Enterprise AI Analysis

Robust Quasi-Experimental Design

This study employed a one-group pretest-posttest quasi-experimental design involving a stratified random sample of 1,000 undergraduate nursing students from the Faculty of Nursing, Sohag University, Egypt. Data were collected via a self-administered questionnaire at baseline and one month post-intervention.

Key Instruments: The AI Knowledge Scale (AIKS-16; 0–32) measured foundational AI literacy. The AI Acceptance Scale (AIA-34; 0–136), aligned with the Technology Acceptance Model (TAM), assessed Perceived Usefulness (PU), Perceived Ease of Use (PEOU), and Attitude/Intention. Both scales demonstrated strong internal consistency (Cronbach's α ≥ 0.85) and underwent rigorous translation and cultural validation. Threats to internal validity were mitigated by a short exposure window, identical delivery across groups, and proctored assessments.

The intervention was a standardized 10-session blended curriculum, delivered in small groups, incorporating micro-lectures, case-based discussions, brief hands-on demonstrations, and guided ethical reflection. Each session was supported by a facilitator script and fidelity checklist.

Significant Gains and Strong Correlations

The curriculum resulted in substantial and equitable improvements across all measures:

• AI Knowledge: Increased significantly from an average of 15.01±4.72 to 30.33±3.11 (p<0.001), demonstrating a large effect size (Cohen's d ≈2.77). This represents a shift from roughly 47% to 94.8% of the maximum score.
• AI Acceptance: Rose dramatically from 67.02±13.47 to 122.33±9.21 (p<0.001), also with a large effect size (Cohen's d ≈3.11). This translates to an increase from approximately 49% to 89.9% of the maximum score.
• TAM Subdomains: All subdomains—Perceived Usefulness, Perceived Ease of Use, and Attitude/Intention—showed significant gains, with PEOU registering the largest standardized change (d ≈2.99).
• Knowledge-Acceptance Correlation: Post-test knowledge was strongly correlated with acceptance (r=0.647, p<0.001), supporting the theoretical pathway that understanding fosters adoption.
• Equitable Impact: ANCOVA revealed no educationally meaningful differences in knowledge or acceptance gains by sex or residence after controlling for baseline scores, indicating uniform program effectiveness.

These results confirm that the blended curriculum effectively enhanced AI literacy and willingness to adopt AI tools among nursing students.

Future-Proofing Nursing with AI Literacy

This study provides a reproducible educational protocol for integrating AI literacy into nursing curricula, moving beyond mere exposure to fostering active, accountable, and ethically grounded engagement with AI. The findings underscore the importance of:

• Concept Scaffolding: Building strong foundational knowledge about AI, its capabilities, limitations, and ethical considerations.
• Guided, Low-Risk Practice: Offering hands-on experience with AI tools in simulated and study contexts to improve perceived ease of use.
• Explicit Ethical Guardrails: Teaching students about bias, data privacy, and the importance of disclosure and verification routines to promote responsible AI use.
• Faculty Development: Equipping instructors with the necessary knowledge and resources to teach AI effectively, using standardized materials and clear grading rubrics.
• Assessment Redesign: Shifting assessment from factual recall to applied judgment, evaluating students' ability to appraise AI outputs and articulate safe use boundaries.

By institutionalizing this model, nursing programs can prepare graduates not just to use AI, but to use it accountably, promoting safety, equity, and person-centred care in an AI-enabled healthcare landscape.