Developing Socially-Constructed Quality Metrics in Agile: A Multi-Faceted Perspective

Abstract

This research proposes development of socially-constructed metrics for quality assessment and improvement in Agile Software Development (ASD) projects. The first phase of our research includes an extensive literature review, which indicates that traditional (outcome-focused) metrics that evaluate quality are not directly transferable to adaptive, ASD projects. We then conduct semi-structured interviews confirming the necessity of considering people and process aspects for quality considerations in agile. We propose three dimensions for composite metrics in ASD, namely, (1) evidence (2) expectation and (3) critical evaluation. This combines quantitative and qualitative information drawn from people, process, and outcome-related factors. The proposed model allows ASD teams to concurrently conduct quality assessment and improvement during their projects, producing innovative metrics, adhering to the core principles of the agile manifesto. In our next research stage, this reference model will be tested and validated in practice.