Organizing Empirical Studies as Learning Iterations in Design Science Research Projects

Abstract

Software Quality is a relevant topic that interests both Academy and Industry. Hence, research on this topic should be aligned with the Industry needs, which demands the adoption of research approaches that enable closer interaction between researchers and practitioners. In this context, Design Science Research (DSR) stands out as a way to reduce the gap between theory and practice. DSR is a methodological approach to building innovative artifacts to solve real-world problems and, at the same time, making a scientific contribution. As a problem-oriented research method, DSR seeks to understand the problem, build and evaluate artifacts that allow transforming situations, changing their conditions to better or desirable states. In DSR projects, empirical studies have been usually applied to evaluate the proposed artifact. However, they can also be used to support other activities. Over the last eight years, we have successfully used empirical studies with different purposes in DSR projects. We organized the studies as learning iterations that provide useful knowledge to understand the problem, ground the artifact, develop, evaluate and improve it. As a result, we have experienced a more fluid DSR process and the proposed artifacts have been better grounded and suitable for solving the aimed problem. In this paper, we share our experience by discussing how we have used empirical studies as learning iterations in DSR projects, presenting our approach to organizing empirical studies in a DSR project according to the study purpose and the knowledge it intends to capture, and summarizing two DSR projects that address software quality issues and were developed by using such approach.