Determinants of continuous use intention

: an empirical study of ChatGPT among students in German higher education

Abstract

The rapid integration of generative AI like ChatGPT into German higher education has outpaced empirical understanding of its long-term use. Beyond initial adoption, a critical gap exists regarding factors driving students to sustainably embed these tools into academic workflows. This thesis addresses this gap by identifying and empirically analyzing the key determinants of students' Continuous Use Intention (CUI) of ChatGPT for academic purposes within this national context. A quantitative, cross-sectional survey was conducted, and a series of linear regression analyses were performed to test an integrated model combining constructs from technology acceptance theories with AI-specific factors. The results revealed a hierarchy of factors predicting CUI. Habit emerged as the most powerful determinant of CUI, followed by Perceived Usefulness and AI Literacy, all showing significant positive effects. Contrary to established models, Satisfaction and Trust in AI did not have a significant direct influence in the final model. However, Perceived Accuracy was a strong predictor of both, establishing it as a foundational prerequisite. This research concludes that sustained use of ChatGPT by German students is primarily a function of routinized behavior, perceived value, and user competence. Findings suggest that traditional drivers like Satisfaction and Trust act as indirect, foundational "gateway" conditions rather than direct drivers in a mature usage context.

Date of Award	23 Oct 2025
Original language	English
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Miguel Rita (Supervisor)