Date: Wednesday, August 13th, 2025

Time: 1:30 - 3:30 EST

Location: CODA 1308 (Cabbagetown) or Join via Zoom

Rachel Lowy

Ph.D. Student in Human-Centered Computing 

School of Interactive Computing 

Georgia Institute of Technology 

Committee 

Jennifer G. Kim (Advisor) - School of Interactive Computing, Georgia Institute of  Technology

Ashok Goel - School of Interactive Computing, Georgia Institute of  Technology

Betsy DiSalvo - School of Interactive Computing, Georgia Institute of  Technology

Gillian Hayes -  School of Information and Computer Science, University of California - Irvine

Abstract 

Inclusive Post-Secondary Education (IPSE) supports students with Intellectual and Developmental Disabilities (IDD) in accessing higher education, emphasizing the development of independence, autonomy, and self-determination. IPSE students are invited to participate in regular college courses on campus. However, the coursework is often not cognitively accessible to them, and instructors may lack the time and resources to adapt assignments for IPSE learners. This research explores how Large Language Models (LLMs), which enable rapid generation of adaptive coursework, can support assignment modification for IPSE learners. Diverging from fully automated AI systems that personalize learning by optimizing performance while sidelining student input, this proposal introduces an LLM-based system that aligns with IPSE values by prioritizing student agency and choice. Grounded in Causal Agency Theory, which emphasizes key decision-making processes essential for self-determination, the system supports students in taking causal action. The system provides adaptation recommendations that align with the Universal Design for Learning (UDL) framework to make assignments accessible. Based on the student’s decision to accept or reject these recommendations, as well as any additional input they request, the system generates a fully adapted assignment. In a semester-long deployment involving IPSE students and their academic support staff, two versions of this system differing in their approach to personalization will be evaluated. The study will assess usage patterns, perceived effectiveness, and sense of student agency in supported decision-making to understand how LLMs can balance personalization with student choice in promoting both academic performance and self-determination in inclusive higher education.

Rachel Lowy (she/her)

PhD Student, School of Interactive Computing

Georgia Institute of Technology