Ordinal Logistic Regression Analysis of Factors Influencing University Students' Use of AI Technologies in Education

Abstract

Through this study, the researchers seek to examine the factors influencing university students' frequency of use of artificial intelligence (AI) tools in education in the Kurdistan Region. Understanding the determinants of effective utilization is essential as AI technologies rapidly integrate into higher education. To address this, a descriptive quantitative design was applied, and data were collected from 240 students across universities using a structured questionnaire. The dependent variable frequency of AI use (never, sometimes, always) was analyzed through ordinal logistic regression (OLR).

The model demonstrated good fit (Nagelkerke's R² = 0.421) and satisfied the proportional odds assumption, achieving a classification accuracy of 73.8%, particularly strong for the sometimes category. Significant predictors of AI use included family economic situation, students' perceptions of AI as a tool to improve learning and solve complex problems, proficiency in educational software, and interest in AI tools. Conversely, ethical concerns and reliance on personal skills negatively influenced students' engagement with AI in education.

The findings highlight the critical role of economic, technical, and behavioral factors for AI adoption. They suggest that enhancing digital literacy and addressing ethical concerns can strengthen the effectiveness of AI integration. The study contributes theoretically by extending research on AI adoption through ordinal regression modeling and provides practical insights for policymakers and universities to foster responsible and equitable AI use in higher education.