Patterns of Engagement in a Flipped Undergraduate Class: Antecedents and Outcomes

• In Event: New Empirical Insights on What Energizes Learners: A Session on Emotions and Engagement

Abstract

Background. Online video lectures are an increasingly popular instructional tool that can be incorporated into face-to-face, online, or flipped courses. Although research suggests that video lectures promote learning outcomes (e.g., Chen & Wu, 2015), a concern about such learning environments is whether students regularly watch lecture-videos outside class. Little research has investigated antecedents or academic outcomes of students’ video-viewing patterns. The purpose of this study is (1) to describe patterns of video views before exams in a flipped anatomy course and (2) to investigate predictors (i.e., perceived competence and task value) and (3) outcomes (i.e., final course grade) of video viewing patterns. This study is situated within expectancy-value theory, which posits that students’ expectations for success (e.g., can I do this?) and subjective task values (e.g., do I want to do this?) are proximal determinants of their achievement-related outcomes (Wigfield & Eccles, 2000). Thus, we focus on perceived competence and task value as relevant antecedents of online video lecture viewing behaviors.
Methods. Participants were 265 undergraduates (75% female, 85% Caucasian) taking a large introductory anatomy course at a Midwestern public university. Students watched video lectures at home and participated in active learning activities during class. We collected self-reported perceived competence (5 items; Midgley et al., 2000; α = .86) and task value (5 items each; Conley, 2012; interest: α = .94; attainment: α = .78; utility: α = .84) during Week 2, log files of students’ lecture-video views, and the final course grade. To understand the nature of lecture-video views prior to a course exam, we examined the period before the first exam, calculating the sum of minutes of videos viewed during each wave (four days), resulting in 12 waves for analyses. We first examined an overall growth curve of the video views using linear mixed models (Fitzmaurice et al., 2004) using R. Next, we investigated how perceived competence and task value predicted random effects of the initial level and growth rates, and how these student-specific trajectories predicted final grades.
Results and discussion. We modeled intercept, linear, and quadratic trends of overall lecture-video views before the first exam. Video views were high for most students during the four-day waves immediately before the exam, were low during the waves further from the exam, and were slightly higher during waves 11-12, about six weeks before an exam (see Figure 1). The slightly higher pattern for waves 11-12 could be attributable to students’ possibly high initial motivation at the beginning of the semester (during the first few days of class). Student-specific intercepts (beta= 3.45**) and linear slopes (beta= 10.39*) positively predicted final grade (p < .001), suggesting that students who watched more video in the final week prior to the exam and students with a less steep pattern of views had higher final course grades. Surprisingly, baseline motivation variables were not significant predictors of video viewing patterns. Results suggest that instructors should encourage students to watch video lectures more regularly throughout the course. Future research will consider other factors that explain viewing patterns.

Authors

• Joshua Michael Rosenberg, The University of Tennessee - Knoxville

• You-kyung Lee

• Kristy A. Robinson, McGill University

• John Ranellucci, Hunter College

• Cary J. Roseth, Michigan State University

• Lisa Linnenbrink-Garcia, Michigan State University