Enhancing the quality of software requirements artifacts with scoring rubrics-assisted reading technique

Abstract

The quality level of software requirements artifacts depends on how adequate esuch artifacts are read. Several reading techniques exist and the most frequently used is the checklist reading technique. Though checklist is an improvement over the ad hoc approach, it is however, laden with some setbacks that impacts on its usability and efficacy. It presents before readers questions that lack detailed guides as to how to carry out the requirements artifacts reading process. In this study, scoring rubrics-assisted reading (SRAR), which over the years has been used in Universiti Utara Malaysia’s School of Computing (Software Engineering sub-department) in the assessment of students’ requirements’ artifacts, was empirically evaluated. SRAR is a checklist-like reading approach that can be used for the evaluation of the quality of software requirements artifacts. The main objective of the study is to evaluate SRAR. Scoring rubrics were employed in the reading of requirements artifacts. The review was done by four independent experts in two iterations. The difference in the mean rubric scores between roundsof reading was used as the performance metrics to measure the improvement in the quality of the artifacts. Initial results indicate that SRAR has efficacy as there was decrease in defects after refinement using SRAR. The paper will be useful to requirements engineers and the research community as the evaluated technique is conceptually a superior alternative to the checklist and ad hoc reading techniques.