From Bytes to Attitudes: Unpacking Students' Cognitive, Affective, and Behavioral Responses to AI

Abstract

This study aims to explore secondary and high school students' attitudes towards artificial intelligence (AI) in affective, behavioural, and cognitive dimensions and to analyse the relationships between these attitudes using multilevel and structural equation modeling. The study was conducted with quantitative data obtained from 702 students and 8 English as a Foreign Language teachers. Multilevel models were constructed considering students' gender, grade level, previous use of AI, and teacher-related classroom effects. The findings demonstrated that students who had previously used AI exhibited more positive attitudes in all attitude dimensions. In addition, cognitive attitudes significantly predicted affective attitudes, affective attitudes significantly predicted behavioural attitudes, and affective attitudes played a partial mediating role in this relationship. Although the teacher effect was limited, it provided a significant contribution to the model. These results reveal that students' attitudes towards AI are influenced not only by individual but also by contextual factors, and emphasize that strategic planning for AI integration in educational environments needs to be designed by considering multilevel structures.­­­­­