GenAI-Driven Game and Viva Assessments for Software Engineering
Year: 2026
Type: conference-paper
Abstract: Generative AI (GenAI) is rapidly reshaping computing education. This paper reports multi-course experiences integrating GenAI into undergraduate and graduate software engineering curricula, including requirements engineering and agile software development. Redesigning assessments—such as timed oral micro-presentations and quiz-bowl-style vivas—to promote authentic learning while reducing AI-assisted plagiarism. Across two academic years, students demonstrated strong conceptual mastery in oral assessments (≈80% earned B or higher) and moderate performance on calibrated take-home essay quizzes (average ≈65%). Classroom applications included AI-assisted persona generation, requirements elicitation, SRS critique, and agent-supported implementation. Findings indicate that thoughtfully designed assessments can leverage GenAI for practice and feedback while preserving rigorous evaluation. The study also discusses human-centric learning, threats to validity, ethics, and design principles for scalable adoption, presenting a framework for integrating GenAI into software engineering education. The redesigns foster sustainable innovation competencies—systems thinking, stakeholder empathy, and risk-aware decision-making—while modeling ethical leadership through transparent AI use and accountable oral defenses.
Author: Soma Datta
Institutions: University of Houston, University of Houston - Clear Lake
Language: English
Subfield: Education
Field: Social Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: hybrid
OpenAlex ID: W7166544403