An empirical investigation into the dimensions of run-time coupling in Java programs

  • Authors:

  • Áine Mitchell;James F. Power

  • Affiliations:

  • National University of Ireland, Maynooth, Co. Kildare, Ireland;National University of Ireland, Maynooth, Co. Kildare, Ireland

  • Venue:
  • Proceedings of the 3rd international symposium on Principles and practice of programming in Java
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

Software quality is an important external software attribute that is difficult to measure objectively. Several studies have identified a clear empirical relationship between static coupling metrics and software quality. However due to the nature of object-oriented programs, static metrics fail to quantify all the underlying dimensions of coupling, as program behaviour is a function of its operational environment as well as the complexity of the source code. In this paper a set of run-time object-oriented coupling metrics are described. A method of collecting such metrics which utilises the Java Platform Debug Architecture is described and a collection of Java programs from the SPECjvm98 benchmark suite are evaluated. A number of statistical techniques including descriptive statistics, a correlation study and principal component analysis are used to assess the fundamental properties of the measures and investigate whether they are redundant with respect to the Chidamber and Kemerer static CBO metric. Results to date indicate that run-time coupling metrics can provide an interesting and informative qualitative analysis of a program and complement existing static coupling metrics.