transfer support for quality requirements technologies is challenged by low strength of evidence. Evaluations of technologies lack descriptions of evaluation design - description of how evaluations have been performed - and validity. There is a need for more and better empirical evaluations of technologies to handle quality requirements. The results of industrial interviews shows the need for more empirical investigations, for example surveys, to identify current industrial practices and technologies able to handle or incorporate quality requirements in projects successfully.