Inside global education – The conceptual constitution of key terms

• In Event: Highlighted session: Rethinking global citizenship education within macro and micro levels

Proposal

Schooling now more than ever seeks to prepare students to take part in the ‘global competition’ for future education and employment destinations, participate in ‘global problem solving’, and, broadly, be better equipped to face the challenges globally connected contemporary societies must engage with (Dill, 2013; Goren & Yemini, 2016; 2017a;b; 2018; Reilly & Niens, 2014; Vidovich, 2004). The grounded manifestations of this new global outlook are manifold and include global citizenship education (GCE), 21st-century skills, intercultural-competencies, cosmopolitan education, and education for global and international mindedness. These terms are all used synonymously in many cases, with policy-makers and scholars taking into account different contextual factors when choosing which to employ, but the factors taken into account are rarely presented, and the meaning associated with each term within each context remains somewhat subjective. This lack of clarity about terms being used and how they are linked can make providing a coherent and accurate review ‘of the field’ challenging and always partial.
In this project, we use an innovative, cross-disciplinary methodology which combines the use of Big Data, artificial intelligence (AI) in the form of natural language processing (NLP), network visualization, and qualitative data analysis to examine the similarities and differences in the academic scholarship surrounding each of the five terms presented above. This analysis enables us to create a comprehensive landscape of a large body of literature that identifies key strands of research within each concept and demonstrates the relationships between them on a much larger scale than can be done using traditional methods of systematic qualitative analysis. The visualizations allow us to view terms that are central to the discourses surrounding each concept for which data were collected, as well as the relationship between the terms themselves, in the form of colored communities (clusters) that indicate terms which are more likely to be mentioned together. The visualisations also allow for a comparison of the clusters and terms that are at the heart each of the bodies of literature that we examined.
Methods
We used the ERIC database to collect between 100-300 academic peer-reviewed papers published in the last five years, concentrating on each of the five terms: GCE (N=226), 21st century skills (N=168), intercultural-competencies (N=191), cosmopolitan education (N=159), and education for global and international mindedness (N=40).
Our method (developed by Blumenfeld-Lieberthal, Serok, & Milner, 2017) use the OpenCalais software, developed by Thomson Reuters , which utilizes machine learning and NLP driven algorithms to derive themes (referred to in OpenCalais as “social tags”) from each of the individual articles in all cohorts.
Once we had assembled the database of all searchable articles and their social tags, we used a co-word methodology (i.e. linking papers through co-occurrences of themes) to create networks where the nodes represent the themes (social tags), and the links (also known as edges) denote co-appearances of the same social tag in the same articles. The weight of the links represents the number of times two different themes appeared in the same article, in other words, the number of articles that included the two themes. We included only nodes with a degree larger than or equal to 10, which means themes that appeared together with at least 10 other themes. This allowed us to concentrate on the essential components and relationships in the network (Kumar, Novak, & Tomkins, 2010). We used the network visualization and exploration software Gephi to perform our network analysis and produce the visualization of the network. Through these visualisations, we emphasized different attributes of each network, such as the different colors of the nodes, which indicate the different communities the nodes are assigned to.
Due to the descriptive nature of the visualisations, which does not provide interpretive value, we supplemented the analysis using a deductive qualitative approach. When examined more closely how the visualisations enable us to view outliers; terms that one may not expect to be connected to one another. We located the instances where these terms appeared together in the same papers in our database and sought to reveal the intricacies of these connections so as to shed light on their origin and meaning. Qualitative analysis is also used to provide examples for each of the clusters in the visualisations, and to highlight the unique characteristics of each body of literature that we examined.
Contribution to knowledge and preliminary findings
The figures created thus far through the network visualisations have enabled to identify the clusters and topics that lie at the heart of the scholarship surrounding global citizenship education and 21st century skills. We limited these networks to include only terms that had a K-core score of 10 and above, meaning they were connected to at least ten other terms in the network. This was done in order to highlight only the terms that are integral to the scholarly discourse. The visualisations show that the GCE scholarship can be divided into four main clusters: Pedagogies; globalisation and interculturalism; education for sustainable development, and higher/international education. When limited by the same criteria, the 21st century skills scholarship visualization reveals quite different clusters which include: Pedagogy and assessment; curricula and learning; educational technology, and finally supranational organisations and the digital divide. One interesting finding that arises from the visualisations is the centrality of UNESCO to the GCE discourses particularly with regards to ESD, a while the 21st century skills visualization includes PISA and the OECD, which have recently begun to incorporate this concept to their professional rationales and measurements. Additionally, the clusters in the GCE scholarship also much more distinct and easier to classify than those that comprise the 21st century skills scholarship, highlighting the contested nature of the meaning of GCE and the plethora of ways in which the term is applied in recent scholarship (Goren & Yemini, 2017a; Oxley & Morris, 2013).

Authors

• Heela Goren, Institute of Education University College London

• Miri Yemini, Tel Aviv University

• Claire Maxwell, Institute of Education University College London