Emotional Profiles in a Massive Open Online Course and Their Relationship With Engagement

• In Event: Emotions in Technology-Rich Learning Environments: Theory, Measurement, and the Future

Abstract

The importance of emotions in learning has been firmly established in face-to-face learning environments (Pekrun, 2006) and now researchers are examining the roles of emotions in online learning environments. A special issue of Internet in Higher Education showed strong similarities between the experience and effects of emotions in online learning environments and traditional classrooms (Daniels & Stupnisky, 2012). However, little is known about the role of emotions in perhaps the newest and arguably least traditional learning environment: massive open online courses (MOOCs). The purpose of this research was to identify profiles of four achievement emotions (enjoyment, relief, boredom, and guilt) over the duration of one MOOC and examine the differences in cognitive, behavioral, and social engagement.

Data from 330 MOOC learners collected at four time points was included in a latent profile analysis (LPA) to identify emotion profiles (MPlus 7.0: Muthén, & Muthén, 2012). The best fitting latent profile solution was chosen based on a set of criteria (e.g., Entropy value and Lo–Mendell–Rubin Test). We used MANOVA to evaluate how the emotion profiles differed in levels of cognitive, behavioral, and social engagement, in light of significant observed inter-correlations.

The LPA results indicated that a three-class solution fit best for guilt, a two-class solution fit best for boredom and enjoyment, and relief was best described as a one-class solution representing a steady increase. For guilt, the first profile (n=228) represented low levels of guilt with a subtle increase, the second profile (n=59) showed a high, stable level of guilt, and the third profile (n=54) indicated a sharp increase during the last two-thirds of the MOOC. The MANOVA result was significant, Wilks’ Lambda F = 4.706, p < .001. More specifically, the first profile (low level of guilt) was associated with significantly higher levels of behavioural engagement than the third profile (sharp increase), p < .01. Regarding boredom, the first profile (n=116) indicated a sharp increase toward the end of the MOOC whereas the second profile (n=225) represented a stable and low level of boredom. These two profiles differed significantly in engagement, Wilks’ Lambda F = 5.55, p = .001. In particular, the second boredom profile shows a significantly higher level of cognitive, p < .001, and behavioural engagement, p = .01 than the first profile. For enjoyment, the first profile represented a high and stable level of enjoyment (n=287), whereas the second was characterized by a gradual decline of enjoyment (n=54). The MANOVA was significant, Wilks’ Lambda F =54.50, showing that the high-stable profile had significantly higher levels of cognitive, behavioural, and social engagement than the declining profile, ps < .001.

Having a high level of enjoyment during MOOCs appears to be adaptive for students’ engagement, whereas escalating feelings of guilt or boredom appear to impede engagement. Also noteworthy, overall it seems that MOOC learners were mostly classified into pleasant emotion profiles, including a steady increase in relief.

Authors

• Virginia Man Chung Tze, University of Alberta

• Lia Marie Daniels, University of Alberta

• Erin Buhr, University of Alberta

• Lily Le, University of Alberta

• Tracy Lyn Durksen, University of New South Wales