Prospects for ChatGPT Application in Higher Education: A Scoping Review of International Research

Abstract

Introduction. The arrival of advanced Generative Pre-Trained Transformer (GPT) in late 2022 resulted in a wide spread of the technology, capable of generating texts of any genres, complexity, and styles. Written communication embraces all spheres of human educational and professional activities that caused a lightning speed at which the technology was introduced into education, science, journalism, mass communication, medicine, etc. The academic community is vigorously discussing the plausible consequences and potential of ChatGPT for various domains, especially education, medicine, and science at large. The review seeks to analyze international publications on ChatGPT’s impact and potential in higher education and to identify the key lines of research. Materials and Methods. Based on the PRISMA methodology the authors selected 190 articles and reviews indexed in the Scopus database with the inclusion criteria in focus (period, languages, field, themes). The publications were analyzed and distributed among seven thematic clusters. Results. As a result of the analysis of the publications, the following clusters were formed: higher education (71 documents), education in various areas (37 documents), writing and language learning (24 documents), assessment in educational systems (22 documents), medical education (18 documents), ethical issues and academic integrity (15 documents), and perception of ChatGPT on part of students and educators (12 documents). Discussion and Conclusion. The analysis of the ChatGPT potential and negative aspects in higher education allows for singling out the subject areas where the most essential transformations will take place: assessment of knowledge and skills, information credibility, scope of competence and skills formed in higher education, ethical issues of research, various aspects of professional training. The authors’ conclusions correlate with the reviews on ChatGPT published earlier. Addressing the new challenges requires the system of education to be transformed, new forms of learning and assessment to be introduced. The review results may be applied in further research on the topic. Given a skyrocketing rise in the number of publications on ChatGPT, researchers ought to keep track of new research in the emerging subject area.

Keywords

• higher education
• artificial intelligence
• chatgpt
• generative pre-trained transformer
• natural language processing
• large language model
• writing