Document Type : Research Paper
Authors
1 Ph.D. Student in Educational Technology / Allameh Tabatabaei University, Tehran / Iran
2 PhD student in English Language Teaching/ Shahrekord University, Shahrekord/Iran
3 PhD student in Educational Management, Shiraz University
Abstract
Given the rapid evolution of artificial intelligence (AI) and the need to assess AI skills, it is vital to validate indigenous instruments to measure AI literacy. This study aimed to psychometrically validate Ng et al.’s (2024) AI literacy questionnaire in a sample of Iranian undergraduates as a non-Western context (Iran). This research employed a descriptive-correlational design with a psychometric approach. The questionnaire was localized through a back-translation process and expert review by specialists in educational technology, computer science, and English language teaching. Initial reliability was evaluated by conducting a pilot study involving 30 students using Cronbach’s alpha and test-retest methods. The final version of the questionnaire was administered to 151 undergraduate English education students at Shahrekord University who were selected via stratified random sampling. To examine construct validity and reliability, confirmatory factor analysis and structural equation modeling based on covariance were conducted using AMOS and SPSS software. The results of the first-order confirmatory factor analysis indicated that the factor loadings of the items were acceptable and significant. In addition, the composite reliability (0.84) and average variance extracted (0.62) indices indicated satisfactory convergent validity of the constructs. The second-order confirmatory factor analysis was run to evaluate the AI literacy scale’s affective, behavioral, cognitive, and ethical dimensions. The model fit indices indicated an acceptable model fit. Overall, the results revealed that the four-dimensional (ABCE model) structure of AI literacy is confirmed in an Iranian sample. Furthermore, the descriptive analysis of the dimensions showed the affective dimension received the highest mean value, highlighting the importance of the motivational and attitudinal factors in forming AI literacy. In sum, the Persian version of the AI literacy questionnaire with 32 items demonstrated acceptable validity and reliability and can be used as an authentic tool for measuring AI literacy among Iranian university students.
Keywords
)