engleski

Learning analytics of YouTube videos linked to LMS Moodle

In education, the COVID crisis requires saving and real-time availability of a high quantity of data, especially video materials. Consequently, some institutions use YouTube to store large amounts of video data and link them to the LMS Moodle. In this paper, we are dealing with analyzing YouTube analytics data and Moodle data to discover new knowledge about students' behavior and identify the factors that influence the students' success. The variables that are analyzed in our approach include: Type of video material, the complexity of the content in the video, the number of spikes, the number of dips, the week of semester, the type of study, absolute duration of the videos, the week of COVID isolation, subject, the number of students on the course, total views, the number of views in different time segments, total watch time, average percentage viewed in different time segments, the number of students that didn't watch the videos, the number of downloads of files associated with videos in different time segments and others. We applied our approach to videos from two courses. Results indicated differences in interactions with videos in different time periods, differences in patterns for different types of videos. Research results could help teachers to monitor video viewing rates at key points throughout the semester to identify level of engagement and act early to prevent students` dropout.

learning analytics ; YouTube ; videos ; Moodle ; COVID

nije evidentirano