AI Literacy & Learner Modeling at colaps

By irenejob opportunities, news
AI-generated image that depicts a dialogue between a human and an LLM

We are excited to open a new PhD position at colaps at the University of Duisburg-Essen (UDE) at the intersection of learner modeling, psychology, philosophy and generative AI.

What’s the project about?

The position is part of MORAL-ALIGN, a DFG-funded project in the context of the DFG Priority Programme 2573 “Rethinking Disinformation” (Re:DIS) exploring how generative AI systems can gradually shift their moral framing to match a user and how that shift can make misinformation more persuasive. Most work on this topic treats “AI literacy” as a fixed, pre-measured trait. This PhD project asks the question: what if AI literacy isn’t static, but something that develops (or erodes) through repeated interaction with an LLM?

The successful candidate will build on the learner modeling paradigm from Intelligent Tutoring Systems and Learning Analytics (think Bayesian Knowledge Tracing and Deep Knowledge Tracing) and apply it to a brand-new target: jointly tracking a person’s moral profile, their AI competencies, and their susceptibility to morally-oriented misinformation as they evolve over an extended human-LLM dialogue.

The result will be a new computational framework that updates its estimate of a user’s AI literacy and vulnerability to misinformation after every round of conversation.

Why this project matters

Three gaps motivate the work:

  • Static measurement: most AI literacy questionnaires are one-shot snapshots that can’t capture how people change over multi-day interactions.
  • No process model of susceptibility: we know what makes people vulnerable to disinformation, but not how those factors evolve dynamically across a conversation.
  • No computational bridge: learner modeling techniques were built for education, and have never been adapted to jointly capture moral susceptibility and AI competence growth together.

This project sits right in that gap, combining Moral Foundations Theory with learner modeling to build something new.

What you’ll be doing

  • Designing and calibrating the framework
  • Running empirical studies on human-LLM dialogue behavior
  • Prototyping an adaptive scaffolding intervention to boost AI literacy in real time
  • Publishing at venues like LAK, CHI, or IJAIED
  • Getting involved in teaching, conference organization, and the broader research community

Who we’re looking for

  • A strong university degree (min. 8 semesters) in computer science or a related field
  • Excellent English
  • Basic research methods know-how and some programming experience
  • Curiosity about the messy overlap of AI, ethics, and learning

The details

Applications (CV, cover letter, and a 2–3 page project proposal) go to Prof. Dr. Irene-Angelica Chounta. Full details and the application outline are in the official posting.

If you’re excited about building AI systems that understand how people grow in their relationship with them — not just what they know at a single point in time — we’d love to hear from you!

References

Atari, M., Haidt, J., Graham, J., Koleva, S., Stevens, S. T., & Dehghani, M. (2023). Morality beyond the WEIRD: How the nomological network of morality varies across cultures. Journal of Personality and Social Psychology, 125(5), 1157–1188. https://doi.org/10.1037/pspp0000470

Corbett, A. T., & Anderson, J. R. (1994). Knowledge tracing: Modeling the acquisition of procedural knowledge. User Modeling and User-Adapted Interaction, 4(4), 253–278.

Effron, D. A., & Helgason, B. A. (2022). The moral psychology of misinformation: Why we excuse dishonesty in a post-truth world. Current Opinion in Psychology, 47, 101375.

Ng, D. T. K., Wu, W., Leung, J. K. L., Chiu, T. K. F., & Chu, S. K. W. (2024). Design and validation of the AI literacy questionnaire. British Journal of Educational Technology, 55(3), 1082–1104.