Introduction

Augmented Reality (AR) has emerged as one of the most promising technologies in the field of education. In mathematics education, AR has the ability to present abstract material in a visual and interactive way, which can enhance students’ understanding of complex mathematical concepts . This technology offers a new approach for students to learn by connecting digital elements to the real world through 3D visualization.

The use of AR in education has grown rapidly over the past decade, particularly in the fields of computer science and higher education in countries such as Spain, the United States, and the United Kingdom .Although research on AR in mathematics education continues to grow, there are still many opportunities to optimize this technology to enhance learning effectiveness. Several studies have shown that AR can increase student engagement and improve conceptual understanding in mathematics learning . However, more research is needed to explore its long-term effectiveness and the application of AR in various mathematics teaching strategies.

Most of the existing literature focuses more on technical aspects or other disciplines, such as early childhood education, where AR has been used to enhance motivation and knowledge acquisition . On the other hand, topics such as science, technology, engineering, and mathematics (STEM), problem-solving, and teacher noticing have become major trends in mathematics education research . However, based on the accessible literature, studies that explicitly integrate these topics with AR remain limited .In fact, AR technology itself has advanced significantly from relying on physical markers to markerless AR, which uses GPS, sensors, and natural feature tracking to overlay digital content in mathematics learning contexts . This technological evolution opens up greater opportunities to integrate AR into more contextual and innovative mathematics teaching approaches.

Although research on AR in mathematics education continues to grow, there are still gaps in in-depth exploration regarding its effectiveness across different educational levels, optimal implementation strategies, and methods for evaluating its impact on student learning outcomes. Previous studies have also not comprehensively mapped publication trends, academic collaboration, and the direction of research development in this field. Therefore, this study contributes by analyzing publication patterns, collaboration networks, and current research focuses on filling gaps in the literature and providing broader insights into the role of AR in enhancing mathematics learning.

The purpose of this study is to conduct a bibliometric analysis of the development of AR use in mathematics education over the past ten years (2015–2024). This time span was chosen because it reflects a period during which AR technology has been increasingly applied in education, with a significant surge in publication trends, allowing for a more comprehensive analysis of patterns and collaborations. This study aims to map trends and collaborations among researchers, institutions, and countries, as well as identify the main emerging topics in the use of AR in mathematics education. In addition, it seeks to explore the long-term impact of AR on mathematics learning, focusing on its effectiveness in enhancing conceptual understanding, student engagement, and learning outcomes. The results of this study are expected to fill gaps in the literature and provide clearer guidance for future research on the implementation of AR in mathematics education.

Methodology

The bibliometric method is a quantitative analysis technique used to evaluate scientific publication patterns based on bibliographic data, such as the number of publications, author collaborations, and research trends . The author employed a bibliometric method to identify publication patterns, author collaborations, and research trends related to the use of AR in mathematics education. The analysis was conducted using data from Scopus for the period 2015–2024, which was selected due to the significant increase in AR-related research in the field of education during this time. Scopus was chosen as the primary source because it includes high-quality journals and provides appropriate metadata for bibliometric analysis. Other databases, such as Web of Science and IEEE Xplore, were not included due to limited access, while Google Scholar was excluded because of its lack of structured metadata.

The research process was carried out in five stages: (1) Search procedure, (2) Bibliographic screening, (3) Bibliographic completeness, (4) Bibliometric analysis, and (5) Research process flow diagram.

Search Procedure

The first step in this study was a literature search using the Scopus database on July 31, 2024. The search used the keywords "augmented reality," "education," "mathematics," "math," and "mathematical" applied to the title, abstract, and keywords of publications. This search yielded 542 documents, including various types of publications such as journal articles, conference papers, and reviews. However, due to the specific keyword combinations used, some publications with more specific terms may not have been captured in this analysis.

Bibliographic Screening

After identifying the documents, the next step was to apply inclusion and exclusion criteria to filter the results. The following table summarizes the criteria used in this study:

Table 1. Inclusion and Exclusion Criteria

The use of English as an inclusion criterion is based on its wide reach in the global academic community and its dominance in bibliometric databases. However, this may result in the exclusion of relevant studies in other languages. After screening, the number of documents analyzed was reduced to 194 journal articles, published between 2015 and 2024. This screening ensured that the documents met clear academic standards and were relevant to the research objectives.

