AI Virtual Human-Augmented Game-Based Teaching to Enhance Emotional Intelligence in Nursing Students: Protocol for a Single-Group Pretest-Posttest Action Research Study.

BACKGROUND: The rapid advancement of IT and the complexity of health care demand innovative nursing education that moves beyond lectures and workshops. Nursing students must acquire clinical knowledge alongside emotional intelligence (EI), empathic communication, and crisis management to respond to patients at risk for suicide. In Taiwan, suicide is the second leading cause of death among university students, underscoring the urgency of suicide prevention training. Generative artificial intelligence (AI) platforms such as AI virtual humans provide immersive, scenario-based simulations that can integrate game-based learning into psychiatric nursing curricula. OBJECTIVE: This study evaluates the effectiveness of combining game-based pedagogy with AI virtual human simulations in enhancing nursing students' EI, empathic communication, and psychological crisis management. Specifically, it aims to (1) foster self-awareness and emotion regulation, (2) strengthen crisis assessment and coping strategies, (3) increase motivation and engagement, and (4) establish an evidence-based framework for nursing education innovation. METHODS: An action research design will be conducted over 2 instructional cycles (36 weeks) within a psychiatric nursing course at a Taiwanese university. Participants are third-year postbaccalaureate nursing students recruited through convenience sampling. In phase 1, five simulation scenarios involving suicide risk presentations-major depressive disorder, bipolar disorder, schizophrenia, substance use disorder, and borderline personality disorder-were developed and validated by experts. Phase 2 applies 2 action research cycles (week 1 to 18 and week 19 to 36) with a single-group pretest-posttest design. Quantitative data include the Adult Emotional Intelligence Scale, administered at baseline and after the intervention. Qualitative data consist of reflection journals and classroom observations. Analysis employs descriptive statistics, paired sample 2-tailed t tests or Wilcoxon signed-rank tests, effect sizes, linear mixed-effects models, and thematic analysis. RESULTS: Scenario development and expert review were completed in phase 1. Cycle 1 (weeks 1-18) has begun from September 2025, with 37 students enrolled, exceeding the sample size target of 32. Postintervention Adult Emotional Intelligence Scale results will be collected by week 16, followed by cycle 2 with a new cohort. Final integration and analysis will occur after week 36 to evaluate intervention effectiveness. CONCLUSIONS: Integrating game-based learning with AI virtual human simulations introduces an innovative, technology-enhanced model for psychiatric nursing education. Anticipated outcomes include improved EI, enhanced empathic communication, and stronger crisis management abilities. Through iterative refinement across 2 cycles, this study will provide empirical evidence on the role of generative AI in nursing education, informing curriculum innovation and supporting AI-driven simulation as a sustainable strategy for competency-based training. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): PRR1-10.2196/80290.