Saved in:
Bibliographic Details
Main Authors: Park, Junsoo, Medhat, Youssef, Wai, Htet Phyo, Thajchayapong, Ploy, Goel, Ashok K.
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.05848
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911573652537344
author Park, Junsoo
Medhat, Youssef
Wai, Htet Phyo
Thajchayapong, Ploy
Goel, Ashok K.
author_facet Park, Junsoo
Medhat, Youssef
Wai, Htet Phyo
Thajchayapong, Ploy
Goel, Ashok K.
contents Learner representations play a central role in educational AI systems, yet it is often unclear whether they preserve meaningful differences between students when instructional outcomes are unavailable or highly context-dependent. This work examines how to evaluate learner representations based on whether they retain separation between learners under a shared comparison rule. We introduce distinctiveness, a representation-level measure that evaluates how each learner differs from others in the cohort using pairwise distances, without requiring clustering, labels, or task-specific evaluation. Using student-authored questions collected through a conversational AI agent in an online learning environment, we compare representations based on individual questions with representations that aggregate patterns across a student's interactions over time. Results show that learner-level representations yield higher separation, stronger clustering structure, and more reliable pairwise discrimination than interaction-level representations. These findings demonstrate that learner representations can be evaluated independently of instructional outcomes and provide a practical pre-deployment criterion using distinctiveness as a diagnostic metric for assessing whether a representation supports differentiated modeling or personalization.
format Preprint
id arxiv_https___arxiv_org_abs_2604_05848
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Evaluating Learner Representations for Differentiation Prior to Instructional Outcomes
Park, Junsoo
Medhat, Youssef
Wai, Htet Phyo
Thajchayapong, Ploy
Goel, Ashok K.
Computation and Language
Artificial Intelligence
Learner representations play a central role in educational AI systems, yet it is often unclear whether they preserve meaningful differences between students when instructional outcomes are unavailable or highly context-dependent. This work examines how to evaluate learner representations based on whether they retain separation between learners under a shared comparison rule. We introduce distinctiveness, a representation-level measure that evaluates how each learner differs from others in the cohort using pairwise distances, without requiring clustering, labels, or task-specific evaluation. Using student-authored questions collected through a conversational AI agent in an online learning environment, we compare representations based on individual questions with representations that aggregate patterns across a student's interactions over time. Results show that learner-level representations yield higher separation, stronger clustering structure, and more reliable pairwise discrimination than interaction-level representations. These findings demonstrate that learner representations can be evaluated independently of instructional outcomes and provide a practical pre-deployment criterion using distinctiveness as a diagnostic metric for assessing whether a representation supports differentiated modeling or personalization.
title Evaluating Learner Representations for Differentiation Prior to Instructional Outcomes
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2604.05848