Software quality is a topic of importance throughout the world. Unfortunately software quality assurance groups are often understaffed for the work at hand, and also undercapitalized and under equipped. This presentation attempts to cover the known factors which influence software quality results, including methodologies, tools, and staffing levels.
The presentation provides empirical data on the impact of major quality approaches, such as requirements modeling, static analysis, formal inspections, the six-sigma approach, PSP/TSP, clean-room methods, ISO certification, the Software Engineering Institute (SEI) capability (CMM) level concept, and other topics that can impact overall quality levels by as much as 1%. The presentation utilizes the function point metric for quantifying quality results. U.S. software averages about 5 defects per function point, with about 85% of these being removed prior to delivery. The best results have defects below 2.0 per function point combined with 99.6% removal efficiency. Empirical data is provided on software quality levels in a number of industries, and in the major industrialized countries.