How digital therapeutic alliances influence the perceived helpfulness of online mental health Q&A: An explainable machine learning approach DOI Creative Commons
Yinghui Huang, Hui Liu, Maomao Chi

et al.

Digital Health, Journal Year: 2025, Volume and Issue: 11

Published: April 1, 2025

Objective This study investigates the role of digital therapeutic alliance (DTA) in predicting and explaining perceived helpfulness responses on online mental health Q&A platforms. Methods constructs a large dataset 19,682 interactions from platforms, employs natural language processing, explainable machine learning, causal inference methods to identify understand factors, particularly DTA, that influence human counselors’ questions. Results The learning-based model for demonstrated strong performance, achieving an root mean square error 0.8234 absolute percentage 22.7288%. explanatory analysis revealed peripheral path-related cues, such as counselor engagement (e.g., word count response time), had highest predictive power. Additionally, central those linked DTA—specifically emotional bonds tasks—significantly influenced were positively impacted by engagement. Conclusion integrates DTA elaboration likelihood theories propose computational framework understanding Findings offer theoretical insights into mechanisms practical guidance optimizing platform design, training counselors, improving user satisfaction through targeted strategies.

Language: Английский

How digital therapeutic alliances influence the perceived helpfulness of online mental health Q&A: An explainable machine learning approach DOI Creative Commons
Yinghui Huang, Hui Liu, Maomao Chi

et al.

Digital Health, Journal Year: 2025, Volume and Issue: 11

Published: April 1, 2025

Objective This study investigates the role of digital therapeutic alliance (DTA) in predicting and explaining perceived helpfulness responses on online mental health Q&A platforms. Methods constructs a large dataset 19,682 interactions from platforms, employs natural language processing, explainable machine learning, causal inference methods to identify understand factors, particularly DTA, that influence human counselors’ questions. Results The learning-based model for demonstrated strong performance, achieving an root mean square error 0.8234 absolute percentage 22.7288%. explanatory analysis revealed peripheral path-related cues, such as counselor engagement (e.g., word count response time), had highest predictive power. Additionally, central those linked DTA—specifically emotional bonds tasks—significantly influenced were positively impacted by engagement. Conclusion integrates DTA elaboration likelihood theories propose computational framework understanding Findings offer theoretical insights into mechanisms practical guidance optimizing platform design, training counselors, improving user satisfaction through targeted strategies.

Language: Английский

Citations

0