In a not so long ago time I dabbled in creativity research, at least through the lens of how children become creative, believe themselves to be creative, and exhibit creative thought, behavior, and ideas. All of that said, the use of social network analysis in educational research is sparse…
Nevertheless, I found an article by Zhang, Zhang, Yu, and Zhao (2015) that caught my eye… and used k-core to help explain the spatial structure of creativity research over a ten year period between 1992 and 2011. They argue that because of the proliferation of creativity research being generated from various disparate fields, it can become difficult to understand where the field as a whole needs to move in the future. Using keywords to express themes and trends in the literature, they used social network analysis to examine the cooccurrence between words to “deeply explore research hotspots, research opportunities, as well as cutting-edge evolution in creativity field” (Zhang et al., 2015, p. 1024). Therefore, their research question, though not explicitly stated, was: Collectively, what are the common themes of research, future research opportunities, and hotspots identifiable from past research and trends in keyword among the extant literature?
The data was collected directly from Web of Science and keywords were extracted and summed using NoteExpress 2.0. The primary keywords were then identified to other keywords in the same paper. In other words, a paper typically has three keywords and these keywords can, if looked at broadly, tell trends of research by their simple connections to each other. Furthermore, using a method to judge the distance between measured objects, multi-dimensional scaling (MDS), they also figured the position of each research topic and schools of thought by employing MDS.
This figure demonstrates the co-occurrence of each work, similar to a correlation matrix.
They used k-core to derive groups of connected keyword paradigms to “pick out the sub-groups according to each keyword degree in the keywords networks” (p. 1029). This resulted in 15 sub-groups, whose node degree ranged from 45 to 22, and in which 89.6% were above 40.
Results indicated that out of 163 top keywords from 4,575 papers on creativity research, keywords can be best described into five main topics: creativity applications into particular areas, pathology of creative generation, individual level creativity, organizational level creativity, and theoretical and methodological studies of creativity research.
Using their MDS analysis, they identified the spatial structure of creativity research. Results there indicated
With the main top keyword of “Creativity” in the center, each quadrant represents key research paradigms found.
From here, they went on to further analyze the five key areas of research and what those fields are up to during the past ten years of research. They then explain the salient areas of future research suggested therein. That is, future research should be focused on the process of creativity and specific sub-processes of the complex creative formation.
Ultimately, this form of analysis, namely k-core, can provide researchers with a metric to determine groups of centrally connected relations in social networks. This further aids researchers in seeing who is important, integral to the network, and important to collectively speak about and examine.
Zhang, W., Zhang, Q., Yu, B., & Zhao, L. (2015). Knowledge map of creativity research based on keywords network and co-word analysis, 1992–2011. Quality & Quantity, 49(3), 1023–1038. https://doi.org/10.1007/s11135-014-0032-9