Student Engagement with Pre-Recorded Lecture Videos in a Flipped-Class Format

This research-based work-in-progress (WIP) addresses the use of flipped class formats (FCFs) in a collegiate environment, particularly student engagement with pre-recorded lecture videos (PLVs). FCFs have gained popularity in the past decade, for the student-centered instruction method provides pupils with additional opportunities to engage with the course material, their peers, and/or instructor, typically more than they would have received through a traditional didactic lecture. Although shown to increase student engagement, satisfaction, and performance, FCFs, in particular on-line lecture videos, have not been scrutinized in terms of student interaction with PLVs. To this end, it is hypothesized that by providing a means of interaction during a lecture video, students will be more engaged with the digital content and subsequently will perform better than their peers who did not engage with online media. During the fall 2022 term, 183 students enrolled in a sophomore-level Statics and Mechanics of Materials course were taught using a FCF. The students viewed a total of 65 PLVs encompassing the content for 23 lectures on the university's approved video platform, Panopto, and integrated in a Learning Management System, Canvas. Oftentimes there were multiple videos for one lecture, as to minimize the length of the video. The PLVs had a total run-time of 271 minutes and 21 seconds; the average video length was 4 minutes and 10 seconds, and the mode of the lecture video length was 3 minutes and 26 seconds. Twenty-one pre-recorded lectures had an accompanying set of questions administered through Top Hat that tested the students' conceptual understanding of the material presented during the videos. These questions took three common forms–multiple-choice, true or false and matching–and totaled 162. There were an average of 7.5 questions per lecture, and the mode of questions per lecture was 8. Hints for incorrect answers were provided to students. Student engagement with the lecture videos was analyzed through quantitative Panopto and Top Hat data. Student engagement metrics such as normalized video viewership and the answering of Top Hat questions (answering habits, correctness to attempts ratio) were determined. Correlations and trends between video viewership, interactions with Top Hat, and course performance were elucidated using statistical methods, namely
$H$
-tests and analyses of variance. Students who answered Top Hat video questions after watching lecture videos performed half a letter grade better than those who answered the questions prior to viewing the videos. The top and middle terciles of normalized video viewership cohorts had a corresponding 6% and 3.6% increase in performance over the lowest tercile group; the more a student watched lecture videos, the better they performed. Students who had high correctness scores in comparison to their attempts on Top Hat video questions (namely the top and middle tercile cohorts) performed 10.6% and 5.2% better than the lowest tercile group, respectively. Lastly, students who answered all Top Hat video questions on average earned a 5% higher score than students who skipped questions.