Beyond Alliance Scores: Construct Transport and Measurement Validity in Mental Health Chatbot Research—A Systematic Scoping Review
Systematic scoping review of how therapeutic alliance has been conceptualized, measured and adapted in mental-health chatbot research. It covers 40 reports from 38 studies, using a COSMIN-informed mapping and a construct-transport rubric that was locked before independent re-rating. It concludes that current alliance scores reflect context-specific user-reported relational experience and do not establish equivalence with alliance in human psychotherapy.
Publisher
International Journal of Human–Computer Interaction (Taylor & Francis)
Published
23 Sept 2026
Added
today
Key Findings
- 40 reports from 38 independent studies were included
- 29 studies used Working Alliance Inventory measures with limited modification
- No study met criteria for a sufficient positive conclusion on any mapped measurement property
- Under the locked primary construct-transport rule, none met the Adequate threshold: 25 were Questionable and 13 Inadequate (sensitivity analyses showed the zero-Adequate result was not absolute)
- Users nonetheless reported favourable alliance-like relational experiences
Methodology Notes
Systematic scoping review with COSMIN-informed appraisal. tandfonline walled. Verified via the Crossref record (published online 2026-09-23; affiliations Shahid Beheshti University and Islamic Azad University) and the OpenAlex abstract (1,237 characters), which appeared after the 09-24 sweep.
Sources
IJHCI article(opens in a new tab) (primary)
Authors
Mohammad Mahdi Akbarzadeh, Mobina Akbarzadeh
Tags
Cite This
APA
Mohammad Mahdi Akbarzadeh, Mobina Akbarzadeh. (2026). Beyond Alliance Scores: Construct Transport and Measurement Validity in Mental Health Chatbot Research—A Systematic Scoping Review. International Journal of Human–Computer Interaction (Taylor & Francis). https://www.tandfonline.com/doi/full/10.1080/10447318.2026.2735780