ChatGPT na educação básica: um estudo bibliométrico dos 100 artigos mais citados

Abstract

The integration of ChatGPT into K-12 education has sparked unprecedented scientific growth since the tool's public launch in 2022. However, the scattered nature and rapid volume of publications make it difficult to identify the most impactful works and the trends shaping the field. This study aimed to identify, characterize, and analyze the 100 most cited scientific articles regarding the use and impact of ChatGPT in K-12 education, revealing the intellectual structure, collaboration networks, and emerging trends within this literature. It is a retrospective, quantitative-qualitative study based on bibliometric analysis and scientific mapping. Data were collected from the Web of Science and Scopus databases, with the 100 most cited studies selected after removing duplicates and screening for scope. Analyses were conducted using Bibliometrix software. Results revealed a 33.33% annual growth in the field's output, dominated by China (n=78) and the United States (n=73), alongside significant under-representation of the Global South. The journal Computers and Education: Artificial Intelligence emerged as the primary dissemination outlet. Thematic analysis identified AI literacy, ChatGPT, and generative AI as established "motor themes," whereas the teacher's role and ethical aspects remain peripheral niches, revealing critical gaps between dominant scientific output and the field's pedagogical demands. Author collaboration networks appeared fragmented, indicating a field in a pre-paradigmatic stage. The study concludes that, while the field demonstrates significant dynamism and impact, it lacks longitudinal research, greater geographic diversity in output, and agendas that position teachers and ethics as central research subjects.