Analyzing the Quality of Domain Models Developed by Novice Systems Analysts

Assuring quality in capturing and representing the systems requirements is extremely important. This study is aimed at understanding errors frequently committed by novice systems analysts in developing domain models using the Unified Modeling Language (UML). Understanding of errors that affect the quality of resulting conceptual models and the relationships among those errors would be beneficial to both practice and research. This paper reports the results from analyzing the class diagrams produced by teams of novice systems analysts as part of different e-business systems requirements specification. The results include errors grouped into different quality categories and some interesting observations on relationships among errors from different categories.