Exploring learning management system usage patterns and their relation to student performance and satisfaction

Blended education (BE) continues to be an increasingly widespread delivery mode of learning and teaching, as was outlined in the recent survey report on digitally enhanced learning and teaching in European Higher Education Area (EHEA) institutions (Gaebel et al., 2021). Defined as the deliberate and integrated combination of online and face-to-face instruction (Prinsen & Terbeek, 2021; Van Valkenburg et al., 2020), BE is the most popular delivery mode in Europe with 75% of EHEA institutions applying it either throughout the institution or in some faculties (Gaebel et al., 2021). Moreover, all of the responding institutions reported providing some form of digitally enhanced education, with 57% using it widely throughout their institutions. BE has been found to be successful in facilitating self-paced learning activities and individualizing students’ learning pathways (Castro, 2019). Other benefits reported in the literature include increased learning involvement and interaction, more flexible and accessible learning opportunities, and cost effectiveness (Boelens et al., 2017; Buhl-Wiggers et al., 2022; Rasheed et al., 2020). However, research into whether and how BE is related to student achievement and student satisfaction is scarce. Also, with the increased accessibility of ICT, and the occurrence of new forms of digitally enhanced education, difficulties arise with conceptualizing BE; there is not one unified vision of what constitutes a blended course (Hrastinski, 2019; Boelens et al., 2015). Previous research has identified a variety of approaches to defining and classifying blended courses, which most often focus on the proportion of online and face-to-face instruction (Boelens et al., 2015; Trigwell (2005). The latter approach is often criticized for being overly reductive. In order to gain insight into the quality of BE and potential areas of improvement it is therefore important to be specific about the types of blended courses that are being studied. As outlined by Park et al. (2016), it may be beneficial to operationalize BE by categorizing the diverse and varied types of blended courses within an institution. These categories can then be used to identify potentially significant predictors that may estimate student academic success within the given context. As described by these authors, clustering academic courses based on learning management system (LMS) data is an emerging approach to such a categorization. As part of a larger study into BE at the University of Amsterdam, the current study aims to use institutionally available LMS data in order to identify distinct categories of courses with an asynchronous online component. Furthermore, the aim is to observe whether these distinct categories could be used to estimate student success and satisfaction, in order to provide targeted and differentiated support. Specifically, this study is guided by the following research questions: 1) Can multiple clusters of Learning Management System (LMS) usage patterns be identified across UvA courses? And if so: 2) Are there differences between the clusters of LMS usage patterns in course averages of student performance and satisfaction? Since a data-driven approach in examining the benefits and drawbacks of BE is emergent and has not often been applied yet, an additional aim of this study was to explore the opportunities and limitations of data-driven research for BE in the higher education context.