Related Experiment Video
Updated: Jun 11, 2026

Virtual Agent for Real-Time Motivational Interviewing by Integrating Adaptive Nonverbal Behavior and Language Models
Published on: December 23, 2025
SleepPathfinder: A Socratic Questioning and Self-Decision-Based Chatbot to Support User Engagement in Digital CBT-I:
Youjin Roh1, Anderson Sungmin Yoon2, Hayoung Oh1
1Department of Applied Artificial Intelligence, Sungkyunkwan University, #32541, Dasan Hall of Economics, Humanities and Social Sciences Campus, Seoul, Republic of Korea, 82 2-740-1785, 82 2-740-1785.
A new chatbot, SleepPathfinder, uses Socratic questioning to improve digital Cognitive Behavioral Therapy for Insomnia (dCBT-I). It enhances user engagement and understanding by supporting cognitive exploration and autonomy in insomnia treatment.
Area of Science:
- Digital health
- Psychology
- Human-computer interaction
Background:
- Chronic insomnia significantly impacts quality of life and mental health.
- Cognitive Behavioral Therapy for Insomnia (CBT-I) is a first-line treatment, with digital CBT-I (dCBT-I) enhancing accessibility.
- Current dCBT-I systems offer limited support for cognitive exploration and meaning-making, potentially hindering long-term adherence.
Purpose of the Study:
- To evaluate the usability and feasibility of SleepPathfinder, a conversational dCBT-I chatbot.
- To assess SleepPathfinder's ability to support user understanding and engagement with CBT-I through Socratic questioning and self-decision.
Main Methods:
- A 4-stage conversational flow: education, Socratic exploration, self-decision, and advice.
- A pilot usability study (n=45) assessed system stability and user experience.
- A 5-day comparative experiment (n=30) compared exploratory vs. directive dialogue conditions, measuring usability, cognitive appraisals, autonomy, and readiness for change.
Main Results:
- The exploratory dialogue condition tended to reduce perceived threat and severity of sleep problems.
- Autonomy-related experiences, including perceived choice and engagement, showed upward trends in the exploratory condition.
- Deeper user self-narratives correlated with greater shifts in cognitive appraisals, while chatbot question frequency did not.
Conclusions:
- Socratic questioning and self-decision conversational structures are usable and feasible within dCBT-I systems.
- This approach complements existing CBT-I protocols by facilitating cognitive exploration and user autonomy.
- Findings provide design-oriented evidence for refining dialogue-supported dCBT-I systems to enhance user engagement.
Related Concept Videos
Sleepwalking and Sleep Talking
Factors that increase the likelihood of sleepwalking include sleep deprivation and alcohol consumption. Contrary to common beliefs, it is safe...
Blind Procedures
Understanding Sleep
The circadian rhythm, a nearly 24-hour cycle, is deeply influenced by environmental light cues. Light exposure directly affects the hypothalamus, which in turn regulates...
