Generative Artificial Intelligence Intervention and Individual Psychological Counseling on Emotional Distress in Young Adults
- Conditions
- Depression, Anxiety
- Registration Number
- NCT06992180
- Lead Sponsor
- Central University of Finance and Economics, China
- Brief Summary
This randomized controlled trial aims to evaluate the effectiveness of an AI-based mental health chatbot, Emohaa, in reducing symptoms of depression, anxiety, insomnia, and academic burnout among young adults experiencing emotional distress. A total of 222 participants will be randomly assigned to one of three groups: (1) AI chatbot intervention, (2) traditional counseling, or (3) waitlist control. The study spans six time points from baseline through one-month follow-up. Primary and secondary outcomes will be measured via validated self-report questionnaires. The intervention is delivered entirely online through a WeChat-integrated platform, ensuring high accessibility and scalability.
- Detailed Description
This three-arm randomized controlled trial will investigate the efficacy of Emohaa, a generative artificial intelligence (AI)-based chatbot designed to deliver empathetic conversational support to young adults experiencing emotional distress. The study will enroll 222 participants, randomly allocated (1:1:1) into one of the following arms:
AI Chatbot Group: Participants will interact weekly with Emohaa via a WeChat-integrated interface for 30-minute sessions over a 6-week period. The chatbot uses fine-tuned large language models to simulate emotionally supportive conversations and provide personalized, non-directive psychological support.
Traditional Counseling Group: Participants will receive six weekly sessions (face-to-face or remote) with trained counselors using structured, supportive counseling protocols derived from cognitive-behavioral therapy (CBT).
Waitlist Control Group: Participants will not receive any intervention during the study period but will be offered access to Emohaa after study completion.
Assessment Schedule and Outcome Measures
Data will be collected via Wenjuanxing online survey links at six time points:
T1: Baseline (before intervention)
T2: After session 2
T3: After session 4
T4: Post-intervention (after session 6)
T5: 2-week follow-up
T6: 1-month follow-up
All outcome measures are validated self-report instruments with established psychometric properties in Chinese populations.
Primary Outcomes (T1-T6) PHQ-9: Patient Health Questionnaire for depressive symptoms
GAD-7: Generalized Anxiety Disorder Scale
ISI: Insomnia Severity Index (7 items)
Academic Burnout Scale (6 items)
Secondary Outcomes (T1-T6) ERQ: Emotion Regulation Questionnaire (10 items)
Loneliness Scale (3 items)
IUS-12: Intolerance of Uncertainty Scale
PSS-10: Perceived Stress Scale
GESS: General Self-Efficacy Scale (10 items)
Additional Measures (T2-T6) CSQ-8: Client Satisfaction Questionnaire
DWAI: Digital Working Alliance Inventory
Data Quality and Collection Procedures All data will be collected electronically using Wenjuanxing (问卷星), which provides timestamped, secure survey distribution and response logging. Participants will receive automatic reminders prior to each time point. The data collection team will perform regular checks to detect and manage missing or inconsistent entries.
To ensure the reliability and validity of the data collection process:
All measurement instruments are widely used, psychometrically validated scales in Chinese.
Survey completion will be monitored in real time.
Study staff will be trained in participant guidance and data integrity monitoring procedures.
Duplicate entries will be flagged using participant ID and contact verification protocols.
Random audits and outlier checks will be conducted at each time point to ensure data quality.
Recruitment & Eligibility
- Status
- NOT_YET_RECRUITING
- Sex
- All
- Target Recruitment
- 222
- Age ranges from 15 to 30
- Suffering from emotional distress (PHQ-9 > 14 or GAD-7 > 12)
- Ability to use the mobile phone to interact with AI
- Consent to participate in the study
- High risk of self-injury and suicide
- Presence of psychotic symptoms (i.e., delusions)
- Serious substance use problems
- Other mental or physical illnesses requiring urgent medical assistance
- Received other psychological or pharmacological interventions in the last month.
Study & Design
- Study Type
- INTERVENTIONAL
- Study Design
- PARALLEL
- Primary Outcome Measures
Name Time Method Depression From enrollment to the end of treatment at 10 weeks The Patient Health Questionnaire-9 (PHQ-9) was employed in this study to measure depressive symptoms. This scale consists of 9 items, each rated on a 4-point Likert scale ranging from 0 (not at all) to 3 (nearly every day). The total score ranges from 0 to 27, with scores of 0-4 indicating no depressive symptoms, 5-9 indicating mild symptoms, 10-14 indicating moderate symptoms, and 15 or above indicating severe depressive symptoms. Previous research has demonstrated good reliability and validity for the Chinese version of the PHQ-9.
Anxiety From enrollment to the end of treatment at 10 weeks The Generalized Anxiety Disorder-7 (GAD-7) was used to assess anxiety symptoms in this study. This scale consists of 7 items, each rated on a 4-point Likert scale ranging from 0 (not at all) to 3 (nearly every day). The total score ranges from 0 to 21, with scores of 0-4 indicating no anxiety symptoms, 5-9 indicating mild symptoms, 10-14 indicating moderate symptoms, and 15 or above indicating severe anxiety symptoms. Prior studies have shown that the Chinese version of the GAD-7 has good reliability and validity.
- Secondary Outcome Measures
Name Time Method Loneliness From enrollment to the end of treatment at 10 weeks The short version of the UCLA Loneliness Scale (ULS-3) was used to assess feelings of loneliness. This scale comprises 3 items, each rated on a 3-point scale ranging from 1 (hardly ever) to 3 (often). The total score ranges from 3 to 9, with higher scores indicating greater loneliness. The short version of the UCLA Loneliness Scale is frequently used in large-scale surveys, and its Chinese version has demonstrated good reliability and validity.