Theoretical Validation and Empirical Evaluation of Object-Oriented Design Metrics

Various object-oriented metrics have been proposed as a way of capturing features of object-oriented software such as encapsulation (information hiding), abstraction and inheritance. A major criticism of past object-oriented metrics is that little attention has been paid to theoretical validation or empirical evaluation of those metrics. By theoretical validation we refer to the process of ensuring that metrics conform to the principles of measurement theory. By empirical evaluation we refer to the study of software in order to characterize, predict, control, manage or improve through qualitative or quantitative analysis. The two approaches complement one another; a valid set of metrics should be both theoretically validated and empirically evaluated before it can be used with confidence. In this article, we describe the validation of objectoriented metrics incorporating both approaches. We begin by looking at theoretical validation.