Behavioral Modeling of Generation Z Students’ ChatGPT Use and Academic Independence Using Supervised Machine Learning Techniques

Abstract

This research investigates the behavioral factors that influence academic independence in Generation Z students when utilizing AI tools, such as ChatGPT, for learning and research purposes. By incorporating constructs from the Technology Acceptance Model (TAM), the Unified Theory of Acceptance and Use of Technology (UTAUT), and the AI Literacy framework, the study utilized Efficient Logistic Regression and efficient linear SVM algorithms to categorize students' levels of independence as High, Moderate, or Low based on behavioral indicators. A dataset comprising student responses to variables such as Perceived Ease of Use (PEOU), Perceived Usefulness (PU), Social Influence (SI), Habitual Engagement (HE), Convenience (CO), Ethical Consideration (EC), and AI Literacy (AI LIT) was examined using 5-fold cross-validation. The results indicated that both models attained high validation accuracies of 96.12% for Logistic Regression and 97.41% for Linear SVM, with ROC–AUC values approaching 1.0, thereby affirming their outstanding discriminative performance. AI Literacy, Ethics Index, and Convenience were identified as the most influential predictors of academic independence through the Kruskal–Wallis feature importance analysis. According to the findings, students who possess greater AI literacy and ethical awareness exhibit enhanced autonomy and responsible use of AI, whereas habitual behaviors and those influenced by social factors have a diminished effect. This research highlights the effectiveness of machine learning in modeling educational behavior. It also adds to the expanding conversation about how to incorporate AI responsibly in higher education.

Ähnliche Arbeiten