Answers are grounded in this paper and cite their sources.
Generative artificial intelligence (AI) is increasingly being integrated into medical education, creating both opportunities and concerns for educators. Responsible implementation requires an understanding of educators’ perceptions of its limitations. This study aimed to develop and psychometrically assess the Generative AI Educator PerspectoMeter Scale (GAIE-PM) to measure medical educators’ perceptions of generative AI limitations in medical education. This descriptive scale-development study was conducted among medical educators at medical sciences universities in Iran in 2024. Following a modified Association for Medical Education in Europe (AMEE) Guide-informed approach, the scale was developed through a systematic literature review, a qualitative study with Iranian experts in three virtual focus groups, item generation, expert validation, pilot testing, and psychometric evaluation. After expert validation, 18 items were retained for pilot testing. Data were collected from 180 medical educators. Principal Factor Analysis (PCA) with Varimax Rotation was used to examine the preliminary factor structure, and Partial Least Squares Structural Equation Modeling (PLS-SEM) was used to assess the measurement model. Test–retest reliability was assessed in a subsample of 30 medical educators over a two-week interval using the intraclass correlation coefficient (ICC) and Pearson correlation coefficient. To ensure a more robust assessment of internal consistency, McDonald’s omega (ω) was calculated alongside Cronbach’s alpha. The preliminary PCA supported a three-factor structure consisting of Pedagogical Limitations, Technical and Access Limitations, and Ethical and Quality Limitations, explaining 48.37% of the total variance. The final PLS-SEM measurement model retained 14 items across the three factors. Internal consistency was acceptable across the three factors, with Cronbach’s alpha values ranging from 0.69 to 0.81, a mean Cronbach’s alpha of 0.75, composite reliability values ranging from 0.81 to 0.86, and a mean McDonald’s omega of 0.81. Average variance extracted values ranged from 0.505 to 0.570, indicating acceptable convergent validity. Discriminant validity was supported using the Fornell–Larcker criterion. Test–retest reliability showed good temporal stability, with an ICC of 0.82 for the total scale score and factor-level ICC values ranging from 0.76 to 0.88. Pearson correlation coefficients between the two administrations ranged from 0.76 to 0.84 ( p < 0.001). The GAIE-PM Scale demonstrated acceptable psychometric properties for assessing medical educators’ perceptions of generative AI limitations in medical education. The scale may help identify pedagogical, technical/access, and ethical/quality-related concerns that should be considered when integrating generative AI into educational settings. Further validation in larger and more diverse samples is recommended.
No takes yet. Share an insight, caveat, or question.
Kohan et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: