Binar Kurnia Prahani, Khoirun Nisa’, Suliyanah Suliyanah, Utama Alan Deta
Abstract
Research focus of reviews trends and research on implementing ChatGPT in science, technology, engineering, arts, and math (STEAM) learning. It emphasizes the importance of deep learning and 21st-century skills in education, highlighting the limitations of ChatGPT in accuracy and credibility. The authors analyzed 204 STEAM education documents, revealing that 65% focused on technology education and less than 3% on art and mathematics education. The articles written in technology scope are the most widely circulated. The most productive region is the United States, which has three productive authors. The most productive authors are Ray (India) and Wang (Macao), who have the highest h-index. The United States and United Kingdom are the most productive affiliations. Many types of research on ChatGPT in STEAM education include a survey with several participants of different education levels. Social science is the most popular subject area. The Journal Nature is the primary source for this research. Several research highlighted artificial intelligence, ChatGPT, and human keywords. This study highlights the potential of ChatGPT in STEAM, suggesting further research on student behavior, learning designs, and credibility concerns. It suggests collaboration with Google Scholar or Web of Science data for in-depth analysis.
