摘要：PURPOSE: Robustness analysis is a technique that can be performed to help ensure the correctness, completeness and consistency of use case and domain models. Robustness analysis also helps bridge the gap between the analysis and design phases by providing a guided approach to identify a first-guess set of objects that will realize scenarios described in use cases. It is necessary to perform robustness analysis in the early phases of the development lifecycle in order to reap its benefits. In particular, robustness analysis needs to be performed by business analysts during the requirements phase to improve the quality of their models as well as help provide a seamless transition to the design phase. However, a core skill that is required to develop robustness diagrams is knowledge of OO concepts which business analysts normally do not have. To overcome this limitation, business analysts acquire brief knowledge of OO concepts via a small learning curve in order to develop and reap the benefits of creating robustness diagrams. However, is this brief knowledge of OO concepts attained through a small learning curve enough to allow business analysts to develop quality robustness diagrams? DESIGN: In this paper we present a controlled student-based experiment to empirically evaluate the requirement of in-depth OO knowledge to produce quality robustness diagrams. FINDINGS: The results show that business analysts can indeed produce quality robustness diagrams without in-depth OO knowledge. ORIGINALITY: The results of this experiment will aid in embracing the technique of robustness analysis amongst business analysts in order to overall improve the software development process.
