Abstract:

After a year since ChatGPT came to light, an evaluation of student perspectives and usage concerning large language models (LLMs) becomes necessary. As LLMs continue to influence education, understanding their effects on student learning experiences emerges as an important aspect. This contribution analyzes the perceptions and utilization of LLMs among students in the fields of Education and Information Technology at the University of Andorra. Using a survey as its primary instrument, this research reveals the different ways in which students interact with artificial intelligence (AI) and their attitudes toward these kind of tools. The study is mainly quantitative, with some questions employing a qualitative approach. Preliminary findings show a range of attitudes, from enthusiastic adoption to skepticism. The surveyed students express advantages in receiving immediate and personalized support for academic aspects, assistance in content generation, and improvement in writing skills. However, concerns about accuracy emerge as a notable weakness. This mixed-methods study illustrates distinct usage patterns between disciplines and academic levels, showing the frequency and purpose of LLMs utilization. Given the bias in the distribution of women and men across the two fields under study, gender data should be interpreted in relation to the specific field rather than gender itself. The results of this research contribute to a deeper understanding of student behavior in regard of content generation tools, providing valuable information for educational institutions seeking to effectively incorporate those tools into their curricula, and it also facilitates further research on the influence of AI in higher education.

Keywords:

Artificial intelligence, Large language model, Higher education.