Abstract

This study aims to map research trends in STEM (Science, Technology, Engineering, and Mathematics) education at the primary school level from 2016 to 2025. The analysis adopts a bibliometric approach utilizing data sourced from Dimensions.ai, processed using VOSviewer software. Over 180,000 publications were evaluated based on annual trends, document types, fields of study, and keyword co-occurrence visualizations. The findings reveal a significant increase in the number of publications, peaking in 2023. Scientific articles emerged as the most dominant document type, followed by book chapters and monographs. Education, social sciences, and psychology were the most frequently represented thematic areas, indicating a strong interdisciplinary orientation. Network visualization identified three major clusters: pedagogical, demographic, and psychosocial. Meanwhile, overlay visualization highlights a shift in research focus from student characteristics to implementation-oriented and contextual issues in the post-pandemic era. These findings underscore the importance of developing adaptive, contextual, and evidence-based STEM learning policies and practices in primary education. This study offers strategic insights for researchers, educators, and policymakers in shaping the future direction of more relevant and sustainable STEM education.

Concepts :

Access to Document

Citations by Year