Mapping ChatGPT trends in educational evaluation: A bibliometric analysis to support SDGs 4

This study conducts a bibliometric analysis to map trends and patterns in the application of ChatGPT for educational evaluation. Data were collected from the Scopus database on June 7, 2025, using a specific search string and filtered through the PRISMA framework. Results show a significant surge in publications in 2024 (1,191 documents) compared to 2023 (351) and a decline in 2025 (682, partial year), indicating strong but possibly stabilizing interest in the topic. Despite the growth in publication volume, citation impact per article decreased markedly over time, with the highest average citations per article observed in 2022 and 2023. Key thematic areas identified include assessment practices, artificial intelligence, and educational measurement. Thematic mapping reveals that while ChatGPT remains central to these themes, integration of AI with equitable and ethical assessment practices is still underdeveloped. The most globally cited article in this field, by Gilson (2023), has accumulated 1162 citations, underscoring its foundational influence. This study provides valuable insights into evolving research landscape of ChatGPT in educational evaluation, identifies existing gaps, and emphasizes the need for sustained high-quality research to foster innovative and inclusive assessment strategies aligned with Sustainable Development Goal 4.