A model and sensitivity analysis of the quality economics of defect-detection techniques
This work addresses software development cost management for practitioners, but it is incremental as it builds on existing economic models by extending them to include more techniques.
The paper tackles the problem of understanding cost and revenue factors in software defect detection by proposing a stochastic economic model that incorporates both dynamic and static techniques, and uses sensitivity analysis to identify key factors for model simplification and research prioritization.
One of the main cost factors in software development is the detection and removal of defects. However, the relationships and influencing factors of the costs and revenues of defect-detection techniques are still not well understood. This paper proposes an analytical, stochastic model of the economics of defect detection and removal to improve this understanding. The model is able to incorporate dynamic as well as static techniques in contrast to most other models of that kind. We especially analyse the model with state-ofthe-art sensitivity analysis methods to (1) identify the most relevant factors for model simplification and (2) prioritise the factors to guide further research and measurements.