A qualitative study on best practices and process of eliciting unambiguous quality attributes in scrum-based projects

Software quality is very crucial for users’ satisfaction and software success in the market. Recent studies found some ambiguous software quality attributes that may lead to low-quality software, and lack of elicitation practices in projects that apply Agile Software Development (ASD) methodology especially Scrum. However,
current ASD methodologies, systematic literature reviews and surveys did not provide explanation of the necessary elicitation practices. Therefore, this qualitative study was essential to achieve two objectives: exploring the best practices and identifying process of eliciting unambiguous quality attributes in Scrum-based
projects. The study used qualitative approach in which data was collected via interviewing eight experienced software practitioners from India and documents analysis that explains documentation of quality attributes in Scrum. For data analysis, the thematic analysis method was used for analysing interviews scripts and documents. The findings resulted in three initial themes that represent three steps in the elicitation process and six sub-themes that represent the elicitation practices. The first step is proactive exposure to quality attributes which consists of two practices: understanding software scope and envisaging potential quality attributes. The second step is mutual learning discussion which consists of two practices: ameliorating technical knowledge of customers and users and compiling details of quality attributes. The third step is verifying common understanding which consists of two practices: utilization of visual artefacts and documentation of quality attributes. The
contribution of the study provides an extension to ASD body of knowledge regarding the effectiveness of disambiguation of terminologies in software domain, simplifying technical terms, representing reusable software artefacts, showing similar software,
drawing mock-up and developing proof of concept in eliciting unambiguous quality attributes. Furthermore, the findings accentuate practical contributions to the software developers such as reducing effort, time and cost of designing and construction of software.