Identifying Diverse Attitudes Towards Artificial Intelligence Among Middle School Students Through Latent Profile Analysis

ABSTRACT This study aims to determine the student profiles of middle school students by examining their attitudes towards artificial intelligence (ATAI) and their computational identity (CI) structures. CI, which can be considered a domain‐specific form of cognitive identity in programming and AI contexts, was examined in four basic dimensions such as participation in programming, belonging, realization, and goal setting. The study examined ATAI as cognitive, affective, and behavioural components. The data obtained from 453 students studying in middle school in Turkey were evaluated with latent profile analysis (LPA). As a result of the study, it was found that the short form of the AI attitude scale also presented strong construct validity. On the other hand, three different student profiles were revealed according to LPA. Profile 1 represents students with a low level of AI attitude and a low level of cognitive identity, while Profile 2 consists of students with a high level of attitude and cognitive identity. Profile 3, on the other hand, consists of students who are in between these two groups and have medium level values. The multinomial logistic regression analysis was conducted as a supplementary within‐model comparison to examine how ATAI and CI components are structurally associated with profile differentiation, rather than to establish predictive relationships. The findings indicated that CI components represent the primary dimensions along which the profiles differ. On the other hand, demographic variables such as age, gender, or AI experience were not significantly associated with profile membership.