MIRA Lab, Metacognitive Instructional Regulation Agents

Epistemic Ownership
in AI‑Assisted Learning

MIRA Lab studies whether students can explain, defend, and revise the work they produce with AI. We build learning platforms that record this reasoning and deploy them in credit-bearing courses at the University of North Texas.

See the platforms Research program

From the director

I study how students keep ownership of their reasoning when AI can produce the work for them. I design and build the lab's platforms myself, and I run them in UNT courses so the evidence comes from real coursework.

Ji Hyun Yu, Ph.D.

Assistant Professor of Learning Technologies, University of North Texas

Editorial board, Educational Technology & Society and Interdisciplinary Journal of Problem-Based Learning
Facilitator, AMIS Community of Research
Lead, AI + Learning Analytics concentration, LTEC M.S.
Affiliated faculty, Data Science

Curriculum vitae Google Scholar jihyun.yu@unt.edu

Research platforms

Designed and built in the lab, and deployed in credit-bearing courses under IRB approval.

All platforms

MIRA-Agent

Asks students to explain their reasoning, and any pasted code, before their code can run.

Programming education

MIRA-Sim

AI stakeholders press students toward their positions across six policy stages.

Civic policy argumentation

MIRA-Py

Each Python task opens only after a concept check and a code check.

Introductory Python

Design mockup

MIRA-Roundtable

Small groups meet live to debate a question without a settled answer while an AI discussant argues the side no one has taken.

Group debate, in development