A framework to measure coupling using static change impact analysis
Subhrakanta Panda, Durga Prasad Mohapatra · International Journal of Business Information Systems · 2016
Software defects can often be the result of many changes made to a software over a period. Incomplete or incorrect comprehension of the program segments affected by the changes may cost regression testing. Therefore, identification and testing of such program segments early in the testing phase can reveal more regression errors. Driven by this motivation, this paper uses a new slicing technique to identify the affected program parts through static change impact analysis. The probability of a program segment being erroneous is then shown by computing the coupling factor of the affected program parts. We propose a novel static approach to calculating the coupling factor of these affected program segments to predict the erroneous parts. The theoretical validation and comparative analysis shows the correctness and preciseness of the approach.