Effects of an
AIGC
‐Enhanced
3C
Feedback Simulation on Critical Thinking in Nursing Education: A Mixed‐Methods Study
Jimei Luo, Lina Ma, Changxiu Li, Yan Wang, Dan Feng, Rong Liu, Luwen Luo, Yue Chen, Hongmin Ran, Qiuhang Hu, Lulu Huang, Hui Zeng ABSTRACT
Aim
Empirical evidence on the use of artificial intelligence‐generated content (AIGC)–enhanced 3C feedback simulation in nursing education remains limited. This study aimed to evaluate its effect on critical thinking disposition among undergraduate nursing students and to explore students' learning experiences with this teaching approach.
Design
A mixed‐methods study with a cluster‐randomised controlled design.
Methods
Seven classes of first‐year undergraduate nursing students from one university were included ( N = 350) and cluster‐randomised by class to the intervention group (four classes, n = 200) or control group (three classes, n = 150). The intervention group received AIGC‐enhanced 3C feedback simulation, incorporating AI‐assisted script generation, critical appraisal of AI‐generated content, and 3C structured reflection, whereas the control group received traditional lecture‐based teaching. The primary outcome was critical thinking disposition measured using the Chinese version of the Critical Thinking Disposition Inventory (CTDI‐CV), and the secondary outcome was theoretical knowledge test performance. Quantitative outcomes were analysed using linear mixed‐effects models to account for class‐level clustering because the theoretical knowledge test was administered only after the intervention; its analysis did not adjust for a baseline knowledge score. After the intervention, four focus groups were conducted with students from the intervention group ( n = 40), and the transcripts were analysed together with 3C reflection prompt card texts using thematic analysis.
Results
Baseline CTDI‐CV total scores did not differ significantly between the two groups. After adjustment for baseline scores and class‐level clustering, the intervention group had a higher post‐intervention CTDI‐CV total score than the control group (β = 38.574, 95% CI 33.287–43.861, p < 0.001). Between‐group differences across all seven CTDI‐CV dimensions remained statistically significant after Holm adjustment. The intervention group also had a higher theoretical knowledge test score after accounting for class‐level clustering (estimated mean difference = 4.80, 95% CI 0.72–8.88, p = 0.021). Five qualitative themes were identified: a shift from knowledge transmission to situational judgement; structured reflection as a thinking scaffold; role switching as a means of promoting multiple‐perspective thinking; questioning the known and exploring the unknown; and AIGC as a touchstone for critical thinking. The qualitative findings complemented the quantitative results by providing insight into students' learning experiences.
Conclusion
After adjustment for baseline levels and class‐level clustering, undergraduate nursing students who received AIGC‐enhanced 3C feedback simulation demonstrated higher post‐intervention critical thinking disposition than those who received traditional teaching. The qualitative findings suggested that situational judgement, structured reflection, role switching, and multiple‐perspective experiences, and critical appraisal of AI‐generated content may represent important learning experiences associated with critical thinking disposition within this teaching approach. Because the primary outcome was a self‐reported measure of critical thinking disposition, the findings should not be interpreted as direct evidence of improved objective clinical reasoning ability.
Implications for the Profession and/or Patient Care
Integrating critical appraisal of AI‐generated content with scenario simulation and structured reflection may provide a teaching approach for further evaluation in the development of critical thinking in nursing education.
Impact
This study addresses the challenge of fostering critical thinking while integrating generative AI into nursing education and provides preliminary empirical evidence to inform the design of AI‐supported simulation‐based teaching.
Reporting Method
The study was reported in accordance with applicable EQUATOR Network reporting guidance.
Patient or Public Contribution
No patients or members of the public were involved in the study design, conduct, analysis, interpretation of the findings, or manuscript preparation. All participants were undergraduate nursing students.