Using Learning Management System Activity Data to Predict Student Performance in Face-to-Face Courses

Literature Review

LMS provide both instructors and students with several advantages, such as efficient course delivery (Lonn & Teasley, 2009), access to rich online material (Coates et al., 2005), providing fast and easy communications with instructors as well as with other students (Stone & Zheng, 2014), and higher levels of functionality (Stantchev et al., 2014). More crucially for this paper, LMS are also a source of advantages that are specific to instructors. West, Waddoups, and Graham (2007) have reported that although instructors initially adopt LMS because they facilitate the management of the courses, they eventually begin using the more interactive features, thus shifting from passive to active learning. Morgan (2003) also reported that instructors began restructuring their courses once they started using LMS. One particular advantage for instructors, which has received considerable attention, is the ability of LMS to gather vast quantities of data. This ability has led to the rise of educational data mining (EDM), which is a subset of data mining (Ferguson, 2012). Data mining (DM) is the automatic extraction of patterns from large quantities of data (Klosgen & Zytkow, 2002). EDM is an interdisciplinary field that deals with the application of DM techniques to educational data (Romero & Ventura, 2010) where the data is visualized and analyzed in order to evaluate the web activity of students (Romero, Ventura, & García, 2008) and to get more objective feedback (Mor & Minguillón, 2004).