Predicting and Optimising the Cost of Quality in Enterprise Software Testing: A Comparative Multivariate Regression and Random Forest Framework
DOI:
https://doi.org/10.31305/trjtm2026.v06.n02.006Keywords:
cost of quality, software testing, multivariate regression, random forest, predictive analytics, quality optimisation, machine learningAbstract
The Cost of Quality (CoQ) aggregates the prevention, appraisal, and failure costs incurred while delivering software, yet most quality assurance (QA) organisations still manage it retrospectively through disconnected dashboards. This study proposes and empirically evaluates a release-level predictive framework that couples an interpretable multivariate linear regression (MLR) model with a non-linear Random Forest (RF) regressor to identify, quantify, and forecast the drivers of CoQ. A structured dataset of thirty consecutive enterprise product releases, each characterised by twenty-six engineering and quality predictors spanning testing intensity, operational overhead, governance, and development quality, was analysed. The MLR model achieved a near-perfect in-sample fit (R² = 0.99) but degraded sharply under leave-one-out cross-validation (LOOCV, R² = 0.20), exposing the overfitting risk of high-dimensional linear models on small release histories. The RF model generalised more robustly (LOOCV R² = 0.41; MAE ≈ 11% lower than MLR) and ranked requirement volume, requirement volatility, and defect leakage as the dominant cost drivers. Triangulating both models reveals that CoQ behaves as a high-cost equilibrium in which scale-driven cost escalators are offset by compensatory stabilising mechanisms. The findings support a strategic shift from volume-driven quality assurance towards intelligence-driven quality optimisation, and the framework offers practitioners a reproducible, low-overhead pathway from reactive reporting to predictive decision-making.
References
Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32. https://doi.org/10.1023/A:1010933404324
Crosby, P. B. (1979). Quality is free: The art of making quality certain. McGraw-Hill.
Feigenbaum, A. V. (1956). Total quality control. Harvard Business Review, 34(6), 93–101.
Fenton, N., & Bieman, J. (2014). Software metrics: A rigorous and practical approach (3rd ed.). CRC Press.
Hastie, T., Tibshirani, R., & Friedman, J. (2009). The elements of statistical learning: Data mining, inference, and prediction (2nd ed.). Springer.
Juran, J. M., & De Feo, J. A. (2016). Juran’s quality handbook: The complete guide to performance excellence (7th ed.). McGraw-Hill Education.
Krasner, H. (2021). The cost of poor software quality in the US: A 2020 report. Consortium for Information & Software Quality (CISQ).
Omri, S., & Sinz, C. (2020). Deep learning for software defect prediction: A survey. In Proceedings of the IEEE/ACM 42nd International Conference on Software Engineering Workshops (pp. 209–214). ACM. https://doi.org/10.1145/3387940.3391463
Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M., & Duchesnay, É. (2011). Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12, 2825–2830.
Slaughter, S. A., Harter, D. E., & Krishnan, M. S. (1998). Evaluating the cost of software quality. Communications of the ACM, 41(8), 67–73. https://doi.org/10.1145/280324.280335
Smith, A. E., & Mason, A. K. (1997). Cost estimation predictive modeling: Regression versus neural network. The Engineering Economist, 42(2), 137–161. https://doi.org/10.1080/00137919708903174
Stone, M. (1974). Cross-validatory choice and assessment of statistical predictions. Journal of the Royal Statistical Society: Series B, 36(2), 111–133.