Bibliographic Completeness

This stage ensures that each selected document contains complete metadata that can be accurately read by bibliometric software such as VOSviewer. Verified metadata includes author names, institutional affiliations, keywords, and publication sources. Documents lacking sufficient metadata or with unreadable entries are excluded from the analysis to maintain the accuracy and reliability of the mapping and visualization process.

Bibliometric Analysis

The final stage of this research method is the bibliometric analysis, conducted by the authors using VOSviewer version 1.6.19. VOSviewer was chosen for its ability to visually map bibliometric data and reveal patterns and trends in AR research within mathematics education, providing readers with a clear overview of themes, collaborations, and global developments. The analysis process involved several key steps: (1) exporting data from Scopus in CSV format, (2) cleaning and filtering the data, (3) importing the data into VOSviewer for network mapping, (4) generating visual analyses of keywords, academic collaborations, and publication trends, and (5) verifying the results by comparing emerging patterns with relevant literature. These visualizations help identify global collaboration patterns and major trends in AR research in mathematics education over the past decade.

Research Process Flow Diagram

Figure 1 illustrates the data extraction and filtration process used in this study, consisting of four main stages: identification, screening, eligibility, and inclusion. From the initial 542 documents, a total of 348 were excluded based on exclusion criteria, including unpublished works (e.g., preprints, conference abstracts), articles in languages other than English, documents not specifically addressing AR in mathematics education, and unsuitable document types such as editorials or book chapters. After the screening process, 194 of the most relevant journal articles were included in the bibliometric analysis.

Figure 1. Four-Stage Data Extraction and Screening Flow Diagram

This process ensures that studies meeting the inclusion criteria are incorporated into the analysis, resulting in more systematic and relevant research findings. The flow diagram provides a clearer overview of research patterns on Augmented Reality (AR) in mathematics education.

Findings/Results

Number of Documents Retrieved and Filtered

The initial search yielded 542 documents relevant to research on AR in mathematics education. After filtering by document type and selecting only journal articles, the number was reduced to 194, as shown in Figure 1. This finding indicates that research on AR in mathematics education has received significant attention in the academic world, particularly through journal articles published between 2015 and 2024.

Analysis Based on Document Types

Figure 2 presents the types of documents analyzed in this study. The majority are journal articles (88.7%), indicating that research on AR in mathematics education is predominantly published in scholarly article form. In addition, reviews account for 7.7%, reflecting efforts to synthesize findings from previous studies. Other document types include errata (2.1%), conference papers (0.5%), data papers (0.5%), and notes (0.5%), which, although limited in number, still contribute to the related literature. Non-article documents detected in the initial dataset were excluded during the screening process to ensure that only peer-reviewed journal articles were included in the analysis.

Figure 2. Document Types

Analysis of Publication Trends by Year

Figure 3. Publication Trends from 2015 to 2024

Figure 3 illustrates the publication trends on AR in mathematics education from 2015 to 2024. While the number of publications was relatively high in 2015 and 2018, stagnation occurred in 2016 and 2017, with no notable growth. However, after 2018, the number of publications began to increase steadily, with a significant surge in 2020 and 2021. This sharp rise can be attributed to the widespread shift toward digital learning during the COVID-19 pandemic, which accelerated the adoption of innovative technologies such as AR in education.

Analysis Based on Journal Sources

Figure 4 presents the journal sources that have published research on AR in mathematics education. Education Sciences and IEEE Access show a higher and more consistent number of publications since 2020, while Multimedia Tools and Applications experienced a notable surge in 2024. Other journals, such as the British Journal of Educational Technology, International Journal of Emerging Technologies in Learning, and Frontiers in Education, have made smaller contributions, publishing approximately 1–2 articles per year.

This trend indicates that AR in mathematics education has begun to attract attention, particularly in educational technology and digital learning journals, although the number of publications remains limited.

