Abstract:
The software developers are facing a major challenge that over 70% of the software development effort is spent in testing and maintenance of software. Software testing is the most common software quality assessment technique. Software quality cannot be added to the software by testing it, instead software must be developed in a way that guarantees that the software has high quality in every phase. High quality software can be assured by applying the appropriate measurement and testing techniques during software development. Quality Attributes such as maintainability, reusability and testability are useful to find out the extent to which software is useful to undergo changes during the usage phase. At unit testing level, measuring how well a software class can be reused and maintained helps programmers to write reusable and maintainable software, and also helps to identify reusableormaintainable class components. It is widely agreed that there is a direct relationship between poor maintainability and high coupling. In software design there can be various kinds of connections which comprise a coupling relationship. One of such hidden connections between any two seemingly unrelated parts of the system gives rise to indirect form of coupling. newline newlineTo investigate the relationship between indirect coupling and maintenance effort of the object oriented software, this research proposes a metric called Indirect Path Coupling that measures coupling of a class through indirect coupling paths to other classes in the system. The proposed metric takes into account all the indirect coupling paths which are formed between any two classes in the software by considering independent, multiple or partially overlapping multiple indirect coupling paths into account. Extensive case studies have been conducted on several releases of nine open source software to provide empirical software maintenance effort. Indirect path coupling for various classes has been computed by taking into account the various indirect paths formed of different.
