Improving Software Testing by Agile Teams Through The CMMI Adoption: A Cross-Industry Study

Authors

  • Idowu Olugbenga ADEWUMI School of Engineering, Federal College of Agriculture, Ibadan, Nigeria Author
  • Wumi AJAYI Babcock University Author
  • Abdurrazaq Olusola OGUNLADE Lead City University image/svg+xml Author
  • Ayoola PHILIPS Lead City University image/svg+xml Author

DOI:

https://doi.org/10.64358/8cmv0m68

Keywords:

Capability Maturity Model Integration (CMMI), Agile Software Testing, Software Quality Assurance (SQA), Defect Detection Efficiency (DDE), Cost of Quality (CoQ), Test Coverage, Defect Leakage

Abstract

The research study empirically evaluates the impact of Capability Maturity Model Integration (CMMI) application on Agile software testing techniques. The study draws its data from 22 automotive software projects (11 CMMI-aligned and 11 non-CMMI-aligned). The quantitative data show that the DDE of CMMI-Agile teams is 85.7%. Non-CMMI teams have a DDE of 71.3%, which is a relative improvement of 20.2%. There was an increase in test coverage from 65.3% to 79.5% (+14.2%) and a reduction in cycle time from 10.1 to 7.9 days (−21.8%). Leakage of defect reduced from 12.5% to 5.0% (−60.0%) while average Cost of Quality, CoQ reduced from USD 14,957 to USD 11,892 (−20.5%).  Furthermore, these teams implemented more automation would be in CMMI-Agile (71.5 versus 45.3 +26.2). Analysis of costs revealed that the prevention and appraisal of the CMMI teams increased by 60.6% and 15.0% respectively, while the internal and external failure costs reduced by 44.5% and 44.6%. Survey of 26 QA leads found that documentation quality, DFT identification and collaboration efficiency were seen to have improved at a rate of 63%, 72% and 56% respectively.  However, cultural resistance was noted at a rate of 41%. According to these findings, aligning maturity significantly improves the results of software quality assurance compared to previous agile non-CMMI metrics. The findings suggest that CMMI improves defect detection by 22%, cycle time by 18%, and failure cost by 45% thus justifying maturity-based Agile quality management in high-stakes areas like automotive software development.

Downloads

Published

2026-03-31