Abstract
This paper uses 3596 core papers in the WOS (Web of Science) database in the past ten years as a data source, and uses CiteSpace knowledge graph software to visually analyze foreign STEAM education. Through cluster analysis, it is found that the themes of foreign STEAM education research mainly focus on three aspects: interdisciplinary integration, thinking training, and gender-based education equity research. This article studies the current situation and hot spots of STEAM education abroad, in order to provide a reference for the development of STEAM education in China and make it more in-depth and comprehensive.
Introduction
While STEAM (Science, Technology, Engineering, Arts, and Mathematics) education has become a global priority, the rapid expansion of academic literature makes it difficult to discern overarching trends and emerging "hot spots." Individual studies often focus on specific classroom interventions, leaving the macro-level evolution of the field obscured. This research addresses this "forest for the trees" problem by providing a comprehensive visual analysis of international STEAM research from 2011 to 2022. The goal is to offer a roadmap for researchers to understand saturated topics and identify the next frontiers of educational innovation.
Methodology and Data Source
The study utilized a curated dataset from the Web of Science (WOS) core database, specifically the Social Sciences Citation Index (SSCI). A total of 3,596 valid literature records were analyzed, spanning from January 2011 to June 2022. The search criteria focused on "STEAM EDUCATION" or "STEM EDUCATION," limited to English-language articles.
The researchers employed bibliometric visual analysis using CiteSpace 5.8.R3 to generate knowledge graphs. To categorize the research, the Log-Likelihood Rate (LLR) algorithm was applied for keyword clustering. The credibility of these clusters was validated using Modularity (Q value > 0.3) and Silhouette (S value > 0.7) metrics. Additionally, citation burst detection was used to identify keywords signaling research frontiers.
Results and Discussion
Growth Trends and Geographical Distribution
The volume of STEAM publications has seen exponential growth over the last decade. In 2021, the number of publications was approximately seven times that of 2011, reflecting a consistent upward trajectory in global interest.
Analysis of the country collaboration network reveals the United States as the undisputed leader in STEAM research, contributing 1,760 papers. Other significant contributors include Australia, England, and China. While Purdue University leads in publication volume, the University of Michigan exhibits the highest "centrality," acting as a critical bridge for international collaboration.
Core Research Pillars
The visual analysis identified ten distinct clusters with an average Silhouette value of 0.905, indicating high consistency. Three dominant research themes emerged:
- Interdisciplinary Integration: The shift from STEM to STEAM emphasizes "artistic integration." However, the field currently lacks a comprehensive theoretical framework for this interdisciplinarity.
- Gender-Based Equity: Research has evolved from simple representation counts to exploring psychological barriers. A key insight is that mathematical self-efficacy—the perception of one's own ability—is a more significant barrier for female students than actual aptitude.
- Computational Thinking (CT): CT is being reframed as a fundamental problem-solving mindset across disciplines, though a unified definition remains elusive.
Conclusion and Future Directions
The analysis highlights that while the STEAM landscape is maturing, particularly in the US, several challenges remain. Future research should focus on:
- International Cooperation: Strengthening ties between developing STEAM programs (such as those in China) and established international institutions.
- Standardizing Definitions: Resolving conceptual confusion surrounding terms like "computational thinking" to facilitate better cross-study comparisons.
- Active Learning: Implementing pedagogical shifts away from traditional models, as active learning has been shown to significantly reduce failure rates in STEM subjects.
By mapping the current state of the field, this study provides a foundation for more targeted and effective STEAM education research globally.
Summary updated