Related Experiment Videos
Cross-cultural adaptation and psychometric evaluation of the Shinners Artificial Intelligence Perception Scale
Meltem Özduyan Kiliç1, Ergün Kaplan2, Fatoş Korkmaz2
1Faculty of Nursing, Ankara University, Ankara, 06080, Turkey. meltem.ozduyan17@gmail.com.
Background:
Integrating artificial intelligence (AI) technologies into nursing practice is expanding rapidly, yet validated instruments to assess nursing students' perceptions of AI remain scarce in non-English-speaking contexts. The Shinners Artificial Intelligence Perception (SAIP) Scale was developed to measure healthcare professionals' perceptions of AI; however, its applicability in Turkish culture has not been tested.
Methods:
This methodological study was conducted between February and July 2025 and involved 412 nursing students from three public nursing faculties in Ankara, Türkiye. The cross-cultural adaptation process followed forward-backward translation, expert review, and content validity assessment using the Davis technique. Construct validity was examined through exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). Reliability was assessed using item-total correlations, Cronbach's alpha, and test-retest analysis with a subgroup of 32 participants.
Results:
Content validity index (CVI) was 0.88. The KMO sample adequacy was 0.802, and Bartlett's test yielded χ2(36) = 651.198 (p < 0.001). EFA supported a two-factor structure, explaining 48.1% of the total variance. CFA supported the two-factor structure, with incremental fit indices (CFI = 0.968, TLI = 0.954, IFI = 0.969) and absolute fit indices (χ²/df = 1.608, GFI = 0.966, SRMR = 0.064) indicating generally acceptable fit. However, the RMSEA value of 0.090 (90% CI: 0.065-0.122) suggests borderline model fit, as the upper limit of the confidence interval slightly exceeds commonly recommended thresholds. Cronbach's alpha values were 0.772 for the overall scale, 0.763 for Factor 1 (Perception of AI's impact on professional role), and 0.723 for Factor 2 (Perceptions of preparedness for AI). Test-retest reliability demonstrated significant positive correlations (r = 0.584-0.752, p < 0.001) and no significant differences between the two measurement points.
Conclusions:
In this study, psychometric evaluation yielded an 8-item Turkish version of the SAIP (SAIP-TR); thus, the adapted instrument should be considered a shortened form of the original scale. The SAIP-TR may provide a useful framework for assessing AI-related perceptions among nursing students and may support educators and researchers in understanding students' readiness for AI integration into nursing practice and in informing future educational strategies. However, further validation studies in more diverse populations are recommended to strengthen the scale's generalizability and conceptual stability.
Clinical Trial Number:
Not applicable.
