Every Click You Make, Every Break You Take, We'll be Watching You

This research-based full paper investigates Pre-recorded Lecture Video (PLV) viewership habits and elucidates correlations between practice and course performance. Using PLVs and Flipped Class Formats (FCFs) is increasingly prevalent in colleges and universities, driven by various motivating factors. Delivering content that was traditionally done didactically in a format that is both accessible and engaging has many benefits to not only the students but also instructors. For example, students can review PLVs multiple times at their convenience until they understand the material. From an instructional standpoint, PLVs and FCFs allow for more classroom interaction, whether through instructor-led examples or peer-to-peer learning activities. As with any new instructional technique, there are best practices that maximize both student performance and satisfaction, often guided by research. Previous studies have investigated both student use and interaction mechanisms of PLVs, specifically their level of viewership and completion of conceptual check-point questions that could be answered while or after watching assigned videos. Students who had substantial viewership and interacted with the PLVs by answering questions witnessed a statistically significant increase in their course grades compared to those who did not. However, viewership and interaction with a PLV do not tell the whole story. To this end, it is crucial to understand how students use PLVs, specifically, what habits they form while watching PLVs and how this behavior affects their performance in the course. During two years, 445 students enrolled in a sophomore-level Statics and Mechanics of Materials course were instructed using PLVs and an FCF. The students were required to view PLVs before class (on the University's approved Learning Management System, Canvas, using Panopto) and answer basic conceptual questions on Top Hat. Student viewership data (date and time watched, minutes delivered, etc.) was collected through Panopto. PLV viewership habits were ascertained: full-watching (e.g., watching and skipping, watching and rewinding), partial-watching (e.g., incomplete watching, skipping), and no viewership. Additionally, normalized viewership and average number of views per video were characterized. Correlations are drawn between viewership habits and final course grades. Students in the lower-performing group displayed poor habits with the PLVs (i.e., they had low normalized viewership and did not view many videos). In contrast, the middle- and higher-performing students displayed much better viewing habits (i.e., the opposite of low-performing students) and earned higher final course grades. It was also seen that the Top Hat cohort had greater upward mobility than their Traditional textbook counterparts.