Empowering nursing students during AI era: educational strategies for enhancing knowledge and acceptance of artificial intelligence

Abstract

Artificial intelligence (AI) is rapidly permeating health systems, yet undergraduate nursing students often report limited AI literacy and uncertainty about safe, appropriate use. To evaluate the effect of a standardized 10-session blended curriculum on nursing students’ AI knowledge and acceptance and to examine theory-consistent associations between knowledge and acceptance. One-group pretest–posttest quasi-experimental study. A stratified random sample of undergraduates (n=1,000) from the Faculty of Nursing, Sohag University (Egypt) completed a self-administered questionnaire at baseline and one month after the program. Outcomes were the AI Knowledge Scale (AIKS-16; 0–32) and the AI Acceptance Scale (AIA-34; 0–136) aligned with Technology Acceptance Model subdomains (Perceived Usefulness [PU], Perceived Ease of Use [PEOU], Attitude/Intention). Analyses used paired t-tests with 95% CIs and Cohen’s d; Pearson correlations; and exploratory ANCOVA adjusting for baseline to probe subgroup differences (sex, residence). Ethics approval: IRB 88-6-2023. Knowledge increased from 15.01±4.72 to 30.33±3.11 and acceptance from 67.02±13.47 to 122.33±9.21 (both p<0.001; large effects). Gains were observed across PU, PEOU, and Attitude/Intention. Post-test knowledge correlated strongly with acceptance (r=0.647, p<0.001). ANCOVA showed no educationally meaningful differences by sex or residence after adjustment (partial η2 ≤0.006). Knowledge-level transitions indicated marked movement from low/moderate to high categories. A standardized, fidelity-checked blended curriculum produced substantial and equitable improvements in AI knowledge and acceptance among undergraduate nursing students. Framed by the Technology Acceptance Model, results suggest that concept scaffolding, brief hands-on practice, and explicit disclosure/verification routines strengthen perceived usefulness and ease of use, supporting accountable AI adoption. Multi-site controlled studies with performance-based outcomes and longer follow-up are warranted. Not applicable. This was an educational pretest–posttest study with no clinical trial component.

Ähnliche Arbeiten