Validating Requirements Engineering Process Improvements - A Case Study

Abstract

The quality of the Requirements Engineering (RE) process plays a critical role in successfully developing software systems. Often, in software organizations, RE processes are assessed and improvements are applied to overcome their deficiency. However, such improvements may not yield desired results for two reasons. First, the assessed deficiency may be inaccurate because of ambiguities in measurement. Second, the improvements are not validated to ascertain their correctness to overcome the process deficiency. Therefore, a Requirements Engineering Process Improvement (REPI) exercise may fail to establish its purpose. A major shortfall in validating RE processes is the difficulty in representing process parameters in some cognitive form. We address this issue with an REPI framework that has both measurement and visual validation properties. The REPI validation method presented is empirically tested based on a case study in a large software organization. The results are promising towards considering this REPI validation method in practice by organizations.