@inproceedings{bediwy-mojoyinola-2026-validating,
title = "Validating {LLM}-Rated Item Features for Explanatory Item Response Models",
author = "Bediwy, Ahmed H. and
Mojoyinola, Mubarak O.",
editor = "Wilson, Joshua and
Ormerod, Christopher and
Beiting-Parrish, Magdalen",
booktitle = "Proceedings of the Artificial Intelligence in Measurement and Education Conference ({AIME}-Con): Works in Progress",
month = oct,
year = "2026",
address = "Wyndham Grand Pittsburgh Downtown, Pittsburgh, Pennsylvania, United States",
publisher = "National Council on Measurement in Education (NCME)",
url = "https://aclanthology.org/2026.aimecon-wip.49/",
pages = "382--388",
ISBN = "979-8-9983004-1-7",
abstract = "Explanatory item response models let test de- velopers anticipate item difficulty from item design, but they depend on subject-matter ex- perts rating every item on every hypothesized feature{---}a step that is slow, costly, and the practical bottleneck limiting how many items can be modeled. We ask whether large lan- guage models can supply those ratings. Four trained experts and six LLMs independently rated 15 grade-5 mathematics items on six psy- chometric features. We evaluate the LLM rat- ings twice: against the human consensus using quadratic-weighted $\kappa$and mixed-effects mod- els, and against 1,452 student responses by us- ing each source{'}s ratings as the design matrix of a linear logistic test model (LLTM) bench- marked against a Rasch baseline. Claude Opus 4.7 performed best in recovering the item dif- ficulty (r = 0.89), however, the human con- sensus against which agreement is measured performs worst at recovering Rasch difficulty (r = 0.33) compared to the rest of the LLMs. The reason is visible in the expert panel it- self: inter-rater Fleiss' $\kappa$is at or below chance on three of six features, so the consensus is a noisy reference rather than ground truth. We also identify two failure modes (zero-variance and perfect collinearity) that agreement statis- tics cannot detect but make a feature unusable as an LLTM covariate. We argue that agree- ment should be reported alongside criterion- referenced recovery of item parameters, not in place of it."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="bediwy-mojoyinola-2026-validating">
<titleInfo>
<title>Validating LLM-Rated Item Features for Explanatory Item Response Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ahmed</namePart>
<namePart type="given">H</namePart>
<namePart type="family">Bediwy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mubarak</namePart>
<namePart type="given">O</namePart>
<namePart type="family">Mojoyinola</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Artificial Intelligence in Measurement and Education Conference (AIME-Con): Works in Progress</title>
</titleInfo>
<name type="personal">
<namePart type="given">Joshua</namePart>
<namePart type="family">Wilson</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Christopher</namePart>
<namePart type="family">Ormerod</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Magdalen</namePart>
<namePart type="family">Beiting-Parrish</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>National Council on Measurement in Education (NCME)</publisher>
<place>
<placeTerm type="text">Wyndham Grand Pittsburgh Downtown, Pittsburgh, Pennsylvania, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-9983004-1-7</identifier>
</relatedItem>
<abstract>Explanatory item response models let test de- velopers anticipate item difficulty from item design, but they depend on subject-matter ex- perts rating every item on every hypothesized feature—a step that is slow, costly, and the practical bottleneck limiting how many items can be modeled. We ask whether large lan- guage models can supply those ratings. Four trained experts and six LLMs independently rated 15 grade-5 mathematics items on six psy- chometric features. We evaluate the LLM rat- ings twice: against the human consensus using quadratic-weighted ąppaand mixed-effects mod- els, and against 1,452 student responses by us- ing each source’s ratings as the design matrix of a linear logistic test model (LLTM) bench- marked against a Rasch baseline. Claude Opus 4.7 performed best in recovering the item dif- ficulty (r = 0.89), however, the human con- sensus against which agreement is measured performs worst at recovering Rasch difficulty (r = 0.33) compared to the rest of the LLMs. The reason is visible in the expert panel it- self: inter-rater Fleiss’ ąppais at or below chance on three of six features, so the consensus is a noisy reference rather than ground truth. We also identify two failure modes (zero-variance and perfect collinearity) that agreement statis- tics cannot detect but make a feature unusable as an LLTM covariate. We argue that agree- ment should be reported alongside criterion- referenced recovery of item parameters, not in place of it.</abstract>
<identifier type="citekey">bediwy-mojoyinola-2026-validating</identifier>
<location>
<url>https://aclanthology.org/2026.aimecon-wip.49/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>382</start>
<end>388</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Validating LLM-Rated Item Features for Explanatory Item Response Models
%A Bediwy, Ahmed H.
%A Mojoyinola, Mubarak O.
%Y Wilson, Joshua
%Y Ormerod, Christopher
%Y Beiting-Parrish, Magdalen
%S Proceedings of the Artificial Intelligence in Measurement and Education Conference (AIME-Con): Works in Progress
%D 2026
%8 October
%I National Council on Measurement in Education (NCME)
%C Wyndham Grand Pittsburgh Downtown, Pittsburgh, Pennsylvania, United States
%@ 979-8-9983004-1-7
%F bediwy-mojoyinola-2026-validating
%X Explanatory item response models let test de- velopers anticipate item difficulty from item design, but they depend on subject-matter ex- perts rating every item on every hypothesized feature—a step that is slow, costly, and the practical bottleneck limiting how many items can be modeled. We ask whether large lan- guage models can supply those ratings. Four trained experts and six LLMs independently rated 15 grade-5 mathematics items on six psy- chometric features. We evaluate the LLM rat- ings twice: against the human consensus using quadratic-weighted ąppaand mixed-effects mod- els, and against 1,452 student responses by us- ing each source’s ratings as the design matrix of a linear logistic test model (LLTM) bench- marked against a Rasch baseline. Claude Opus 4.7 performed best in recovering the item dif- ficulty (r = 0.89), however, the human con- sensus against which agreement is measured performs worst at recovering Rasch difficulty (r = 0.33) compared to the rest of the LLMs. The reason is visible in the expert panel it- self: inter-rater Fleiss’ ąppais at or below chance on three of six features, so the consensus is a noisy reference rather than ground truth. We also identify two failure modes (zero-variance and perfect collinearity) that agreement statis- tics cannot detect but make a feature unusable as an LLTM covariate. We argue that agree- ment should be reported alongside criterion- referenced recovery of item parameters, not in place of it.
%U https://aclanthology.org/2026.aimecon-wip.49/
%P 382-388
Markdown (Informal)
[Validating LLM-Rated Item Features for Explanatory Item Response Models](https://aclanthology.org/2026.aimecon-wip.49/) (Bediwy & Mojoyinola, AIME-Con 2026)
ACL
- Ahmed H. Bediwy and Mubarak O. Mojoyinola. 2026. Validating LLM-Rated Item Features for Explanatory Item Response Models. In Proceedings of the Artificial Intelligence in Measurement and Education Conference (AIME-Con): Works in Progress, pages 382–388, Wyndham Grand Pittsburgh Downtown, Pittsburgh, Pennsylvania, United States. National Council on Measurement in Education (NCME).