@inproceedings{bn-etal-2025-real,
title = "How Real Are Synthetic Therapy Conversations? Evaluating Fidelity in Prolonged Exposure Dialogues",
author = "Bn, Suhas and
Mattioli, Dominik O. and
Sherrill, Andrew M. and
Arriaga, Rosa I. and
Wiese, Christopher and
Abdullah, Saeed",
editor = "Christodoulopoulos, Christos and
Chakraborty, Tanmoy and
Rose, Carolyn and
Peng, Violet",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
month = nov,
year = "2025",
address = "Suzhou, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-emnlp.1144/",
doi = "10.18653/v1/2025.findings-emnlp.1144",
pages = "20986--20995",
ISBN = "979-8-89176-335-7",
abstract = "Synthetic data adoption in healthcare is driven by privacy concerns, data access limitations, and high annotation costs. We explore synthetic Prolonged Exposure (PE) therapy conversations for PTSD as a scalable alternative for training clinical models. We systematically compare real and synthetic dialogues using linguistic, structural, and protocol-specific metrics like turn-taking and treatment fidelity. We introduce and evaluate PE-specific metrics, offering a novel framework for assessing clinical fidelity beyond surface fluency. Our findings show that while synthetic data successfully mitigates data scarcity and protects privacy, capturing the most subtle therapeutic dynamics remains a complex challenge. Synthetic dialogues successfully replicate key linguistic features of real conversations, for instance, achieving a similar Readability Score (89.2 vs. 88.1), while showing differences in some key fidelity markers like distress monitoring. This comparison highlights the need for fidelity-aware metrics that go beyond surface fluency to identify clinically significant nuances. Our model-agnostic framework is a critical tool for developers and clinicians to benchmark generative model fidelity before deployment in sensitive applications. Our findings help clarify where synthetic data can effectively complement real-world datasets, while also identifying areas for future refinement."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="bn-etal-2025-real">
<titleInfo>
<title>How Real Are Synthetic Therapy Conversations? Evaluating Fidelity in Prolonged Exposure Dialogues</title>
</titleInfo>
<name type="personal">
<namePart type="given">Suhas</namePart>
<namePart type="family">Bn</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dominik</namePart>
<namePart type="given">O</namePart>
<namePart type="family">Mattioli</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Andrew</namePart>
<namePart type="given">M</namePart>
<namePart type="family">Sherrill</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rosa</namePart>
<namePart type="given">I</namePart>
<namePart type="family">Arriaga</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Christopher</namePart>
<namePart type="family">Wiese</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Saeed</namePart>
<namePart type="family">Abdullah</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Christos</namePart>
<namePart type="family">Christodoulopoulos</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tanmoy</namePart>
<namePart type="family">Chakraborty</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Carolyn</namePart>
<namePart type="family">Rose</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Violet</namePart>
<namePart type="family">Peng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Suzhou, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-335-7</identifier>
</relatedItem>
<abstract>Synthetic data adoption in healthcare is driven by privacy concerns, data access limitations, and high annotation costs. We explore synthetic Prolonged Exposure (PE) therapy conversations for PTSD as a scalable alternative for training clinical models. We systematically compare real and synthetic dialogues using linguistic, structural, and protocol-specific metrics like turn-taking and treatment fidelity. We introduce and evaluate PE-specific metrics, offering a novel framework for assessing clinical fidelity beyond surface fluency. Our findings show that while synthetic data successfully mitigates data scarcity and protects privacy, capturing the most subtle therapeutic dynamics remains a complex challenge. Synthetic dialogues successfully replicate key linguistic features of real conversations, for instance, achieving a similar Readability Score (89.2 vs. 88.1), while showing differences in some key fidelity markers like distress monitoring. This comparison highlights the need for fidelity-aware metrics that go beyond surface fluency to identify clinically significant nuances. Our model-agnostic framework is a critical tool for developers and clinicians to benchmark generative model fidelity before deployment in sensitive applications. Our findings help clarify where synthetic data can effectively complement real-world datasets, while also identifying areas for future refinement.</abstract>
<identifier type="citekey">bn-etal-2025-real</identifier>
<identifier type="doi">10.18653/v1/2025.findings-emnlp.1144</identifier>
<location>
<url>https://aclanthology.org/2025.findings-emnlp.1144/</url>
</location>
<part>
<date>2025-11</date>
<extent unit="page">
<start>20986</start>
<end>20995</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T How Real Are Synthetic Therapy Conversations? Evaluating Fidelity in Prolonged Exposure Dialogues
%A Bn, Suhas
%A Mattioli, Dominik O.
%A Sherrill, Andrew M.
%A Arriaga, Rosa I.
%A Wiese, Christopher
%A Abdullah, Saeed
%Y Christodoulopoulos, Christos
%Y Chakraborty, Tanmoy
%Y Rose, Carolyn
%Y Peng, Violet
%S Findings of the Association for Computational Linguistics: EMNLP 2025
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%@ 979-8-89176-335-7
%F bn-etal-2025-real
%X Synthetic data adoption in healthcare is driven by privacy concerns, data access limitations, and high annotation costs. We explore synthetic Prolonged Exposure (PE) therapy conversations for PTSD as a scalable alternative for training clinical models. We systematically compare real and synthetic dialogues using linguistic, structural, and protocol-specific metrics like turn-taking and treatment fidelity. We introduce and evaluate PE-specific metrics, offering a novel framework for assessing clinical fidelity beyond surface fluency. Our findings show that while synthetic data successfully mitigates data scarcity and protects privacy, capturing the most subtle therapeutic dynamics remains a complex challenge. Synthetic dialogues successfully replicate key linguistic features of real conversations, for instance, achieving a similar Readability Score (89.2 vs. 88.1), while showing differences in some key fidelity markers like distress monitoring. This comparison highlights the need for fidelity-aware metrics that go beyond surface fluency to identify clinically significant nuances. Our model-agnostic framework is a critical tool for developers and clinicians to benchmark generative model fidelity before deployment in sensitive applications. Our findings help clarify where synthetic data can effectively complement real-world datasets, while also identifying areas for future refinement.
%R 10.18653/v1/2025.findings-emnlp.1144
%U https://aclanthology.org/2025.findings-emnlp.1144/
%U https://doi.org/10.18653/v1/2025.findings-emnlp.1144
%P 20986-20995
Markdown (Informal)
[How Real Are Synthetic Therapy Conversations? Evaluating Fidelity in Prolonged Exposure Dialogues](https://aclanthology.org/2025.findings-emnlp.1144/) (Bn et al., Findings 2025)
ACL