Research Trends in English Writing: A Semantic Network Analysis

Abstract

This study explores the shifting landscape of English writing research from 2001 to 2020, utilizing a large-scale, data-driven methodology. Data from 1495 articles published in Korea over these two decades was collected and preprocessed with the Biblio data collector, then analyzed with Netminer. The approach involved four stages of semantic analysis: frequency, centrality, network, and modularity analyses. The findings highlight dynamic shifts in research focus. While keywords like ‘test’, ‘college students’, ‘vocabulary’, and ‘level-based’ remained consistent, the 2010s saw emerging themes like ‘task’, ‘textbook analysis’, ‘corpus analysis,’ 'peer feedback’, and ‘genre-based approach’. Centrality analysis showed that in the 2000s, alongside ‘sentence’, ‘questionnaire’ held a central position with multiple nodes linked to it. In the 2010s, ‘questionnaire’ persisted as a central theme but was joined by ‘relationship.’ Network maps generated with Spring 2D and PFnet depicted these evolving interconnections. In the second period, ‘feedback’ emerged as a central theme, yet directly connected to only two nodes: ‘error’ and ‘peer feedback.’ Modularity analysis identified six research groups in each period, with the ‘questionnaire group’ being most significant in the 2000s and the ‘peer feedback group’ gaining prominence in the 2010s. This research illuminates the evolving trends in English writing research, underscoring the potential of big data-driven approaches to uncover key insights and patterns.