Empirical software engineering

Abstract

Software engineering as a discipline exists since the 1960s, when participants of the NATO Software Engineering Conference in 1968 at Garmisch, Germany, recognised that there was a "software crisis" due to the increased complexity of the systems and of the software running (on) these systems. The software crisis led to the acknowledgement that software engineering is more than computing theories and efficiency of code and that it requires dedicated research. Thus, this crisis was the starting point of software engineering research. Software engineering research acknowledged early that software engineering is fundamentally an empirical discipline, thus further distinguishing computer science from software engineering, because (1) software is immaterial and does not obey physical laws and (2) software is written by people for people. In this chapter, we first introduce the concepts and principles on which empirical software engineering is based. Then, using these concepts and principles, we describe seminal works that led to the inception and popularisation of empirical software engineering research. We use these seminal works to discuss some idioms, patterns, and styles in empirical software engineering before discussing some challenges that empirical software engineering must overcome in the (near) future. Finally, we conclude and suggest further readings and future directions.