Designing a Lightweight AI Tutor for Pedagogically Grounded Classroom Use
Erica Perseghin;Gian Luca Foresti
2026-01-01
File in questo prodotto:
| File | Dimensione | Formato | |
|---|---|---|---|
|
paper_1456 (2).pdf
non disponibili
Descrizione: This doctoral research investigates how curriculum-based generative AI (GenAI) tutors can be designed for responsible classroom integration in secondary education. Adopting a design-based research (DBR) approach, the study builds on prior work on teachers’ and students’ AI literacy to inform the iterative design of a lightweight Retrieval-Augmented Generation (RAG) tutor guided by three theory-informed principles: curriculum alignment, delayed feedback as temporal scaffolding, and explicit signaling of system limitations. An exploratory classroom pilot (n = 10, ages 18–19) in SQL learning provides preliminary indications that deferring responses creates temporal space for peer discussion, while content grounding and limitation signaling may support instructional coherence and more reflective engagement with AI. The expected contribution is empirically grounded design knowledge on how GenAI tutors can function as pedagogical structures that support collaborative sensemaking, informing research and practice in AI in Education (AIEd).
Tipologia:
Abstract
Licenza:
Non pubblico
Dimensione
464.33 kB
Formato
Adobe PDF
|
464.33 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
