Use and perception of LLM by Higher Education students and teachers - Systematic Literature Review

Abstract

Artificial Intelligence (AI), especially Large Language Models (LLM), is currently a crucial topic in education, generating intense debates. Its generative capacity drives studies and discussions, making it crucial to understand the knowledge, perception, and use of AI by students and teachers. The aim of this Systematic Literature Review (SLR) is to explore scientific studies that have been developed in the context of Higher Education regarding the perception and use of LLM by students and teachers. We seek to identify: i) the teachers and students’ perception of LLM; and ii) the use of these models by teachers and students. Adopting the PRISMA protocol, we identified 56 works. After applying the selection criteria, we focused on 14 articles. Data sources spanned DOAJ, EBSCO, ERIC, SCOPUS, and Web of Science and Google Scholar. The analysis of these 14 articles revealed that students and teachers are aware of LLM and use them out of curiosity, to generate content, to inspire themselves or to perform tutoring roles. The results also revealed that students show greater receptivity to their use. As for perceptions, they recognise the opportunities and challenges associated with LLM. However, teachers demonstrate scepticism, fearing an inappropriate use of generative AI, mainly in relation to integrity, ethics, safety, and privacy.