Software quality with predictable numbers
Quality Assurance (QA) and Quality Control (QC) have as their ultimate objective the minimizing of customer complaints for a specific deliverable. This article presents a method of estimating the number of errors that need to be located and fixed in a software product. Industry-wide, defects average fifty to eighty per 1,000 lines of code, but organizations can use other approaches to determine the number for their product then determine the effort required to uncover and fix those defects. Data from numerous projects can be put into a historical database and fine-tuned. This concept can be applied in any situation where deliverables are produced.