Can a survey show whether nurse educators are ready for AI?
Researchers tested a three-part AI competency scale with 470 nurse educators in Egypt. The structure held across two split samples, but it measures perceived skills, one subscale had borderline reliability and the authors warn against certification or staff appraisal.
Editorial responsibility: The Impact of AI Editorial Desk · Report a factual concern
Research topic
Whether a new self-report instrument provides an initially valid three-domain measure of perceived AI competency among nurse educators
At a glance
- 1The methodological cross-sectional study surveyed 470 nurse educators from multiple Egyptian nursing faculties between mid-October 2025 and February 2026, then split them randomly into exploratory and confirmatory samples of 235 each.
- 2A three-domain structure explained 56.10% of variance and showed strong confirmatory fit, but the domains were nearly uncorrelated and should not be combined into one total score.
- 3The instrument measures perceived—not demonstrated—competency. The authors restrict its use to group needs assessment, research, programme evaluation and self-reflection, not appraisal, certification or high-stakes decisions.
The Impact Brief
Keep the evidence trail, not the noise.
Get the most consequential AI developments with direct sources and clear limits.
Living evidence record
Impact record IAI-0BWAYSS
Evidence stage
Studied
Confidence
Supported
Reporting basis
Source analysis
Independent support
Present
Record status
Monitoring
Last checked
4 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.
The study addresses a measurement gap, not clinical performance
Nursing schools need to decide what educators should understand about AI, where faculty development is needed and whether a training programme changes confidence or practice. Generic digital-literacy surveys may miss clinical teaching, ethical judgement and the professional development involved in AI-supported education. The new Artificial Intelligence Competency Scale for Nurse Educators, or AI-CSNE, was designed for that planning problem.
Its target is perceived competency. Respondents report how they see their own knowledge and practice; they do not diagnose a patient, detect a model error or teach a class under observation. That distinction is central. Self-assessment can reveal confidence and perceived needs, but it can be influenced by familiarity, social desirability and inaccurate calibration. A high score is not proof that an educator can safely evaluate or use an AI system, while a low score may reflect caution rather than inability.[1]
Four hundred and seventy educators supplied a split-sample test
The researchers conducted a methodological cross-sectional study across multiple nursing faculties in Egypt from mid-October 2025 to February 2026. They recruited 470 nurse educators and randomly divided the completed responses into two independent groups. Exploratory factor analysis used the first 235 to identify a possible structure. Confirmatory factor analysis used the other 235 to test whether that structure fit new respondents from the same study population.
The questionnaire was developed from UNESCO’s AI Competency Framework for Teachers and relevant literature, then reviewed by experts and piloted. This sequence is stronger than writing items and testing them on one undivided sample, because the confirmatory half does not determine the model it is asked to test. It still remains a single-country convenience-period study. The design cannot show whether answers predict teaching quality, safe AI use or student outcomes, and the article reports no longitudinal retest of the same educators.[1]
Three distinct domains emerged
The exploratory analysis supported three domains: Foundational AI Knowledge and Instructional Application; Ethical and Human-Centred AI Use; and AI-Supported Professional Growth. Item loadings ranged from 0.661 to 0.822, and the three factors together explained 56.10% of the observed variation in responses. These domains separate practical understanding from ethical orientation and continuing development rather than treating readiness as one interchangeable skill.
Internal-consistency estimates were 0.842 for the foundational domain, 0.810 for ethical and human-centred use, and 0.697 for professional growth in the exploratory sample, with comparable values in the confirmatory sample. The first two are comfortably above the commonly used 0.70 rule of thumb. The third is just below it and should be treated as borderline, not disguised by describing every coefficient as equally strong. Alpha also depends on the number and similarity of items; it does not prove that a scale is valid or stable over time.[1]
Model fit was strong, but a single readiness score would be misleading
The confirmatory model produced a Comparative Fit Index of 0.989, Tucker–Lewis Index of 0.986, root mean square error of approximation of 0.023 and standardised root mean square residual of 0.046. Those values indicate an excellent fit to the authors’ proposed three-factor structure in the second half of their sample. Standardised loadings were strongest in the foundational domain and somewhat lower, though described as acceptable, in the ethical domain.
The three domains were almost uncorrelated. The authors therefore say they should be interpreted separately rather than added into one total score. That is practically important. An educator could report confidence using AI in instruction but less confidence identifying bias, protecting privacy or planning professional learning. A composite score would hide those differences and could make a targeted curriculum less useful. Institutions adopting the scale should display domain profiles, not rank people by a single number.[1]
The permitted uses are intentionally modest
The paper positions AI-CSNE for exploratory group-level needs assessment, comparisons in research, programme evaluation and self-reflective development. A nursing faculty could use anonymous aggregated responses to choose workshop topics, then examine whether domain scores change after training. Researchers could compare patterns across settings after checking that translated versions measure the same constructs. Individual educators could use items as prompts for development conversations they control.
