Enhancing Software Process Management through Control Charts

Abstract

In software development life cycle, Software Process Management (SPM) acts as a significant part throughout the execution of project. In this study, the application of control chart for analyzing the stability of software process and defects in the software product is discussed. This paper will discuss the analyzing impact or collision of rework effort, defect density, inspection performance and productivity by using control charts. This paper also explains the benefits and challenges of using control charts in software organization.

Measures Suitable for SPC: A Systematic Mapping

“…Measures related to defects are often used in SPC for two main reasons: (i) processes addressing defects-related measures are directly related to software quality and, thus, are critical to organizations and natural candidates to be submmited to SPC, since critical processes are the ones indicated to be statistically controlled [Tarhan and Demirors 2008;SEI 2010;Barcellos et al 2013]; (ii) these processes are performed many times along projects, favoring data collection and geting the amount of data required to apply a measure in SPC. Defect density was the most cited measure, being used in 33 publications (66%), such as [Florence 2001;Jacob and Pillai 2003;Weller and Card 2008;Vijaya and Arumugam 2010;Tarhan and Demirors 2012;Alhassan and Jawawi 2014;Vashisht 2014]. In some studies, this measure is used to quantify different types of defects (e.g., code defect density and file defect density [Zhu et al 2009]).…”

“…However, although publications cite development process, measures are, in fact, related to pieces of the process and, thus, can be suitable for SPC. For instance, the measures effort and productivity [Card 1994;Baldassarre et al 2005;Wang and Li 2005;Boffoli 2006;Wang, Qing et al 2006;Chang and Chu 2008;Gou et al 2009;Tarhan and Demirors 2012;Vashisht 2014] are obtained for each task, activity or phase, producing data usefull to describe the process/subprocess performance. Considering that small processes are more suitable for SPC, some measures are related to parts of processes.…”