Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

ALBA : Adaptive Language-Based Assessments for Mental Health

Varadarajan, Vasudha LU ; Sikström, Sverker LU orcid ; Kjell, Oscar N.E. LU orcid and Schwartz, H. Andrew (2024) 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL 2024 1. p.2466-2478
Abstract

Mental health issues differ widely among individuals, with varied signs and symptoms. Recently, language-based assessments have shown promise in capturing this diversity, but they require a substantial sample of words per person for accuracy. This work introduces the task of Adaptive Language-Based Assessment (ALBA), which involves adaptively ordering questions while also scoring an individual’s latent psychological trait using limited language responses to previous questions. To this end, we develop adaptive testing methods under two psychometric measurement theories: Classical Test Theory and Item Response Theory. We empirically evaluate ordering and scoring strategies, organizing into two new methods: a semi-supervised item response... (More)

Mental health issues differ widely among individuals, with varied signs and symptoms. Recently, language-based assessments have shown promise in capturing this diversity, but they require a substantial sample of words per person for accuracy. This work introduces the task of Adaptive Language-Based Assessment (ALBA), which involves adaptively ordering questions while also scoring an individual’s latent psychological trait using limited language responses to previous questions. To this end, we develop adaptive testing methods under two psychometric measurement theories: Classical Test Theory and Item Response Theory. We empirically evaluate ordering and scoring strategies, organizing into two new methods: a semi-supervised item response theory-based method (ALIRT) and a supervised Actor-Critic model. While we found both methods to improve over non-adaptive baselines, We found ALIRT to be the most accurate and scalable, achieving the highest accuracy with fewer questions (e.g., Pearson r ≈ 0.93 after only 3 questions as compared to typically needing at least 7 questions). In general, adaptive language-based assessments of depression and anxiety were able to utilize a smaller sample of language without compromising validity or large computational costs.

(Less)
Please use this url to cite or link to this publication:
author
; ; and
organization
publishing date
type
Chapter in Book/Report/Conference proceeding
publication status
published
subject
host publication
Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies
editor
Duh, Kevin ; Gomez, Helena and Bethard, Steven
volume
1
pages
13 pages
publisher
Association for Computational Linguistics (ACL)
conference name
2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL 2024
conference location
Hybrid, Mexico City, Mexico
conference dates
2024-06-16 - 2024-06-21
external identifiers
  • pmid:40093858
  • scopus:85200248617
ISBN
9798891761148
DOI
10.18653/v1/2024.naacl-long.136
language
English
LU publication?
yes
id
d2ec26f2-9f35-4bed-8785-72fdb46248bd
date added to LUP
2024-09-17 11:26:34
date last changed
2025-07-09 15:06:46
@inproceedings{d2ec26f2-9f35-4bed-8785-72fdb46248bd,
  abstract     = {{<p>Mental health issues differ widely among individuals, with varied signs and symptoms. Recently, language-based assessments have shown promise in capturing this diversity, but they require a substantial sample of words per person for accuracy. This work introduces the task of Adaptive Language-Based Assessment (ALBA), which involves adaptively ordering questions while also scoring an individual’s latent psychological trait using limited language responses to previous questions. To this end, we develop adaptive testing methods under two psychometric measurement theories: Classical Test Theory and Item Response Theory. We empirically evaluate ordering and scoring strategies, organizing into two new methods: a semi-supervised item response theory-based method (ALIRT) and a supervised Actor-Critic model. While we found both methods to improve over non-adaptive baselines, We found ALIRT to be the most accurate and scalable, achieving the highest accuracy with fewer questions (e.g., Pearson r ≈ 0.93 after only 3 questions as compared to typically needing at least 7 questions). In general, adaptive language-based assessments of depression and anxiety were able to utilize a smaller sample of language without compromising validity or large computational costs.</p>}},
  author       = {{Varadarajan, Vasudha and Sikström, Sverker and Kjell, Oscar N.E. and Schwartz, H. Andrew}},
  booktitle    = {{Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies}},
  editor       = {{Duh, Kevin and Gomez, Helena and Bethard, Steven}},
  isbn         = {{9798891761148}},
  language     = {{eng}},
  pages        = {{2466--2478}},
  publisher    = {{Association for Computational Linguistics (ACL)}},
  title        = {{ALBA : Adaptive Language-Based Assessments for Mental Health}},
  url          = {{http://dx.doi.org/10.18653/v1/2024.naacl-long.136}},
  doi          = {{10.18653/v1/2024.naacl-long.136}},
  volume       = {{1}},
  year         = {{2024}},
}