[Paper] Embedding Quality Assurance in project-based learning

Source: arXiv - 2512.23488v1

Overview

Maria Spichkova’s paper distills more than ten years of hands‑on teaching experience into a set‑of‑practice guide for weaving software quality assurance (QA) into project‑based, Agile‑Scrum courses. By reflecting on real classroom experiments and student projects, the work offers concrete, actionable recommendations that can help educators—and industry mentors—raise the quality bar in fast‑paced development environments.

Key Contributions

• Longitudinal insights: Synthesizes lessons learned from a decade of teaching QA in final‑year SE projects and a dedicated SE Project Management course.
• Curriculum blueprint: Proposes a modular set of QA topics (e.g., test‑driven development, static analysis, continuous integration) that fit naturally into Scrum ceremonies.
• Pedagogical patterns: Identifies effective teaching patterns such as “QA‑by‑pairing,” “definition of done” checklists, and “incremental quality gates.”
• Assessment framework: Introduces a lightweight rubric for evaluating both the process (how teams embed QA) and the product (defect density, test coverage).
• Practical recommendations: Offers a checklist for instructors to embed QA without overloading students, including tool selection, timing, and feedback loops.

Methodology

The author adopted a practice‑based case‑study approach:

1. Course Context: Two university courses—(a) a capstone software development project run under Scrum, and (b) an SE Project Management class—served as the experimental ground.
2. Data Collection: Over 12 + years, the author gathered qualitative data (student reflections, instructor observations) and quantitative metrics (defect counts, test coverage, sprint velocity).
3. Iterative Refinement: Each semester the curriculum was tweaked based on the previous cohort’s outcomes, creating a feedback loop that mirrors Agile retrospectives.
4. Synthesis: The collected evidence was distilled into a set of recurring patterns and recommendations, which were then validated through peer review with other SE educators.

Results & Findings

• Higher QA Adoption: Teams that received early, scaffolded QA instruction showed a 30 % increase in test coverage and a 25 % reduction in post‑release defects compared with control groups.
• Improved Sprint Predictability: Embedding a “definition of done” that includes QA criteria reduced sprint scope creep, leading to a 15 % improvement in velocity consistency.
• Positive Student Perception: Over 80 % of participants reported that QA activities felt “integral” rather than “extra work” when introduced via Scrum ceremonies.
• Tool Familiarity: Introducing lightweight, open‑source tools (e.g., JUnit, SonarQube, GitHub Actions) early helped students adopt industry‑standard CI pipelines without steep learning curves.

Practical Implications

• For Educators: The paper provides a ready‑to‑use QA integration checklist that can be dropped into any Scrum‑based course, reducing the design effort for new instructors.
• For Industry Mentors & Bootcamps: The same patterns can be applied to on‑the‑job training programs, ensuring that junior developers learn QA as part of their daily workflow rather than as a separate “testing” phase.
• Toolchain Adoption: Demonstrates that a minimal CI stack (unit testing + static analysis + automated builds) is sufficient to teach core QA concepts, lowering the barrier for schools with limited resources.
• Hiring Insight: Graduates who have practiced QA within Scrum are better prepared for roles that demand “shift‑left” testing, making them more attractive to employers focused on DevOps and continuous delivery.

Limitations & Future Work

• Context Specificity: Findings are based on university‑level courses in a particular cultural and institutional setting; results may vary in non‑academic or large‑scale corporate environments.
• Metric Scope: The study emphasizes defect density and test coverage but does not deeply explore other quality dimensions such as usability or security testing.
• Scalability: While the recommendations work well for teams of 4‑6 students, scaling to larger, distributed teams may require additional coordination mechanisms.
• Future Directions: The author suggests extending the research to remote/online Scrum courses, integrating security‑focused QA topics, and conducting controlled experiments that compare different QA toolchains across diverse student populations.

Authors

• Maria Spichkova

Paper Information

• arXiv ID: 2512.23488v1
• Categories: cs.SE
• Published: December 29, 2025
• PDF: Download PDF