Did an AI course improve nursing students' skills? Self-rated literacy rose—but the study did not test clinical performance
After a 10-week blended course, 72 first-year nursing students reported higher AI literacy than 70 second-year controls. Academic year and group were inseparable, outcomes were self-reported and there was no long-term or patient-care test.
The short version
- What happened
- The intervention combined 12 hours of online lessons with six hours of workshops on generation, evaluation, ethics and simple agent-building over 10 weeks.
- Who it affects
- Student nurses may benefit from supervised practice that includes verification and ethics, but educators still need objective evidence that those lessons transfer to clinical decisions and patient safety.
- What remains unproven
- Single university and 142 analysed students
Keep in mind
Single university and 142 analysed students
Convenience sample with non-random allocation
Academic year completely confounded with group
Routine rather than attention-matched control
Immediate self-report outcomes
No objective knowledge, clinical-performance or patient outcome measure
Source analysis · 1 linked source. Primary analysed source date: 27 October 2026.
Keep following the evidence
Join The Impact Brief founding list for selected AI reports, their effect on people and the evidence limits. Read a sample first.
Read the full report · 4 min The evidence, methods, implications and limitations
The course mixed concepts, practice and responsible-use teaching
A peer-reviewed quasi-experiment at Lishui University in China found higher self-rated AI literacy after a structured course for nursing undergraduates. The programme gave 72 first-year students 12 hours of online learning and six hours of supervised workshops over 10 weeks. Topics included generative text and images, simple agent development, nursing teaching materials, report writing, privacy, bias and checking AI output. Seventy second-year students continued the routine nursing curriculum and could use public AI tools independently.
The researchers recruited 160 students and analysed the 142 who completed valid pre- and post-course questionnaires. Trained assistants who were not involved in analysis collected the surveys. The comparison therefore measured a supported AI curriculum against normal teaching, not AI access against no access and not a single chatbot against another teaching method. Nearly half of the analysed students already reported daily AI use, so the result is about structure and supervision rather than first exposure.[1]
Self-rated AI literacy showed the clearest difference
On the 12-to-84 AI Literacy Scale, the intervention group's mean rose from 55.67 to 65.01, while the control mean moved from 53.01 to 54.19. After adjusting for baseline score, previous AI training, use frequency and digital literacy, adjusted post-course means were 64.92 and 53.91: a difference of 11.01 points (95% confidence interval 8.55 to 13.47). The within-group effect size for the intervention was 0.66, and the adjusted model's partial eta-squared was 0.367.
Attitude scores also favoured the course by 4.60 points on a 15-to-75 scale. That should not be simplified to greater enthusiasm. Exploratory results indicated stronger awareness of risks as well as uses, which could reflect a more critical stance. The study assessed questionnaires immediately after teaching; it did not ask students to identify hallucinations in a blinded task, protect simulated patient data, make a clinical decision or demonstrate safer care.[1]
The distinction that matters
What was measured—and what was not
The strongest result was a questionnaire score, not an observed nursing outcome.
Course
72 first-year students; 18 hours
Self-rated literacy increased
Control
70 second-year students; routine curriculum
No comparable literacy change
Missing test
No randomization or clinical task
Causality and workplace transfer unknown
The design cannot establish that the course caused the gap
Allocation followed the timetable rather than chance: first-year students received the course and second-year students formed the control group. Academic year was therefore completely confounded with the intervention and could not be adjusted away. The groups looked similar on measured baseline characteristics, but experience, maturity, workload or adaptation to university could still account for part of the result. Eighteen recruited students were not in the final analysis, and the paper does not turn that attrition into an intention-to-treat comparison. A routine-curriculum control also received less structured attention than the course group.
Autonomous-learning scores illustrate the problem. They did not significantly improve in the intervention group; the adjusted difference arose because control scores fell sharply. No formal group-by-time interaction was tested, so the paper does not establish that the course protected autonomy. For nursing schools, the study offers a reproducible 18-hour curriculum to test, not proof of clinical readiness. Randomized or year-matched studies with objective tasks, delayed follow-up and observed placement performance would provide stronger evidence.[1]
Watch the background, then check the evidence
Illustrative photo · not a video frame
UNESCO · 8 min 30 sec · Background explainer
Teacher to Teacher: Is AI reshaping education for the better?
YouTube loads only when you choose; playback starts when you press play.
A conversation about how teachers can guide AI use while preserving pupils’ judgement. Watch for the difference between helping a learner think and supplying an answer.
Evidence check: Teachers’ perspectives provide context. They do not establish that a particular AI tool improves learning outcomes.
Published 2025-01-23. Source checked 14 October 2026. Check captions on the publisher’s video page.
Read the education evidencePhoto: Vitaly Gariev / Unsplash · Licence
Continue the story
Read next
Continue exploring generative ai: compare the finding and its evidence limits.
How are Kuwaiti health students using AI to write in English?
In a peer-reviewed survey of 241 nursing and health-sciences undergraduates, 94.6% reported previous generative-AI use and 83.4% used it for academic work at least sometimes. Students also reported accuracy and dependency concerns, while the single-institution self-report design cannot show that AI improved writing or learning.
Read the next reportExplore the full source and evidence record
What this means for people
- Student nurses may benefit from supervised practice that includes verification and ethics, but educators still need objective evidence that those lessons transfer to clinical decisions and patient safety.
Global context
The course was designed for one Chinese nursing programme and used locally available platforms. Curricula, clinical responsibilities, data rules and access to generative-AI systems differ across countries.
What the evidence does not yet show
- Single university and 142 analysed students
- Convenience sample with non-random allocation
- Academic year completely confounded with group
- Routine rather than attention-matched control
- Immediate self-report outcomes
- No objective knowledge, clinical-performance or patient outcome measure
What to watch next
- Randomized or year-matched replications
- Objective output-checking and privacy tasks
- Delayed retention tests
- Performance during clinical placements
- Collaborative-learning outcomes and patient-safety measures
Living evidence record
Impact record IAI-15X2ZNE
Evidence stage
Studied
Confidence
Supported
Reporting basis
Source analysis
Independent or research support
Present
Record status
Monitoring
Last checked
27 October 2026
Source trail
1 direct source across 1 source type.
People impact
Documented in this record.
Uncertainty
Limits and next checks are explicit.
Stages describe the evidence available—not whether a technology is good or bad. See the public method.
Single-source reporting disclosure
This record analyses one direct source. It can establish what BMC Nursing published or reported, but it is not independent corroboration of every performance claim or predicted outcome. The confidence label will change only when broader evidence is added.
Evidence trail
Sources used for this report
Links checked 27 October 2026
Research question, topics and regions
Whether a short structured AI and generative-AI course changed nursing students' self-reported literacy, attitudes and autonomous learning
Topics: Nursing education · AI literacy · Generative AI · Clinical training · Responsible use
Regions: China · Asia · Global
This report is labelled source analysis. We summarise and analyse source material in our own words; company statements remain attributed claims until independently supported. Translated summaries preserve the meaning of the original source and link back to it. Read our editorial standards.
Reader commentary
What does this mean where you are?
Tell us what you have seen at work, in your community or in everyday life. A question is welcome too. No account is needed; we review notes before publishing them.
Explore commentary across the portal →Published reader notes
0No published reader notes yet. You can start the evidence-led discussion above.
Prefer a private correction or response? Contact the newsroom.