AI and Scientific Thinking: A Bibliometric Study
TL;DR
A bibliometric analysis was conducted to examine publication trends, disciplinary distributions, international collaborations, and their alignment with the United Nations Sustainable Development Goals (SDGs) to indicate a rapid acceleration of research output post-2019.
Abstract
This study investigates the evolution and conceptual framework of research connecting artificial intelligence (AI) with scientific thinking in science education. Utilizing a dataset of 21,874 publications retrieved from the Web of Science, a bibliometric analysis was conducted to examine publication trends, disciplinary distributions, international collaborations, and their alignment with the United Nations (UN) Sustainable Development Goals (SDGs). The results indicate a rapid acceleration of research output post-2019, marking the consolidation of AI-related inquiry within educational and cognitive domains. Although the field remains anchored in computer science and engineering, there is a notable rise in interdisciplinarity through contributions from education, psychology, and philosophy. Global participation is concentrated in technologically advanced regions, while institutional patterns highlight collaborative networks among leading universities and research centers. SDG classifications reveal connections with health, education, innovation, and environmental sustainability. Overall, the study depicts a dynamically expanding domain where AI functions not only as a technological innovation but also as an epistemic framework that reshapes reasoning, inquiry, and reflective judgment in science education. The findings offer both a conceptual and empirical basis for linking technological capabilities with the epistemic and ethical dimensions of learning in an AI-driven world.