Domestic Research Collaboration Networks of Korean Universities in the Top and Less-Prestigious Journals

• In Event: Higher Education in East Asia: Gender, Agency, Research Collaboration and Assessment

Proposal

Introduction
Globalization has increased international publications by higher education institutions (hereafter, institutions) from non-English-speaking countries, including South Korea (hereafter, Korea). Korean researchers are increasingly required to publish papers in English-language journals, which are often regarded as more prestigious (Lee & Lee, 2013). As research has become a core function of universities (Baker, 2014), traditionally non-research universities (TNRs) in Korea have significantly expanded their scientific activities over time (Kim, 2019). However, studies have also reported that the current knowledge production system is systematically biased in favor of research universities, which dominate international journal publications. For TNRs in Korea, publishing in English-language journals presents significant challenges. Structural bias may have reinforced existing inequalities within the academic landscape by leading them to publish in less prestigious journals. Accordingly, this study aims to investigate the research collaboration networks of Korean universities, with particular attention to the prestige level of journals. Specifically, it examines domestic collaboration networks to assess whether Korean TNRs, during the early stages of their development, established collaborative ties with domestic peers to facilitate publication in renowned journals.

Literature Review
Some studies in domestic research collaborations focused on the U.S case. First, U.S. universities steeply increased domestic collaborations in the last few decades (Adams et al., 2005; Jones et al., 2008). Second, in both prestigious and less-prestigious journals, research universities were the dominant actors in domestic collaboration networks in the U.S. (Yu, 2022). A similar pattern was observed in international collaboration of U.S. universities in top journals (Yu, 2023). In addition, Korean universities’ publication production in international journals also increased significantly and research universities were the dominant actors in international collaboration (Kim, 2019).

Method
This research followed the analytical approach that Yu (2022, 2023) adopted. This study analyzed the Science Productivity Higher Education and Research Economy Project (SPHERE) data. This contains publication data of the Web of Science’s Science Citation Index Expanded (SCIE) data (Baker et al., 2015). This study measured the prestigiousness of the journals with their impact factors because highly cited journals are often regarded high-quality journals (Glänzel & de Lange, 2002; Kim et al., 2010). This paper included the top and bottom 25% of journals in terms of impact factors, which the dataset provided.

This study focuses on the early development of the higher education system in Korea. Significant expansion and the implementation of several research policies occurred during the 1990s. The study examines how international publication evolved alongside the growth of the higher education sector. Specifically, it analyzes research collaboration patterns in 1995, 2001, and 2011. The analysis begins with the 1995 dataset, as domestic collaboration prior to 1995 was relatively limited. Additionally, the expansion of Korea’s higher education system began to slow after 2011 (Yu, 2024).

This study conducted social network analysis. Actors of this study were Korean universities. The mission of universities was categorized by adopting Shin (2009)’s classification (Kim, 2019). Secondly, when Korean universities co-published a research paper, a tie was created.

One-mode, symmetrical, undirected, and dichotomous matrices were constructed. A value of 1 was assigned to a pair of Korean universities if they co-authored at least one article; a value of 0 was assigned if they did not. In this study, degree centrality represents the number of peer Korean universities with which a given university collaborated.

This study examined density, subgroups of the networks, degree centrality, betweenness centrality, and eigenvector centrality. For subgroup analysis, this study utilized the Clauset-Newman-Moore method (Clauset et al., 2004). The networks would be visualized with the Harel-Koren algorithm (Harel & Koren, 2002).

Result
In 1995, articles from 69 Korean universities were published in journals ranked within the bottom 25% by impact factor. Approximately 38% of universities were TNRs. On average, TNRs collaborated with two institutions. In all networks explored, Seoul National University (hereafter, SNU) was the central collaborator. SNU worked with 31 institutions in this network. Conversely, among the 54 Korean universities publishing in the top 25% of journals by impact factor, 26% were TNRs. On average, TNRs worked with less than two peer institutions in Korea. SNU collaborated with 28 institutions in the top 25% journals.

In 2001, in the bottom 25% journals, 154 Korean universities had a collaborative tie. Approximately 69% of them were TNRs. On average, TNRs worked with three and four Korean universities. SNU worked with 61 universities. In the top 25% journals, 93 Korean universities published papers and 48% of them were TNRs. On average, TNRs collaborated with four to five Korean universities. SNU worked with 58 universities.

In 2011, 240 Korean universities published articles in the bottom 25% journals. Almost 80% of them were TNRs. On average, while TNRs collaborated with nine peer universities in Korea, research universities worked with 55 institutions. SNU worked with 103 Korean institutions. On the contrary, 144 Korean universities published papers in the top 25% journals and 66% of them were TNRs. On average, TNRs collaborated with eight universities but research universities collaborated with 46 universities. SNU collaborated with 91 peer institutions.

Over the years, although the number of institutions increased, network density remained low across all journal tiers. The number of subgroups within the networks varied, ranging from five to seven.

Conclusion
This study yielded findings consistent with previous research (Yu, 2022, 2023). The number of Korean universities contributing to international publications increased across both tiers of journals. However, more Korean universities appeared in journals with lower impact factors. Most TNRs’ collaborative ties were concentrated in less prestigious outlets. TNRs exhibited fewer collaborative connections compared to research universities. These results reinforce the argument that the current knowledge production system systematically disadvantages TNRs (Kwiek, 2021). TNRs continue to face significant challenges in publishing in international journals, particularly those with high citation impact. Research universities maintained dominance across all collaboration networks, regardless of journal tier, indicating a persistent imbalance. The findings suggest that Korean TNRs were inadequately prepared for the demands of international scholarly publishing. The prevailing norms and structures of global knowledge production remain unfavorable—and in some cases detrimental—to the development of TNRs.

Author

• Wan Yu, Chung Ang University