Validating object-oriented software metrics in the agile and framework iterative software processes

Object-oriented (OO) software metrics are mainly used to quantitatively evaluate and predict certain Software Engineering (SE) aspects including process and products. This dissertation empirically validates some OO metrics' predictive capabilities in two iterative processes: the short-cycled agile process and the long-cycled framework evolution process. This dissertation also validates the System Design Instability (SDI) metric by testing its project-progress indicative capability in the short-cycled agile process and comparing the result with a previous study.; We found that OO metrics are effective in predicting added, changed, and deleted source lines of code as well as comment lines in the short-cycled agile process and ineffective in predicting the same aspects in the long-cycled framework process. Object-oriented metrics are also effective in predicting error-fix and refactoring efforts in the agile process. These results lead us to believe that OO metrics' predictive capabilities are limited to the design and implementation changes during the development iterations, not the long-term evolution of an established system in different releases.; We found that the SDI metric can reasonably indicate project progress with evolutionary trends in the agile process. We also found that the SDI metric is correlated with Extreme Programming (XP) activities: new design, error-fix, and refactoring. We observed two jumps in the SDI metric values in early and late development phases consistently in the two systems, a result that agrees with a previous study.