Graphical and statistical analysis of the software evolution using coupling and cohesion metrics — An exploratory study

Developing software is expensive; thus keep it useful to its users is important. On the other hand, due to constant maintenance performed to meet the changing needs of users, software undergoes degradation of its internal structure, particularly in coupling and cohesion. Monitoring the development of software by using some of its versions can aid Software Engineer with relevant information to guide your maintenance activities. In this paper, we presented a view of the evolution of versions of software. For this, a study was conducted in 10 versions of FindBugs using coupling and cohesion metrics calculated from VizzMaintenance and Metric plug-ins. In this study, we applied the Pearson linear correlation analysis among measurements. The result showed that there is some correlation between these metrics, because coupling metrics directly influenced the cohesion metrics, with undesirable characteristics such as high coupling and low cohesion compromising software quality.