Cross-cultural Adaptation and Psychometric Evaluation of the Artificial Intelligence Ethical Reflection Scale

The rapid integration of artificial intelligence (AI) into health care has heightened the importance of ethical reflection, accountability, and responsible decision-making in nursing practice. Accordingly, there is a need for valid and reliable instruments to assess AI-related ethical reflection. This study aimed to adapt the Artificial Intelligence Ethical Reflection Scale (AIERS) into Turkish and evaluate its psychometric properties in a sample of nursing students. The study was conducted with 427 students enrolled in an undergraduate nursing program in Turkey. The sample was randomly divided into 2 subgroups, and construct validity analyses were performed on separate data sets using exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). Initial CFA findings indicated that the original 3-factor structure was not adequately supported in the Turkish sample. Therefore, the latent structure of the scale was reexamined using EFA. The revised analyses supported a unidimensional 9-item structure, accounting for 53.71% of the total variance. CFA findings demonstrated acceptable model fit (χ2/df= 2.344; GFI = 0.926; AGFI = 0.887; IFI = 0.958; CFI = 0.957; TLI = 0.945; RMSEA = 0.077; SRMR = 0.049). Internal consistency reliability was high (Cronbach's α = 0.912; McDonald's ω = 0.915). Test-retest reliability (n = 10; 4-wk interval between administrations) was also found to be good (ICC = 0.82), although these findings should be interpreted with caution given the very small retest sample size. The findings provide initial psychometric evidence supporting the validity and reliability of the revised AIERS in a sample of Turkish nursing students for assessing AI-related ethical reflection. Further validation studies in diverse samples and analyses of measurement invariance are recommended.