Challenges of software process and product quality improvement: catalyzing defect root-cause investigation by process enactment data analysis


Soylemez M., KOLUKISA TARHAN A.

SOFTWARE QUALITY JOURNAL, vol.26, no.2, pp.779-807, 2018 (Journal Indexed in SCI) identifier identifier

  • Publication Type: Article / Article
  • Volume: 26 Issue: 2
  • Publication Date: 2018
  • Doi Number: 10.1007/s11219-016-9334-6
  • Title of Journal : SOFTWARE QUALITY JOURNAL
  • Page Numbers: pp.779-807

Abstract

It is claimed by software quality management that the quality of a software product is highly influenced by the quality of the software process followed to develop it. Since measurement of the software process is a challenging task, it is frequently the defects in the software product that are used to measure development quality. By extracting semantic information from defect records, practitioners can investigate and address root causes of software defects to improve development process and product quality. Investigating root causes requires effort for a detailed analysis into the components of the development process that originated the software defects, and is therefore encouraged only at higher maturity levels by most known process improvement models such as Capability Maturity Model Integration (CMMI). This practice, however, postpones the benefits that root-cause analysis would bring in gaining process awareness to improve the software development process and product quality in emergent organizations or organizations residing at lower maturity levels (MLs). In this article, we present a method for and results from applying root-cause analysis for software defects recorded in a software-intensive project of a CMMI ML3 certified institute. The suggested method combines process enactment data collection and analysis with Orthogonal Defect Classification which is a known technique in defect root-cause analysis. Prior to and after implementing the method in the study, defect attributes were analyzed and compared in order to understand any improvements in development performance and product quality. The results of the comparison indicate that the suggested method was efficient in the effort it required and effective in improving development performance and product quality. Defect triggers have become more active in identifying software defects in the earlier phases of software development, and the cost of quality due to software defects has decreased in consequence.