When Is It Cost-Effective to Use Formal Software Inspections”, Crosstalk, Volume 17 #3 (March, 2004), On-Line Article. This article presents a method quantitatively to determine the parametric limits to costeffectiveness of software code inspections. The

The purpose of this article is to present a way quantitatively to determine the parametric limits to cost effectiveness of software inspections based on a previously published model. This analysis leads to the conclusion that it is cost effective to inspect both original code and most modifications to the code after initial coding. Any exceptions should be carefully considered based on quantitative analysis of the projected impact of the exceptions. Further, any proposed substitution for rigorous inspections should be carefully evaluated for cost effectiveness prior to replacing or modifying the process. In this author’s experience, the two following issues are discussed qualitatively when program management decides what to inspect and what data to collect: • Deciding whether or not to inspect work products based on a qualitative understanding of various limiting parameters such as work product size, preparation rate, or expected defect density. • Deciding whether or not to collect and to analyze inspection data based on a qualitative understanding of the return on investment (ROI) on collecting, analyzing, and using inspection metrics. This article improves on this practice by estimating quantitatively, from the perspective o