A Series of Controlled Experiments on Software Maintenance

Software maintenance entails the comprehension of often large, complex systems under constant change, and consumes the majority of software development resources [39]. The ISO 9126 model defines maintainability as “a set of attributes that bear on the effort needed to make specified modifications.” It can also be viewed as a two-dimensional characteristic involving both the effort expended on implementing changes and the resulting quality of the changes [3]. The effort required to make correct changes to a software system depends on many factors, including characteristics of the software system itself (e.g., code, design, architecture, and documentation), the development environment and tools, the software engineering process used, and human skills and experience. To better understand how maintenance performance is affected by such a complex combinations of factors, empirical studies are needed. As will be demonstrated in this chapter, once a credible body of empirical evidence has been developed, it can be used to propose improvements that can result in large cost savings and, as it turns out, carried out by quite simple means.