@inproceedings{lior-etal-2026-comparing,
title = "Comparing the Framing Effect in Humans and {LLM}s on Naturally Occurring Texts",
author = "Lior, Gili and
Naccache, Liron and
Stanovsky, Gabriel",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.41/",
doi = "10.18653/2026.findings-emnlp.41",
pages = "825--841",
ISBN = "979-8-89176-462-0",
abstract = "Humans are influenced by how information is presented, a phenomenon known as the framing effect. Prior work suggests that LLMs may also be susceptible to framing, but it has relied on synthetic data and did not compare to human behavior. To address this gap, we introduce WildFrame - a dataset for evaluating LLM responses to positive and negative framing in naturally-occurring sentences, alongside human responses on the same data. WildFrame consists of 1,000 real-world texts selected to convey a clear sentiment; we then reframe each text in either a positive or negative light and collect human sentiment annotations. Evaluating eleven LLMs on WildFrame, we find that all models respond to reframing in a human-like manner ($r\geq 0.52$), and that both humans and models are influenced more by positive than negative reframing. Notably, GPT models are the least correlated with human behavior among all tested models. These findings raise a discussion around the goals of state-of-the-art LLM development and whether models should align closely with human behavior, to preserve cognitive phenomena such as the framing effect, or instead mitigate such biases in favor of fairness and consistency."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="lior-etal-2026-comparing">
<titleInfo>
<title>Comparing the Framing Effect in Humans and LLMs on Naturally Occurring Texts</title>
</titleInfo>
<name type="personal">
<namePart type="given">Gili</namePart>
<namePart type="family">Lior</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Liron</namePart>
<namePart type="family">Naccache</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gabriel</namePart>
<namePart type="family">Stanovsky</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>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Humans are influenced by how information is presented, a phenomenon known as the framing effect. Prior work suggests that LLMs may also be susceptible to framing, but it has relied on synthetic data and did not compare to human behavior. To address this gap, we introduce WildFrame - a dataset for evaluating LLM responses to positive and negative framing in naturally-occurring sentences, alongside human responses on the same data. WildFrame consists of 1,000 real-world texts selected to convey a clear sentiment; we then reframe each text in either a positive or negative light and collect human sentiment annotations. Evaluating eleven LLMs on WildFrame, we find that all models respond to reframing in a human-like manner (r\geq 0.52), and that both humans and models are influenced more by positive than negative reframing. Notably, GPT models are the least correlated with human behavior among all tested models. These findings raise a discussion around the goals of state-of-the-art LLM development and whether models should align closely with human behavior, to preserve cognitive phenomena such as the framing effect, or instead mitigate such biases in favor of fairness and consistency.</abstract>
<identifier type="citekey">lior-etal-2026-comparing</identifier>
<identifier type="doi">10.18653/2026.findings-emnlp.41</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.41/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>825</start>
<end>841</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Comparing the Framing Effect in Humans and LLMs on Naturally Occurring Texts
%A Lior, Gili
%A Naccache, Liron
%A Stanovsky, Gabriel
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F lior-etal-2026-comparing
%X Humans are influenced by how information is presented, a phenomenon known as the framing effect. Prior work suggests that LLMs may also be susceptible to framing, but it has relied on synthetic data and did not compare to human behavior. To address this gap, we introduce WildFrame - a dataset for evaluating LLM responses to positive and negative framing in naturally-occurring sentences, alongside human responses on the same data. WildFrame consists of 1,000 real-world texts selected to convey a clear sentiment; we then reframe each text in either a positive or negative light and collect human sentiment annotations. Evaluating eleven LLMs on WildFrame, we find that all models respond to reframing in a human-like manner (r\geq 0.52), and that both humans and models are influenced more by positive than negative reframing. Notably, GPT models are the least correlated with human behavior among all tested models. These findings raise a discussion around the goals of state-of-the-art LLM development and whether models should align closely with human behavior, to preserve cognitive phenomena such as the framing effect, or instead mitigate such biases in favor of fairness and consistency.
%R 10.18653/2026.findings-emnlp.41
%U https://aclanthology.org/2026.findings-emnlp.41/
%U https://doi.org/10.18653/2026.findings-emnlp.41
%P 825-841
Markdown (Informal)
[Comparing the Framing Effect in Humans and LLMs on Naturally Occurring Texts](https://aclanthology.org/2026.findings-emnlp.41/) (Lior et al., Findings 2026)
ACL