Effectiveness of Traditional Software Metrics for Object-Oriented Systems

Author(s): David Tegarden, Steven Sheetz, and David Monarchi
Venue: Proceedings of the Twenty-Fifth Hawaii International Conference on System Sciences
Date: 1992

Type of Experiement: Case Study



This paper investigates the different possible indicators of complexity including cyclomatic complexity, lines of code, and software science. It then continues on to describe the relationship between inheritance and polymorphism and complexity. The results show that the traditional metrics are able to describe the complexity of an OO system fairly well, but only in direction. The magnitude of the resulting numbers are suspect, but the complexity is at least pointing in the right direction. As for the results for polymorphism and inheritance this study found that the minimal complexity measure based on the traditional metrics was obtained through the use of both of these, but still there was significant decrease in complexity when either one was used.