Technology-enhanced learning in higher education: A bibliometric analysis with latent semantic approach

Technology-enhanced learning (TEL) is the terminology used to refer to the benefit of utilizing information and communication technologies (ICTs) in learning and teaching. In general, the development of ICT has changed the nature of communication, socializing, entertainment, purchasing, and learning (Abdullah, Ward, & Ahmed, 2016). ICT, being flexible, accessible, affordable, and without the temporal or spatial limitations of more traditional resources, has been positioned as a fundamental part of teaching and learning. Accordingly, pedagogies are increasingly dynamic as they have to responsively evolve with innovative teaching and learning tools with the help of ICT (Becker & Ravitz, 1999). Hence TEL applications fulfill a need and as such are increasingly adopted in higher education (Lin, Lu, & Liu, 2013). For instance, massive open online courses use social media platforms to engage students, which enable them to co-create knowledge in a collective learning process (Shen & Kuo, 2015). Interactive whiteboards are used as a teaching tool in educational settings from pre-schools to universities (Sumak & Sorgo, 2016) and traditional educational resources are increasingly supplemented by online video technology (Nagy, 2018). Also, learning management system (LMS) has been successfully used to enhance the quality of education (Findik-Coskuncay, Alkis, & Ozkan-Yildirim, 2018) and the use of tablet computers has been found to improve students’ performances and levels of engagement (Wakefield, Frawley, Tyler, & Dyson, 2018). The TEL applications have changed conventional learning mechanisms and can be regarded as both innovative and disruptive.

The burgeoning use of TEL in higher education has also gained the attention of academics, and thousands of TEL-related research articles offer the necessary information to explore the themes and details regarding TEL development as a whole. However, the published research papers either examine a research field through a specific focal point or review a limited number of articles on a small scale, and the subject related to TEL development as a whole is rare. For example, Jump (2011) used 16 case studies as a basis from which to explore how ICT enhances learning from the perspective of teachers’ motivations and aims, and Tamim, Bernard, Borokhovski, Abrami, and Schmid (2011) provided a meta-analysis to compare students’ achievements in formal face-to-face classrooms both with and without the use of ICT. In addition, Kirkwood and Price (2014) reviewed 47 TEL articles to clarify how researchers perceive the meaning of enhancement. Besides, almost all of the studies regarding TEL development need close-reading from experts in order to comprehend the themes of research (Clegg et al., 2003, Hanson, 2009, Price and Kirkwood, 2014). This is a challenging task to accomplish in term of time and quality for analyzing such massive textual data, as the TEL relevant studies have exploded in the last decade.

Therefore, in order to find a means to enable researchers to rapidly scrutinize research containing extensive information that would be beyond their capability to extract manually, we propose a hybrid bibliometric approach, which combines both direct citation network approach with cluster analysis and text mining with latent semantic analysis (LSA) to examine TEL development in higher education from the related research articles indexed in the Web of Science database, which is the oldest well-established tool for citation analysis and has been applied extensively in many bibliometrics research (Olijnyk, 2015). The authors’ citations of articles was first utilized as a well-defined and dominant feature from which to identify the critical articles and cluster the principal streams in the research area. Temporal development of TEL was also analyzed by a direct citation network approach with cluster analysis according to year of publication. The result of direct citation network analysis has been identified as being the most accurate representation of knowledge taxonomies compared to co-citation or bibliographic coupling analysis, because it has the least risk of missing emerging research domain (Klavans & Boyack, 2017). Meanwhile, the clustering analysis outlined the main streams and identified the critical articles. Researchers can scrutinize the identified critical articles to gain an overview of TEL research development in each stream. In addition, we further examined the dataset by the topic modeling method in text mining, which utilized automatic procedures to group documents into meaningful categories representing the main topics being discussed in the documents. Latent semantic analysis was applied to identify the subgroup topics of each leading stream, which give the analyzed result in more detailed scales. The subsequent LSA extracted the subgroup topics based on all the articles in each stream. Researchers can further review the subgroup topics and the articles containing these subgroup topics to innovate their research questions. The proposed hybrid bibliometric approach could help the researchers to monitor the development and recognize trends and changing patterns of the TEL research field in a comprehensive way.

The rest of this study is organized as follows. Section 2 reviews the literature on TEL development and its application in higher education. Section 3 describes the methods and algorithms we used to retrieve TEL-related articles, analyze the articles’ direct citation network, and identify topics. Section 4 provides statistical descriptions of TEL articles and the leading streams in the TEL research area and their associated topics over time. Section 5 presents the implications and contributions of this study, and offers suggestions for future research.