The Learning Institute funds and supports research projects that aim to deepen our understanding of the learning sciences and learning technologies through interdisciplinary faculty collaboration.

Current pilot projects can be found below, to be completed by August 31, 2027. Another round of selected projects will be announced by the end of April 2026.

Human-Guided AI for Curriculum Innovation: Transforming Faculty Expertise into Ethical AI-Enhanced Learning Resources

Principal Investigator: Ryan K. Boettger, Department of Technical Communication.

This project develops a transparent, faculty-led workflow for creating ethical GenAI-enhanced learning resources in TECM 2700: Technical Writing. Drawing on existing course materials and instructor/student input, our team will build a curated corpus and use it to generate draft textbook chapters and a course-specific chatbot focused on high-impact job-searching and professional branding tasks (resumes, LinkedIn profiles and employment outlook reports). Using design-based research, we will iteratively validate AI drafts with subject-matter experts, then test usability with instructors and students — including scenario-based studies in our eye-tracking lab — to refine accuracy, clarity and bias safeguards. Outcomes include a customized textbook, a supportive chatbot and a replicable model other programs can adapt to responsibly integrate AI into curriculum development.

Empowering AI Competency in Math Education

Principal Investigator: Deborah Cockerham, Department of Learning Technologies.

As K-12 teachers aim to empower students with skills needed for future success, they must support student development of skills in AI understanding and usage. Yet teachers may lack knowledge or confidence with AI, and schools may provide few training opportunities to help teachers understand, select and implement AI tools in the classroom.
 
In this project, researchers are developing a multi-tiered AI-based professional development program for secondary math teachers and will gather data to study its impact. Expected outcomes include improved teacher confidence and competency with using AI tools in math education, effective classroom implementation of AI tools and strategies, and opportunities for teacher leadership. By empowering teachers to effectively integrate AI tools and mentor peers, the project will strengthen instructional practices and foster implementation of AI tools into secondary mathematics education.
Making Semiconductor Chip Design Accessible to Students with Disabilities

Principal Investigator: Gayatri Mehta, Department of Electrical Engineering.

Students with disabilities are natural problem solvers who navigate a world built for non-disabled people and often use technology in their everyday lives without realizing the importance of semiconductor chips, the fundamental blocks of the technology. Students with disabilities are significantly less likely to pursue post-secondary education than those with no disability, and employment rates among people with disabilities are significantly lower than those of non-disabled people. In this pilot project, we propose to develop an interactive learning framework driven by the needs of students with disabilities to make semiconductor chip design broadly accessible. Incorporating this unique perspective enhances the STEM field and engineering solutions with broader accessibility. Our interdisciplinary team will introduce fundamental concepts from semiconductor chip design at a level that requires no prior engineering background. We will present real-world, complex problems in relevant and meaningful contexts to connect the importance of engineering solutions for problems related to social science, to spark interest in STEM pathways, and help students realize how their contributions can benefit society. 

Artificial Intelligence and the Classical Entomologist: Designing Interdisciplinary Pedagogy to Revive an Essential, but Dying Art

Principal Investigator: Vanessa Macias, Department of Biological Sciences.

The formal study of insects, entomology, has long enabled scientists to pursue their fascination with the insect world and find work in public, government and academic sectors.  Modern entomology is transforming due to advancements in molecular biology and AI. However, pedagogy in entomology has not kept pace with advancing technologies, leaving the field without interested students and capable experts.  For this reason, we have designed a research internship and a set of course modules to innovate the learning environment for entomology at UNT to bolster hands-on, research-based exploration of the application of emerging computational technologies to local insect problems. We are integrating a course unit on insect identification using AI and will choose one student to participate in a summer internship to partner with the ongoing West Nile Surveillance efforts to develop a machine-learning platform to identify local mosquitoes. We expect this preliminary implementation to support the establishment of a cohesive entomology track at UNT that will contribute highly capable and competitive graduates to an important field.

Partner, Not Crutch: Designing a Metacognitive Nudge to Promote AI Co-Regulation

Principal Investigator: Ji Hyun Yu, Department of Learning Technologies.

As generative AI becomes a common tool for students, it risks acting as a "crutch" that encourages uncritical cognitive offloading. This interdisciplinary project bridges Learning Sciences and Data Science to transform AI into a true cognitive partner. We are deploying MIRA (Metacognitive Intelligence for Regulated Analytics), a custom AI agent, within a data science course. MIRA doesn't just provide answers; it enforces self-regulated learning through intentional friction points. These include a "Planning Gate," requiring students to explain their logic before receiving code, and a "Veracity Gate," where students must diagnose intentionally flawed AI outputs. Additionally, MIRA features a "Cognitive Mirror", a real-time dashboard reflecting the student's learning behaviors. By capturing detailed interaction data to train a custom BERT classifier, this pilot will deliver a scalable "metacognitive nudge" tool designed to foster productive AI co-regulation and deeper learning in higher education.