ChatGPT-Assisted Learning and Academic Performance Among University Students

Abstract

This study investigates the link between the use of ChatGPT in learning and students' academic performance at the University of Guyana. The rise of generative artificial intelligence (AI) has transformed the way students seek information, work on assignments, develop ideas, and handle academic tasks. This study employed a mixed-methods design comprising surveys, interview responses, and academic performance records. It is guided by the following: the Technology Acceptance Model and the Constructivist Learning Theory. The key areas of concern for this analysis are students' frequency of using ChatGPT, perceived usefulness, learning motivation, understanding of course content, academic achievement, and potential risks (overdependence, decreased critical thinking, and academic dishonesty). The results highlight potential for using ChatGPT thoughtfully and strategically to aid comprehension, productivity, writing preparation, and personalized learning. The results reveal that students who use ChatGPT for learning support feel more engaged and more confident in their learning, whereas those who use it as a crutch for copying answers perceive it as plagiarism. Students report higher engagement and confidence in their learning when using ChatGPT as a learning aid, and higher rates of plagiarism when using it as a learning crutch. It also demonstrates, however, that this is where dependency and ethical issues can occur if it is not used properly. The authors recommend that universities incorporate ChatGPT via clear policies, AI literacy courses, policy overhauls, and pedagogical support. While ChatGPT can be a valuable tool for academic success, it is essential to use it responsibly, in conjunction with thoughtfully engaging and critical thinking, and with honesty.