Active Learning in a Massive Open Online Course: Intentions, Self-Regulation, and the Trajectory of Utilization

• In Event: Self-Regulated Learning Analytics: Aligning Data and Their Treatment to the Assumptions of Theory

Abstract

Utilizing MOOCs designed specifically for adult professional development (MOOC-Ed, 2014), we apply psychological theories of active learning and self-regulation to further understanding of the user in the free-choice educational environment. Active learning environments entail that: 1) the learner has responsibility for learning decisions, such as choosing learning activities and monitoring and judging progress (Bell & Kozlowski, 2008); 2) the active learning approach promotes a learning process in which individuals explore and experiment with tasks in order to figure out the rules, principles, and strategies for effective performance (Frese, Brodbeck, Heinbokel, Mooser, Schlieffenbaum & Theimann, 1991; Smith, Ford, & Kozlowski, 1997). Self-regulation is central to active learning as it involves setting specific goals, utilizing task strategies, high intrinsic interest, self-monitoring, and self-reflecting on performance outcomes (Zimmerman & Schunk, 2008).
In the research presented in this symposium, data mining approaches and feature engineering decisions are outlined that enable the researcher to overcome the noise of participant log data and distill it into meaningful variables. With these modifications to the data, we assess various predicators of course behavior. Prior research has shown that level of educational attainment can be used as a proxy for one’s experience in applying self-regulatory processes (Artino & Stephens, 2009). In addition, research on the impact of affective dimensions on cognitive outcomes have made clear connections between motivation and academic outcomes (Elliot & Dweck, 2013; Schunk & Zimmerman, 2012). Participant use over time is modeled longitudinally in the professional development MOOCs via latent growth curve modeling (LGM), a special application of structural equation modeling. We examine the impact of intention to complete and education level (as a proxy of self-regulation experience) on course resource usage over time, through two latent factors: intercept and slope, where intercept represents baseline usage level and slope the growth in usage level (Duncan, Duncan & Strycker, 2013).
Our initial findings indicated that intention to complete the MOOC was positively related to the intercept. That is, the higher the intent, the higher the general level of course resource utilization. Our second key finding was that level of education (as a proxy for self-regulation experience) related not to participants’ general level of use but rather to the rate at which use declined over time. Specifically, the higher the level of educational attainment, the more stable individuals’ use trends remained throughout the course. Lower education level resulted in steeper decline in use with time. These findings reinforce the characterization of MOOCs as a user-centric active learning environment, whose benefits are largely determined by the motivation and engagement of the learner.
Future work will continue to apply models of self-regulation to the free-choice learning environments of MOOCs. Paired with LGM and other person-centric mixture modeling techniques, a richer set of trace data will be used to develop more sophisticated learner profiles. These profiles, in turn, can be used to guide both human instructors and machine based adaptive support systems.

Authors

• Eric N. Wiebe, North Carolina State University

• Isaac Benjamin Thompson, North Carolina State University