Figure 4. Analysis Based on Journal Sources

Subject-Based Analysis

Figure 5 illustrates the distribution of research subjects related to the use of Augmented Reality (AR) in mathematics education. Social sciences (27.8%) and computer science (25.7%) dominate as the main fields, indicating that AR research focuses not only on technical aspects but also on its social impact and application in digital learning environments. Engineering (14.0%) also plays a significant role, reflecting interest in developing AR-based tools and systems.

Additionally, fields such as mathematics (6.5%), psychology (3.9%), and health professions (3.1%) show that AR is also being used in special education and health-related education. Other fields like materials science (2.6%), physics and astronomy (2.1%), and arts and humanities (1.8%) contribute smaller portions, yet remain relevant for exploring AR’s potential across disciplines. The ‘Other’ category (10.6%) includes diverse subjects that are not specifically classified within the main categories, highlighting the interdisciplinary nature of AR research. This indicates that AR research in mathematics education is inherently interdisciplinary. The subject distribution in Figure 5 reflects the diversity of scholarly approaches panning technology, education, and social sciences which enriches both the understanding and application of AR in educational contexts.

Figure 5. Subject Area Distribution

Analysis Based on Institutions

Figure 6 illustrates the institutions most actively publishing research on Augmented Reality (AR) in mathematics education. Chitkara University, Punjab, India, leads with the highest number of publications, followed by Johannes Kepler University in Austria and Alanya Alaaddin Keykubat University in Türkiye. Other significant contributors include University College Dublin (Ireland), Universiti Malaysia Sarawak, and Universidad de Murcia (Spain).

Additional active institutions in this field include UniversitiKebangsaan Malaysia, Tecnológico de Monterrey (Mexico), Beijing Normal University (China), and Universidad de Salamanca (Spain). The participation of institutions from various regions reflects the global expansion of AR research in mathematics education, with valuable contributions coming from both developed and developing countries.

Figure 6. Institutional Analysis

Country-Based Analysis

Figure 7 displays the distribution of publications by country. The United States leads with the highest number of publications, followed by Malaysia and Spain, indicating that these countries are major hubs for research on AR in mathematics education. China, Germany, and Türkiye also make significant contributions to the development of AR technology in the education sector.

In addition, several other countries such as Indonesia, India, Taiwan, and Italy are also involved in this research, although with a lower number of publications. This global participation indicates that research on AR in mathematics education spans across both developed and developing countries, reflecting its worldwide relevance and growing academic interest.

Figure 7. Document Distribution by Country

Keyword Analysis

Figure 8 presents the keyword mapping in AR research within mathematics education, highlighting dominant terms such as "augmented reality," "mathematics education," and "teaching and learning." Other keywords like "engineering education," "students," and "immersive learning" indicate that AR is also applied across various fields, including engineering and digital pedagogy. This mapping also reveals emerging research themes and clusters, such as learning engagement, STEM education, conceptual understanding, and the integration of AI in learning environments.

Figure 8. Keyword Visualization

To further understand the application of AR in mathematics education, Figure 9 provides a more detailed visualization by focusing on the keyword "mathematics education" from the previous keyword mapping. The results reveal that research in this field revolves around three major clusters. The first cluster focuses on learning methods, reflected in keywords such as "student engagement" and "active learning." The second cluster centers on mathematical concepts, with terms like "geometry" and "problem solving." The third cluster highlights technological innovations in mathematics education, represented by keywords such as "virtual reality" and "interactive simulations." These clusters illustrate the diverse research directions and the integration of AR into pedagogical practices.

Figure 9. Keyword Visualization

Overall, Figure 8 provides a broad overview of AR research trends, while Figure 9 offers a more specific focus on mathematics education. It shows that AR is used not only to enhance interaction in learning but also to support students’ understanding of more complex mathematical concepts.

Figure 9 illustrates that research in AR-based mathematics education is categorized into three main clusters. The first cluster focuses on learning methods, highlighted by keywords such as student engagement and active learning. For example, studies have shown that the integration of AR in active learning strategies significantly enhances student motivation and engagement in STEAM education . The second cluster centers on mathematical concepts, including geometry and problem solving. Several studies revealed that AR supports students in understanding mathematical concepts such as algebra and geometry, while also helping to reduce math anxiety, particularly among students with high anxiety levels . The third cluster highlights technological innovations in mathematics education, represented by keywords like virtual reality and interactive simulations. Research indicates that the integration of AR and VR can improve students’ conceptual understanding of mathematics, although no significant difference was found between the two in terms of effectiveness .

