Development of artificial intelligence literacy scale for gifted secondary school level gifted students

This study aimed to develop a valid and reliable measurement instrument to assess the artificial intelligence literacy of gifted middle school students. During the scale development process, the relevant literature was reviewed and an initial item pool consisting of 30 items was generated. Content validity was established through expert opinions, followed by psychometric analyses conducted during the pilot and main application phases. The study was carried out with a total of 381 gifted middle school students enrolled in 32 Science and Art Centers (BİLSEM) selected through purposive sampling, with at least one province represented from each of Türkiye's seven geographical regions. The pilot study included 261 students (114 females, 147 males), while the main application involved 120 students (50 males, 70 females). Data were collected during the 2024–2025 academic year. Data collection instruments consisted of a "Personal Information Form" and the "Artificial Intelligence Literacy Scale for Gifted Middle School Students" Construct validity of the scale was examined using item–total correlations, Exploratory Factor Analysis (EFA), and Confirmatory Factor Analysis (CFA). Reliability analyses were conducted using Cronbach's alpha and split-half reliability coefficients, while item discrimination was confirmed through upper 27%–lower 27% group comparisons. As a result of the analyses, the scale was developed as a 5-point Likert-type instrument consisting of two subdimensions—"Evaluation" with 5 items and "Awareness" with 6 items—comprising a total of 11 items. Cronbach's alpha reliability coefficients were calculated as α = .85 for the Evaluation subdimension, α = .77 for the Awareness subdimension, and α = .87 for the overall scale. Split-half reliability analyses based on both pilot and main application data indicated that inter-half correlations, Spearman–Brown, and Guttman coefficients were high across both applications, demonstrating strong internal consistency of the scale. In addition, Average Variance Extracted (AVE) and Composite Reliability (CR) values calculated following CFA supported the acceptable levels of convergent validity and reliability of the scale. The scale is scored on a range from 1 (Strongly Disagree) to 5 (Strongly Agree), with total scores ranging from 11 to 55. The results indicate that the developed scale is a valid and reliable instrument for measuring the artificial intelligence literacy of gifted middle school students. Furthermore, artificial intelligence literacy levels were found not to differ significantly according to gender, grade level, participation in coding education, or participation in software education.