Discovering Objects: Which Identification and Refinement Strategies Do Analysts Really Use?

In the near future industry experts predict the object paradigm will be adopted as the dominant approach to systems analysis, design, and implementation. The promise of the object-oriented approach to systems analysis hinges on correctly partitioning the problem domain into essential classes and objects. Most developers agree that this is no easy task. Limited research has been done to date regarding class and object identification and refinement. This paper presents the findings of an empirical study that (a) assesses the frequency of use of various strategies for identifying and refining classes and objects during systems analysis and (b) tests the hypothesized relationships between strategy usage and several respondent characteristics. Results suggest analysts should learn a base strategic repertoire complemented by selected strategies targeted to the application type and the analyst’s previous background.