PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
March 7, 2026Journal of Technology in Behavioral Science0 citationsOpen Access

Reducing Depressive Symptoms through AI-Guided Narrative Self-Films: A Formative Randomized Controlled Trial

EYElvin YaoHYHarry Yang

Key Points

  • This study aims to evaluate the effectiveness of AI-guided narrative self-films in reducing depressive symptoms and enhancing psychological well-being.
  • Conducted a factorial randomized controlled trial with 181 Chinese undergraduates.
  • Participants created AI-generated videos about their past, present, and future selves over five days.
  • Measured depression, hope, narrative coherence, and self-continuity before and after the intervention.
  • Conflict-resolution storytelling significantly reduced depressive symptoms on the PHQ-9 compared to ideal-only storytelling.
  • The greatest improvement was observed in the interdependent and conflict-resolution group.
  • Participants in the conflict-resolution group showed notable increases in hope, narrative coherence, and self-continuity.

Abstract

Abstract Depression remains a leading global health concern, yet theory-based applications of generative artificial intelligence (AI) to mental-health support have received little empirical scrutiny. The present study tested a five-day, AI-assisted “self-film” intervention grounded in narrative-identity theory. In a 3 (value framing: independent, interdependent, neutral) × 2 (narrative arc: ideal-only, conflict-resolution) factorial randomized controlled trial, 181 Chinese undergraduates with mild-to-moderate depressive symptoms produced short AI-generated videos depicting their past, present, and future selves and then compiled them into personalised self-films. Depression, state hope, narrative coherence, and future self-continuity were measured before and after the programme. Conflict-resolution storytelling reduced depressive symptoms on the PHQ-9 compared with ideal-only conditions; the strongest decline occurred in the interdependent/conflict-resolution group. Conflict-resolution participants also recorded significant gains in hope, narrative coherence, and self-continuity. These results provide preliminary evidence that a brief, culturally attuned, AI-supported storytelling exercise can enhance multiple indices of psychological well-being, suggesting promise for scalable digital interventions beyond traditional clinical settings.Depression remains a leading global health concern, yet theory-based applications of generative artificial intelligence (AI) To mental-health support Have Received Little Empirical scrutiny. The Present Study Tested a five-day, AI-assisted “self-film” Intervention Grounded in Narrative-identity theory. In a 3 (value Framing: independent, interdependent, neutral) × 2 (narrative Arc: ideal-only, conflict-resolution) Factorial Randomized Controlled trial, 181 Chinese Undergraduates with mild-to-moderate Depressive Symptoms Produced Short AI-generated Videos Depicting their past, Present, and Future Selves and Then Compiled Them into Personalized self-films. Depression, State hope, Narrative coherence, and Future self-continuity Were Measured before and after the Study. Conflict-resolution Storytelling Reduced Depressive Symptoms on the PHQ-9 Compared with ideal-only conditions; the Strongest Decline Occurred in the interdependent/conflict-resolution group. Conflict-resolution Participants Also Recorded Significant Gains in hope, Narrative coherence, and self-continuity. These Results Provide Preliminary Evidence that a brief, Culturally attuned, AI-supported Storytelling Exercise Can Enhance Multiple Indices of Psychological well-being, Suggesting Promise for Scalable Digital Interventions Beyond Traditional Clinical Settings.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Yao et al. (2026) studied this question.

synapsesocial.com/papers/69abc1845af8044f7a4ea4c6https://doi.org/10.1007/s41347-026-00596-5
Ask AI
Helpful
Bookmark
Share
View Full Paper