Common Factors in Psychotherapy: Enhancing Provider-to-Patient Dynamics to Improve Patient Outcomes
Track: Non-traditional track
Keywords: clinical benchmarks, common factors, health equity, machine learning, artificial intelligence
TL;DR: We describe our approach to benchmarking Common Factors Therapy Skills of Empathy and Collaboration in LLM development.
Abstract: For this paper, we describe our approach to benchmarking Common Factors of Empathy and Collaboration on the HOPE dataset—a publicly available dataset comprising 12.8k utterances from 212 therapy sessions involving a therapist and client dyad. Malhotra et al. (2022) conducted thorough processing of the HOPE dataset to eliminate noise and transcription errors. Common Factors Theory encompasses factors from (1) the client, (2) provider and, (3) therapeutic context; we specifically focus on provider behaviors in this paper. Our central research question: Can we produce a scalable, consistent, and unbiased way to assess the occurrences of reflective listening, appreciation, and confrontation–markers of empathy and collaboration, the core features of Common Factors Theory–using natural language processing and AI methods to augment provider communications?
**Could not add my co-authors in the portal.
Our full team: Alison Cerezo, PhD,1,2, Vijaykumar Palat, MS1, Amber Jolley-Paige, PhD1, Sarah Peregrine Lord, PsyD1,3
(1 mpathic.ai; 2 University of California Santa Barbara, 3 University of Washington)
Presentation And Attendance Policy: I have read and agree with the symposium's policy on behalf of myself and my co-authors.
Ethics Board Approval: No, our research does not involve datasets that need IRB approval or its equivalent.
Data And Code Availability: Yes, we will make data and code available upon acceptance.
Primary Area: Clinical foundation models
Student First Author: No, the primary author of the manuscript is NOT a student.
Submission Number: 25
Loading