Research

Using a RAG-enhanced large language model in a virtual teaching assistant role: Experiences from a pilot project in statistics education

Hungarian Statistical Review, Volume 7 (2024)

Renáta Németh – Annamária Tátrai – Miklós Szabó – Árpád Tamási

The role of artificial intelligence (AI) in education is expected to grow, but how it transforms teaching and learning remains unclear. This study explores the use of an AI tutor that is similar to ChatGPT enhanced with retrieval-augmented generation (RAG), in a pilot project at the Faculty of Social Sciences of Eötvös Loránd University in Budapest, Hungary. The tutor provided a searchable knowledge base for students preparing for admission to the MSc in Survey Statistics and Data Analytics. Instructor feedback highlighted the tutor’s ability to deliver accurate, textbook-based responses, but noted limitations in addressing real-world complexities. Student feedback, which was gathered through focus groups and surveys, showed high satisfaction and many used the tool for active learning such as comparing concepts and organising material.

Exploring the Use of Retrieval-Augmented Generation Models in Higher Education

Social Sciences & Humanities Open

Renáta Németh – Annamária Tátrai – Miklós Szabó – Péter Tibor Zaletnyik – Árpád Tamási

The role of artificial intelligence in education is growing, but its impact on teaching and learning remains unclear. This paper examines an artificial intelligence tutor like ChatGPT, enhanced with retrieval-augmented generation, in a pilot project at two universities. Four courses from three programs with different learning objectives and pedagogical methods were involved to provide as diverse a context as possible. We created an artificial intelligence tutor for each course, capable of answering students’ questions by referring to the educational resources provided. Both students and lecturers reported positive experiences. Qualitative analysis of the chats shows a high level of engagement and motivation. Based on expert coding of a random sample, 1.5% of answers were incorrect and 16.5% were outside the context provided to the large language model. We found that retrieval-augmented generation reduces hallucinations for topics that are sufficiently explained in the material. A comparison of different courses shows the same tool can work differently in different contexts. Our results show that the artificial intelligence tutor does not simply impart knowledge, but mediates the relationship between student, lecturer, and course material. This mediation can facilitate learning, but it can also pose new challenges.

Anthropomorphization of AI Tutors in Educational Contexts: Insights from Interdisciplinary University Courses

ELTE Faculty of Social Science and University of Pécs Faculty of Humanities

This ongoing research examines the anthropomorphization of AI-based educational assistants by students in higher education. The study is grounded in the analysis of three university courses conducted during the fall semester of the 2024–2025 academic year, covering distinct disciplinary fields: theoretical social sciences, applied social sciences, and IT system architectures. The project is a collaborative effort between Renáta Németh (ELTE Faculty of Social Sciences, Department of Statistics), Viktor Berger (University of Pécs, Faculty of Humanities), and Miklós Szabó (ELTE Faculty of Social Sciences, Department of Minority Sociology). All three courses integrated the GoSchool.ai platform—a retrieval-augmented, course-constrained AI system designed to assist students with course-related tasks. The research explores how students interact with the AI assistant, with particular attention to moments when human-like attributes are projected onto the system, such as perceived intentionality, understanding, or personality. Through the collection and analysis of anonymized interaction logs, course context data, and student reflections, the study aims to contribute to a deeper understanding of human-AI relational dynamics in educational settings.

AI-Enhanced Learning in Practice

Oslo Metropolitan University Department of Socail Work, Child Welfare

This ongoing study investigates the pedagogical and experiential implications of integrating GoSchool.ai—an AI-powered educational assistant—into higher education at Oslo Metropolitan University. The platform was piloted during the 2024–2025 academic year in both a BA-level and an MA-level course, in collaboration with Blanka Støren-Vaczy and Mats Eirik Lillehagen. This research explores how students interact with the assistant in support of learning and assignment completion, and how the system may shape classroom dynamics, expectations of knowledge. Preliminary observations point to student engagement with the platform and indicate that structured support and training may be necessary to maximize effective use. From the instructional side, the system’s chat logs provide insight into student comprehension, while its integration into seminars and lectures opens pathways for pedagogical experimentation. The project contributes to ongoing discussions around AI literacy, instructional design, and the role of intelligent systems in academic environments.

Investigating the Educational Integration of Generative AI

Interdisciplinary Social Research Doctoral Program, Faculty of Social Sciences, Eötvös Loránd University (ELTE)

Funded PhD Researcher in Interdisciplinary Social Research by Árpád Tamási

Using GoSchool.ai as a live testbed, we focus on understanding the impact of AI-based tutoring on student learning, teaching roles, and institutional practices. How does AI-supported tutoring affect student performance, self-regulation, and cognitive engagement?
In what ways do teaching roles and educator competencies evolve when AI tools are introduced? What social, infrastructural, and individual factors influence students’ and instructors’ willingness to adopt generative AI systems (based on TAM and UTAUT-2 models)?
Which design choices reduce “metacognitive laziness” and support sustained, reflective learning? Our methodology is grounded in Design-Based Research (DBR), structured into short, iterative micro-cycles of design → intervention → analysis → reflection. By involving students and instructors across various fields (e.g., statistics, informatics, law, anthropology), the study aims to generate both theoretical insights and actionable recommendations for AI-enhanced pedagogy.