Case Study Research in Software Engineering: Guidelines and Examples

CHAPTER 7

SCALING UP CASE STUDY RESEARCH TO REAL-WORLD SOFTWARE PRACTICE

7.1 INTRODUCTION

Empirical software engineering research is often conducted in real-world settings with complex and multifaceted characteristics. These characteristics will be difficult to capture and understand in depth if the case study is scoped too narrowly, for example, scoped to a single, small-scale unit of analysis.

Many real-world software engineering cases involve software artifacts, such as code repositories, design descriptions, test suites, and requirements documents, and each of these artifacts may be very large and complex. Investigating these artifacts can therefore produce immense amounts of data to analyze. If one also adds the potential complexity and size of industrial software processes, organizations, methods, tools, and so on, case study research may need to be scaled up from small single case investigations to a combination of multiple large-scale case studies, in order to achieve deeper empirical understanding of real-world software engineering. Thus, one of the greater aims that we, as a software engineering research community can embrace, is to make our research truly relevant by not limiting its investigations to small and convenient “toy” examples. Instead the community should extend our methodological tools to reach for a deeper understanding of the real complexity of real-world, large-scale software engineering.

While embracing this important aim to large-scale studies, we must ...