<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="niu-etal-2025-bias">
    <titleInfo>
        <title>Bias and Reliability in AI Safety Assessment: Multi-Facet Rasch Analysis of Human Moderators</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Chunling</namePart>
        <namePart type="family">Niu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Kelly</namePart>
        <namePart type="family">Bradley</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Biao</namePart>
        <namePart type="family">Ma</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Brian</namePart>
        <namePart type="family">Waltman</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Loren</namePart>
        <namePart type="family">Cossette</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Rui</namePart>
        <namePart type="family">Jin</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2025-10</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the Artificial Intelligence in Measurement and Education Conference (AIME-Con): Full Papers</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">979-8-218-84228-4</identifier>
    </relatedItem>
    <abstract>Using Multi-Facet Rasch Modeling on 36,400 safety ratings of AI-generated conversations, we reveal significant racial disparities (Asian 39.1%, White 28.7% detection rates) and content-specific bias patterns. Simulations show that diverse teams of 8-10 members achieve 70%+ reliability versus 62% for smaller homogeneous teams, providing evidence-based guidelines for AI-generated content moderation.</abstract>
    <identifier type="citekey">niu-etal-2025-bias</identifier>
    <location>
        <url>https://aclanthology.org/2025.aimecon-main.42/</url>
    </location>
    <part>
        <date>2025-10</date>
        <extent unit="page">
            <start>393</start>
            <end>397</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Bias and Reliability in AI Safety Assessment: Multi-Facet Rasch Analysis of Human Moderators
%A Niu, Chunling
%A Bradley, Kelly
%A Ma, Biao
%A Waltman, Brian
%A Cossette, Loren
%A Jin, Rui
%Y Wilson, Joshua
%Y Ormerod, Christopher
%Y Beiting Parrish, Magdalen
%S Proceedings of the Artificial Intelligence in Measurement and Education Conference (AIME-Con): Full Papers
%D 2025
%8 October
%I National Council on Measurement in Education (NCME)
%C Wyndham Grand Pittsburgh, Downtown, Pittsburgh, Pennsylvania, United States
%@ 979-8-218-84228-4
%F niu-etal-2025-bias
%X Using Multi-Facet Rasch Modeling on 36,400 safety ratings of AI-generated conversations, we reveal significant racial disparities (Asian 39.1%, White 28.7% detection rates) and content-specific bias patterns. Simulations show that diverse teams of 8-10 members achieve 70%+ reliability versus 62% for smaller homogeneous teams, providing evidence-based guidelines for AI-generated content moderation.
%U https://aclanthology.org/2025.aimecon-main.42/
%P 393-397

Markdown (Informal)

[Bias and Reliability in AI Safety Assessment: Multi-Facet Rasch Analysis of Human Moderators](https://aclanthology.org/2025.aimecon-main.42/) (Niu et al., AIME-Con 2025)

• Bias and Reliability in AI Safety Assessment: Multi-Facet Rasch Analysis of Human Moderators (Niu et al., AIME-Con 2025)

ACL

• Chunling Niu, Kelly Bradley, Biao Ma, Brian Waltman, Loren Cossette, and Rui Jin. 2025. Bias and Reliability in AI Safety Assessment: Multi-Facet Rasch Analysis of Human Moderators. In Proceedings of the Artificial Intelligence in Measurement and Education Conference (AIME-Con): Full Papers, pages 393–397, Wyndham Grand Pittsburgh, Downtown, Pittsburgh, Pennsylvania, United States. National Council on Measurement in Education (NCME).