Epistemic curiosity & question-asking
What makes children and students ask good questions — and how question-asking skills can be measured, modelled, and trained in real classrooms.
Cognitive science · Human–AI interaction · Education
Rania Abdelghani
Postdoctoral researcher at the Hector Research Institute of Education Sciences and Psychology, University of Tübingen
I study how learners use AI to gain knowledge, how to better support these interactions, and how active learning strategies (e.g., curiosity and question-asking) can optimize them.
About
I am a cognitive scientist working at the intersection of education, developmental psychology, and human–AI interaction. My research asks a simple question with deep consequences: how do we keep learners motivated, and willing to invest the effort that real learning takes? The question matters more than ever at a time when instantly accessible information is reshaping our relationship with knowledge.
I study the mechanisms of epistemic curiosity and question-asking in students, and design technologies — from conversational agents to classroom interventions — that train these skills rather than replace them. I also develop methods for studying how student–AI interactions unfold: how they evolve over time, which of their features actually predict long-term learning, and how to support learners without taking away their productive struggle. Across this work I focus on the roles of metacognitive monitoring, self-regulation, question-asking, and AI literacy. My research is interdisciplinary by nature: I draw on fundamental findings from psychology and cognitive science to design controlled classroom experiments, and I use AI-supported tools to build both experimental platforms and data-analysis pipelines.
I completed my PhD in cognitive science at Inria and the University of Bordeaux (Flowers team) in 2024, in collaboration with the EdTech startup EvidenceB, advised by Dr. Pierre-Yves Oudeyer and Dr. Hélène Sauzéon, with the thesis “Guiding the minds of tomorrow: conversational agents to train curiosity and metacognition in young learners”. Along the way I was a visiting researcher in Celeste Kidd's lab at UC Berkeley, where I explored new research environments and built new collaborations. Before research, I trained as an engineer and worked as a research engineer intern in companies in both Tunisia and France.
Since November 2024 I have been a postdoctoral researcher in Prof. Dr. Kou Murayama's Motivation Science Lab at the Hector Research Institute in Tübingen, where I investigate how students regulate their learning when working with large language models.
Research themes
Metacognition & self-regulated learning
How learners monitor what they know and don't know, and how metacognitive practice turns curiosity into effective, autonomous learning.
Generative AI in education
Whether students can remain active learners with ChatGPT-like tools: measuring how they query, evaluate, and (too often) over-trust AI answers, and what that means for how learning with AI should be designed.
Publications
Full list on Google Scholar
2026
2024
Talks
Selected oral presentations
- Jul 2026The illusion of understanding: how middle-school students fail to regulate inquiry with ChatGPT — European Conference on Personality, Edinburgh
- Jun 2026From prompting to epistemic proactivity — AIED, Seoul, South Korea
- Jan 2025Question-asking in the digital classroom — Cogni'classes webinar, Paris
- Jul 2024Generative AI for promoting active learning strategies — AIED, Recife, Brazil
- Jun 2023Interactive environments for training children's curiosity — IDC, New Orleans
- Jan 2023Curiosity-driven learning through metacognition — CEU Conference on Cognitive Development, Budapest
- Mar 2022Conversational agents for training children's curiosity-driven behaviour — EAET, Cambridge · best presentation award
- Jan 2022Curiosity-driven design in sustainable learning technologies — IDC, Braga
Experience & education
Positions
- 2024–nowPostdoctoral researcher, Hector Research Institute, University of Tübingen — Motivation Science Lab (Kou Murayama)
- 2024Host lecturer, Université Côte d'Azur, Nice
- 2023Visiting researcher, Kidd Lab, UC Berkeley (host: Celeste Kidd)
- 2022–23Lecturer, University of Bordeaux
- 2020Research engineer, EvidenceB, Paris · research intern, ISIR, Paris
- 2019Research engineer intern, Therapanacea, Paris
Education
- 2024PhD, Cognitive Science, Inria / Université de Bordeaux — advisors P.-Y. Oudeyer & H. Sauzéon
- 2020MSc, Engineering Sciences, Sorbonne University, Paris
- 2019M.Eng, Intelligent Systems, INSAT, Tunis
Contact
The best way to reach me is by email:
Hector Research Institute of Education Sciences and Psychology
University of Tübingen, Europastraße 6, 72072 Tübingen, Germany