A primer on object-oriented measurement

Many object-oriented metrics have been proposed over the last decade. A few of these metrics have undergone some form of empirical validation, and some are actually being used by a number of organizations as part of an effort to manage software quality. There is evidence that using object-oriented metrics can be beneficial to conducting business and to the bottom line. This is encouraging, as estimates and prediction models using design metrics can be used to allocate maintenance resources and for obtaining assurances about software quality. The purpose of this paper is to present an overview of object-oriented metrics, their use and their (empirical) validation. The focus is on static class-level metrics, since these have been studied the most