Collaborative Author Analysis

Figure 10 illustrates the collaboration network among authors in the field of AR in mathematics education, highlighting Zsolt Lavicza and A. Mantri as key contributors. Lavicza’s work focuses on integrating AR into STEM education to enhance spatial understanding, geometric visualization, and flipped learning approaches. Meanwhile, Mantri has developed AR-based learning environments for geometry by merging virtual information with physical objects to boost student engagement.

Their studies also explore the evolution of AR approaches in education, highlighting a shift toward competency based learning models. These contributions not only advance AR technology but also emphasize pedagogical effectiveness, student engagement, and integration of AR in tech-enhanced learning models. Systematic reviews show a rise in AR related research, particularly in geometry, problem-solving, and critical thinking skills development underscoring AR’s role beyond technological novelty toward supporting more effective teaching methods.

Figure 10 also maps author clusters, revealing several main groups. Some focus on the impact of AR on student engagement and learning effectiveness, while others investigate system development and integration with AI and VR. However, some research groups remain isolated, suggesting the need for broader academic collaboration. This reinforces the importance of cross-country and interdisciplinary cooperation in advancing AR research. Beyond identifying trends and key actors, the analysis also highlights challenges such as standardizing AR technologies in curricula and adapting them across different educational levels.

Figure 10. Author Collaboration

Highly Cited Studies on AR in Mathematics Education

Table 2 presents the top ten most-cited articles on AR in mathematics education between 2015 and 2024, based on citation data from the Scopus database. This table is included to highlight the most influential works that have shaped the direction of research in this field.

Analyzing these highly cited studies helps identify the prevailing theoretical frameworks, research methods, and educational contexts that have received significant academic attention. It also allows for a better understanding of how AR has been applied in mathematics education, what aspects have gained prominence, and where potential research gaps remain.

By examining the impact and content of these frequently referenced works, this section contributes to the overall argument of the study by mapping the foundational literature and identifying influential contributions that inform current and future research on AR integration in mathematics learning.

Table 2. Most-Cited Articles on AR in Mathematics Education (2015–2024)

Table 2. Continued

Table 2 summarizes the ten most-cited articles discussing the use of AR in mathematics education between 2015 and 2024. AR has been shown to enhance conceptual understanding in STEM learning, although it may increase cognitive load , and it supports active learning and visualization despite technical limitations . AR-based games like Pokémon GO also contribute to improving student attention and social interaction .

In higher education, the use of AR remains limited particularly in mathematics and technology highlighting the need for instructional model-based development . Mobile AR is effective in boosting motivation and reducing math anxiety , and along with VR, can enhance student interest in learning . AR also supports active learning, STEAM skill development, and student inclusion , while aiding high self-efficacy students in grasping complex concepts . Additionally, AR-based dissection lessons improve student interest in STEM fields , and AR applications have proven beneficial for students with special needs by increasing their motivation in applied mathematics learning . These findings highlight AR’s significant contribution to improving mathematics education across various levels and educational contexts.

Conclusion

This study shows that the use of AR in mathematics education has experienced significant growth, with a steady increase in publications since 2016 and a peak in 2024. In terms of document types, journal articles dominate, accounting for 88.7% of the total publications. This reflects a strong preference for peer-reviewed scholarly publications in discussing AR in mathematics education, although further analysis is needed to understand the extent to which this topic has been explored in depth. In addition, leading journals such as Education Sciences and IEEE Access have become key platforms for publishing studies on AR in mathematics education, reflecting the growing academic interest in this topic.

The analysis of research collaboration reveals that institutions such as Chitkara University, Johannes Kepler University Linz, and UniversitiKebangsaan Malaysia are among the leading contributors to the development of studies on AR in mathematics education. Geographically, the United States, Malaysia, and Spain have the highest publication output, indicating strong global interest in integrating AR into educational contexts.

