Hybrid Reasoning AI System for Nursing Ethics Education: Development and Randomized Controlled Trial

Objective: Evaluate the effectiveness of NurseEthAI in enhancing the ethical decision-making ability of nursing students and compare it with traditional case-based teaching methods. Methods: Using a randomized controlled trial design, 80 third year nursing students from a certain university were included and randomly divided into an experimental group and a control group. Evaluate based on the nursing ethics decision-making power scale, and compare the intervention effect through pre-test and post test analysis. Results: The total score of ethical decision-making in the experimental group significantly increased from $310.49 \pm$ 25.33 points to $355.73 \pm 24.57$ points ($\mathrm{t}=8.37, \mathrm{p}<0.001$), while the control group increased from $306.74 \pm 26.18$ points to $324.39 \pm 25.62$ points ($\mathrm{t}=3.45, \mathrm{p}=0.001$). The improvement in ethical action dimension is particularly significant ($\mathbf{p}<\mathbf{0. 0 0 1}$). The distribution of ability levels shows that the proportion of high-level decision-makers in the experimental group increased from 30.00 % to 80.00 %, while the control group only increased from 27.50 % to 45.00 %. Conclusion: The NurseEthAI system can significantly enhance the ethical decision-making ability of nursing students, especially in the dimension of ethical action, providing an efficient tool for clinical nursing education. In the future, it is necessary to combine clinical practice to further optimize system functions and promote the transformation from “ethical cognition” to “ethical behavior”.