This study maps the development and intellectual structure of math learning anxiety research using a bibliometric approach. Data were obtained from the Dimensions database (indexed from Scopus) covering publications from 2020 to 2025. Following the PRISMA framework (identification, screening, eligibility), a total of 127 peer-reviewed articles were initially identified, of which 94 articles met the final inclusion criteria. VOSviewer version 1.6.20 was used to generate co-occurrence network, overlay, and density visualizations. Results show a consistent annual increase in publications. Network visualization reveals a strong association between math anxiety and learning outcomes, confirming its role as a significant affective factor. Overlay analysis indicates a thematic shift from early keywords (test, outcome, ability) toward more recent topics (engagement, performance, self-efficacy, academic achievement). Density visualization confirms that student, mathematics, and mathematics anxiety are the most dominant terms, while technology and engagement appear as emerging but low-density themes. These findings confirm that bibliometric visualization effectively maps research trends and intellectual structures, offering a strategic basis for future research agendas.
Copyrights © 2026