The authors explicitly advise against individual appraisal, certification or other high-stakes evaluation. That caution protects both staff and institutions. Using a self-report scale for promotion or compliance would encourage strategic answers and exceed the evidence. It could also turn healthy uncertainty into an apparent deficit, discouraging educators from admitting where they need support. Any programme evaluation should pair the questionnaire with demonstrations, observed teaching practice, scenario-based ethical reasoning and learner outcomes.[1]
Validity beyond structure still needs to be established
Factor structure and internal consistency are only initial forms of evidence. The study’s tests of convergent, criterion and known-groups validity did not find significant subgroup differences, and the authors call for further validation. It is not yet clear whether scores track an established external measure, distinguish groups expected to differ, predict safer AI-supported teaching or respond reliably to useful training. Near-zero correlations between domains also deserve replication rather than being assumed universal.
Translation and cultural adaptation are particularly important. AI policy, nursing curricula, clinical infrastructure and access to tools differ across countries and institutions. Items that make sense in Egyptian faculties may be interpreted differently where educators have less access to AI systems or different professional duties. Future studies should test measurement invariance across languages and settings, test–retest reliability, sensitivity to change, and associations with performance-based assessments and student outcomes.[1]
Ethics and disclosure are clear, while generalisability is limited
Tanta University’s ethics committee approved the study under reference 819-10-2025. Participants consented electronically, answered an anonymous Google Forms survey and received no incentive. The authors say data were kept on password-protected university servers, with no names, email addresses or contact information collected. They report no specific grant from a public, private or nonprofit funder and declare no competing interests. Authors were affiliated with institutions in Egypt, Somalia and the United Arab Emirates.
The article was published on 3 October as a peer-reviewed, accepted early version with a permanent DOI; copyediting remains pending. Its strongest conclusion is that this three-domain instrument has promising initial structural evidence in 470 Egyptian nurse educators. Evidence that would change the assessment includes independent replication, pre-registered cross-cultural validation and a demonstrated relationship with observable teaching or safety outcomes. Until then, it is a planning and reflection tool—not a licence to practise with AI.[1]
What this means for people
- Educators may receive more targeted training when faculties can separate foundational, ethical and professional-growth needs.
- Staff could be harmed if a self-report development tool is repurposed for ranking, promotion or compliance decisions.
- Students and patients benefit only if perceived readiness is supplemented by demonstrated competence, safe practice and evaluated teaching outcomes.
Global context
The study was conducted in Egypt by authors affiliated with institutions in Egypt, Somalia and the United Arab Emirates. It adds African and Middle Eastern evidence to a field often dominated by high-income Western settings, but one national sample cannot represent those regions. Nursing scope, educator preparation, language, regulation and access to AI vary widely, so cross-cultural validation should precede international comparison.
What the evidence does not yet show
- The cross-sectional, self-report design measures perceived competency, not observed teaching, clinical judgement, student outcomes or safe AI use.
- All 470 participants came from Egyptian nursing faculties; transfer to other countries, languages, professions and institutional contexts has not been established.
- The AI-Supported Professional Growth subscale had Cronbach’s alpha of 0.697 in the exploratory sample, just below the conventional 0.70 heuristic.
- Convergent, criterion and known-groups evidence remains incomplete, and no longitudinal test–retest or predictive validation was reported.
- The early-access article is peer reviewed and citable but still subject to copyediting before the final Version of Record.
What to watch next
- Independent replication and measurement-invariance testing across countries, languages and nursing-education systems.
- Performance-based comparisons showing whether domain scores relate to demonstrated AI literacy, ethical reasoning and teaching quality.
- Test–retest reliability and sensitivity to meaningful faculty-development programmes.
- Whether institutions follow the authors’ warning against certification, appraisal or other high-stakes individual use.
Evidence trail
Sources used for this report
Links checked 4 October 2026
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.
Continue the story
Related reporting
Education
Does the way students use AI matter?
A peer-reviewed survey of 713 Saudi undergraduates found that supportive and creative AI use was associated with stronger self-reported creativity and metacognitive awareness, while substitutive use showed negative associations. The one-time self-report design cannot establish what AI caused—or whether skills improved.
8 min · 1 source
Education
What should teachers learn about AI? UNESCO reviews 36 competency frameworks
A new UNESCO IITE analysis compares national, institutional and academic frameworks for teacher AI competence. It offers a policy checklist, but does not test whether any framework improves classroom outcomes.
5 min · 2 sources
Education
PISA analysis separates frequent AI use from effective learning
New OECD analysis of PISA evidence explores how students use AI and digital tools alongside reading skills, warning that access or frequency alone does not establish deeper understanding.
4 min · 1 source
Reader commentary
Add evidence, experience or a question
No account is required. Reader notes are published after a brief civility, relevance and safety check; disagreement is welcome.
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.