Psychometric properties of the Persian version of attitudes towards artificial intelligence in work, healthcare, and education (ATTARI-WHE)

Abstract

<title>Abstract</title> Introduction Artificial intelligence (AI) is increasingly shaping work, healthcare, and education. For further evaluation of AI in Work, Healthcare, and Education, we aimed to translate, culturally adapt, and evaluate the psychometric properties of the Persian version of ATTARI-WHE. Methods This cross-sectional study conducted between November 2024 and March 2025 among 118 participants, including medical students, faculty members, staff, and employees from four Iranian medical universities. The forward–backward translation method was applied, followed by expert content validation and student face validation. Construct validity was examined using confirmatory factor analysis (CFA). Reliability was assessed with Cronbach’s alpha, while convergent validity was evaluated through average variance extracted (AVE) and composite reliability (CR) Results Content validity ratio (CVR = 0.73) and content validity index (CVI = 0.92) confirmed item adequacy and clarity. Face validation showed high comprehensibility. CFA supported a three-factor structure aligned with the original model, although only the cognitive item of the health domain loaded independently. the adequacy of sampling was acceptable (KMO = 0.864), and Bartlett’s test was significant. Factor loadings exceeded 0.5, model fit indices (SRMR &lt; 0.08, explained variance &gt; 60%) indicated good fit Conclusion The Persian version of ATTARI-WHE demonstrates satisfactory validity and reliability, making it a suitable tool for assessing attitudes toward AI in work, healthcare, and education within persian contexts.

Ähnliche Arbeiten