On the need for a process for making reliable quality comparisons with industrial data

  • Authors:
  • Laurie Williams

  • Affiliations:
  • North Carolina State University

  • Venue:
  • ACM SIGSOFT Software Engineering Notes
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

Many factors influence quality data obtained from industrial case studies making comparisons difficult. In this paper, two longitudinal industrial case study experiences are shared which illustrate the complications that can arise. The first is a case study of an IBM team that transitioned to the use of test-driven development. The primary quality measure was functional verification test defects normalized by lines of code. The second case study was performed with an Extreme Programming team at Sabre Airline Solutions. Both test defects and field defects were compared. In both case studies, differences existed which made the comparisons indicative but not absolute.