Abstract:

This paper presents a teaching experience conducted in a third-year course “Reactive Systems” in the Artificial Intelligence degree (n = 39 students) where quizzes were used as a learning tool to support the study of the Go programming language. The quizzes were designed to expose students to programming syntax and to improve code reading fluency, complementing subsequent project-based work in the course rather than serving as a formal assessment instrument.

A mastery-based approach was adopted. Students progressed through a sequence of quiz levels and were allowed to repeat each test several times. Advancement to the next level required obtaining a score of at least 9 out of 10. This rule encouraged students to review the material and repeat the activity until the underlying concept was well understood.

Generative AI was used to produce sets of related questions for each level. Quiz generation was performed using the OpenAI ChatGPT 5.2 model and all items were reviewed by instructors to ensure correctness, proper GIFT formatting, and to avoid topic repetition bias. The system created new variants of the quizzes while preserving the conceptual focus of the exercise. This allowed the generation of chained quizzes covering different topics in Go programming, including basic syntax, control flow, data structures and introductory concurrency patterns. The approach made it possible to provide repeated practice without presenting identical questions in each attempt.

Student perception was collected through a Likert-scale questionnaire including items on feedback usefulness, autonomy and perceived learnin. Results show a consistent positive perception of the activity, especially in feedback quality, progressive structure, and perceived improvement in code understanding, while time constraints were identified as the main limitation.

Keywords:

Generative AI, mastery learning, programming education, Go programming language, quiz-based learning.