The primary focus of AR research in mathematics education includes student interaction, immersive learning experiences, and the integration of AR with AI and virtual reality (VR) to enhance learning effectiveness. Prominent scholars such as Z. Lavicza, A. Mantri, and B. Haas have played significant roles in expanding international collaboration in this field. Despite the rapid increase in publications, further studies are needed to examine the long-term impact of AR on student learning outcomes, the sustainability of learning systems, and the adaptability of AR across diverse educational contexts.

The use of AI technology in mathematics education has a significant impact on the learning process, enhancing students' understanding, motivation, and engagement, while also aiding them in grasping advanced concepts and reducing math anxiety. AI further supports students with special needs by improving visualization and the effective application of mathematical concepts. These positive effects include improvements in the quality of science education, increased interest in STEM fields, and better performance in mathematics, particularly in algebra and geometry, while fostering an adaptive, independent, and active learning environment. Based on thematic analysis of Table 2, the following framework illustrates the relationship between the implementation of AI, its impacts, and the resulting outcomes, highlighting AI's tremendous potential to create innovative and effective learning experiences (Figure 11). This framework confirms that AI plays a crucial role in transforming the paradigm of mathematics education into a more adaptive, inclusive, and effective approach, empowering students to reach their full potential in science and technology fields.

Figure 11. AI Framework in Mathematics Education – Inputs, Impact, and Outcomes

In order to provide a more synthesized understanding of the implications of AI including AR in mathematics education, we propose the AI Framework in Mathematics Education as shown in Figure 11. This framework is developed based on a thematic analysis of the top 10 most-cited articles identified in this bibliometric study. The framework outlines the key inputs, impacts, and outcomes of AI-based learning interventions, particularly those involving AR, as highlighted across the literature. While Figures 8 and 9 present a network-level overview of trends and clusters, Figure 11 offers a conceptual structure that reflects how AR and related technologies contribute to deeper student engagement, improved conceptual understanding, reduced math anxiety, and enhanced STEM motivation. This framework is intended to guide future research and practical implementation of AI tools in mathematics education.

Limitations and Recommendations

This study has several limitations. It only includes publications indexed in Scopus, which means that regional trends or non-English publications may not be captured. Furthermore, the study focuses on the period from 2015 to 2024 and analyzes keyword trends without evaluating the methodological quality of each study.

Based on the results of the bibliometric analysis, several strategic recommendations can be made. First, there is a need to develop AR features based on structured instructional models to more effectively enhance students’ conceptual understanding in mathematics. Second, interdisciplinary and international collaboration remains limited and should be strengthened to broaden perspectives and foster innovation in AR development in education.

In addition, the implementation of AR should consider pedagogical and ethical aspects to ensure that its use is not only visually appealing but also generates meaningful positive impacts on students’ learning outcomes. Developing more adaptive AR implementation models is also necessary, taking into account teachers' readiness through the technological pedagogical content knowledge (TPACK) framework. Future research is encouraged to explore innovative instructional strategies such as flipped learning, problem-based learning, and gamification, as well as to design specialized training programs to help teachers integrate AR effectively into classroom teaching.

Funding

The authors express their deepest gratitude to the Government of the Republic of Indonesia, the Ministry of Education, Technology, Research, and Higher Education for the financial support through the PDD grant No. 069/E5/PG.02.00.PL/2024. Sincere appreciation is also extended to the Rector of Universitas Negeri Padang for the funding provided through grant No. 2660/UN35.15/LT/2024. Heartfelt thanks are also given to all individuals and organizations who have contributed to the successful completion of this research.

Generative AI Statement

As the authors of this work, we used the AI tool ChatGPT to enhance the clarity, structure, and descriptive quality of the manuscript in order to meet academic standards. Following the use of this AI tool, we thoroughly reviewed and verified the final version of our work. We, as the authors, take full responsibility for the content of the published manuscript.

Authorship Contribution Statement

Gusteti: Conceptualization, research design, data collection from bibliographic databases, and bibliometric analysis. Musdi: Manuscript revision, interpretation of analysis results, and technical supervision of bibliometric analysis. Dewata: Critical manuscript revision. Rasli: Final review.