Validation and Reliability of the Turkish Adaptation of the Artificial Intelligence Literacy Scale (AILS) for Healthcare Professionals

Artificial intelligence (AI) literacy is the ability to understand, use, monitor, and critically evaluate AI applications without needing to create AI models. As many professionals outside technical fields frequently engage with AI, this literacy is essential. This study evaluates the validity and reliability of the Turkish adaptation of the Artificial Intelligence Literacy Scale (AILS) to measure AI literacy among healthcare professionals in Türkiye. The study included 210 healthcare professionals—physicians, dentists, nurses, and midwives—aged 18 and older. AILS is a seven-point Likert scale with 12 items divided into four factors: “awareness,” “usage,” “evaluation,” and “ethics.” Cronbach’s alpha indicated good internal consistency for the scale, with a coefficient of 0.85. Confirmatory Factor Analysis (CFA) showed satisfactory fit indices: <i>χ</i>²/<i>df</i> = 1.665, Comparative Fit Index = 0.968, Goodness of Fit Index = 0.944, Tucker-Lewis Index = 0.955, Standardized Root Mean Square Residual = 0.040, and Root Mean Square Error Estimate = 0.056. The 12-item, 4-factor AILS shows a strong fit and robust structure for the sample data. Our study validates the Turkish version of the AILS as a reliable tool for assessing artificial intelligence literacy among Turkish healthcare professionals.

人工智能素养（Artificial Intelligence Literacy）指无需构建人工智能模型，即可理解、应用、监测并批判性评价人工智能应用的能力。鉴于众多非技术领域专业人员时常接触人工智能，该素养显得尤为必要。本研究旨在评估经土耳其本土化改编的人工智能素养量表（Artificial Intelligence Literacy Scale，AILS）用于测量土耳其医疗卫生专业人员人工智能素养的效度与信度。本研究共纳入210名年满18岁的医疗卫生专业人员，涵盖医师、牙医、护士及助产士。AILS为7点李克特（Likert）量表，包含12个条目，分为“认知”“应用”“评估”“伦理”四个维度。克隆巴赫α系数（Cronbach’s α）为0.85，表明该量表具备良好的内部一致性。验证性因子分析（Confirmatory Factor Analysis，CFA）结果显示各项拟合指标均表现优异：卡方自由度比（χ²/df）为1.665，比较拟合指数（Comparative Fit Index，CFI）为0.968，拟合优度指数（Goodness of Fit Index，GFI）为0.944，塔克-刘易斯指数（Tucker-Lewis Index，TLI）为0.955，标准化残差均方根（Standardized Root Mean Square Residual，SRMR）为0.040，近似误差均方根（Root Mean Square Error of Approximation，RMSEA）为0.056。包含12个条目、4个维度的AILS在本次样本数据中展现出优异的拟合效果与稳定的结构效度。本研究证实，经土耳其本土化改编的AILS可作为评估土耳其医疗卫生专业人员人工智能素养的可靠工具。