A Software Complexity Prediction Model using Coupling Metrics: A Statistical Analysis

Abstract: OO programming has become the most popular technology in software development environment as it is been proven that the maintenance of OO software is comparatively lesser than the other programming languages. But still the burden of software maintenance is not completely eradicated. One popular software maintenance approach is the reduction of software maintenance cost by imposing the software evaluation metrics during the development phase of the life cycle. Software metrics helps in identifying the potential problem areas in the code. Many novel metrics have been proposed and only few are validated. The objective of this research is to experimentally explore the two novel OO coupling metrics namely Subclass Coupling Factor (SCF) and Temporal Coupling Factor (TCF) to evaluate their ability to predict the complexity of the built software through statistical validation.

Keywords: OO metrics, software maintenance, SCF, TCF and software complexity.