An AI-Supported Simulation-Based Serious Game in Digital Health Management Education: Design and Formative Evaluation
DOI:
https://doi.org/10.34190/ecel.25.1.5513Keywords:
Digital Health Management, Digital Transformation, Serious Games, Education, Artificial IntelligenceAbstract
Background Teaching digital transformation in healthcare management requires students to address complex organizational challenges, multiple stakeholder perspectives, and decision-making under uncertainty. Traditional case-based teaching provides limited experience of real-world decision-making complexity and feedback. Objectives This study presents the design and formative evaluation of the first version of an artificial intelligence (AI)-supported digital serious game (DSG) in Digital Health Management (DHM) education. It examined students’ evaluations of the integrated management methods, generative AI (GenAI)-supported stakeholder interactions, educational design, and AI-generated model solutions. Methods Design of the DSG was informed by the co.LAB generic framework and resulted in a scenario-based learning environment integrating management methods, collaborative problem-solving, GenAI-supported stakeholder interactions, and AI-generated model solutions. The formative evaluation used an embedded mixed-methods design across bachelor’s and master’s degree courses at two higher education institutions. Twenty-one students participated in the DSG, of whom 15 completed the interim evaluation, while the number of method-specific evaluations varied across learning activities. Qualitative reflections were collected at group level. Results The browser-based DSG combined healthcare digitalization case scenarios, sequential management methods, GenAI-supported stakeholder interactions, collaborative problem-solving, and AI-generated model solutions. The management methods received consistently positive ratings (M = 4.19–4.55/5). Students also positively evaluated the GenAI-supported stakeholder interactions (M = 5.95/7), their usability (M = 4.18/5), the teaching-learning experience (M = 4.98/6), and the AI-generated model solutions (M = 3.98/5). Internal consistency ranged from acceptable to excellent (α = .72 to .91), while Spearman-Brown coefficients for two-item subscales ranged from .66 to .92. The qualitative findings suggest that the AI-generated model solutions supported validation, extension, and critical comparison of students’ own analyses. The Net Promoter Score (NPS) was 40. Discussion The findings suggest that GenAI can support scenario-based learning in DHM when GenAI-supported stakeholder interactions and feedback are embedded in a coherent, collaborative learning design. Future studies should use a prospectively defined evaluation framework to compare different forms of AI-generated feedback and assess reflection, objective learning outcomes, competency development, and transfer in larger samples. Conclusion The study provides an initial design and empirical basis for the further development of AI-supported DSGs in DHM education. The findings indicate that the AI-supported components may support perspective taking, comparison, reflection, and method-based decision-making within a coherent learning design.