Learning Context Matters: Measuring and Diagnosing Personalization Gaps in LLM-Based Instructional Design

Johaun Hatchett, Debshila Basu Mallick, Brittany C. Bradford, Richard G. Baraniuk · 2026

Read Learning Context Matters: Measuring and Diagnosing Personalization Gaps in LLM-Based Instructional Design by Johaun Hatchett, Debshila Basu Mallick, Brittany C. Bradford, Richard G. Baraniuk (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-04972.txt.

Description

Language models are entering classrooms with the promise of personal support for each learner. The paper comes from Johaun Hatchett and 3 co-authors. Real personalization, though, needs learning context: who the student is, what they are trying to grasp, and how they take in the material.

The team built a framework that compares tutoring decisions made with and without that context, using synthetic learner profiles grounded in measurement science. Giving the model context did shift its choices toward those of subject experts, but a large gap remained. A relevance-impact analysis then showed which learner traits the model used, ignored, or wrongly leaned on. The authors argue context shapes planning without guaranteeing sound teaching, and they offer their framework as a basis for principled evaluation.

Catalog record

CREATORJohaun Hatchett, Debshila Basu Mallick, Brittany C. Bradford, Richard G. Baraniuk
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE32 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602-04972

Browse the full library · About the Tower of Babel