Abstract:

Massive Open Online Courses (MOOCs) provide a unique form for lifelong learning and skills development. Courses vary in covered topics, duration, types of offered learning materials, numbers of assessments, conditions to complete a course, etc. Not less important is demographic and social background of learners. Due to people’s diversity goals and activities vary. Some prefer to work with all available materials and thereby get wider knowledge, others do just enough to pass a course. There are different activities and learners spend different amount of time on them. The first decade of MOOCs existence has almost passed but characteristics of completers are still contradictory. Consideration of different MOOCs’ activities can help understand learners’ needs.

In Estonia a MOOC “About programming” has been organized several times since 2014. The 4-weeks programming MOOC is basically designed for adults but schoolkids can also participate in it. Our course is an introduction to the world of programming with a brief overview of IT related subjects and examples like self-driving cars, programmable domestic appliances, labyrinths, etc. During this course, learners do their first steps in programming in Python. The MOOC on programming with completion rate about 60% is based on individual work. A forum and helpdesk with a quick response in 8 hours (even at weekends) are used to support learners. Another assistance mechanism used in the course is troubleshooters. They contain hints of certain aspects and are provided for every programming task. To pass the course, all mandatory tests and tasks should be successfully completed.

The current paper aims to study engagement clusters based on gathered data from completers of MOOC on programming. For this purpose, we conducted pre- and post-questionnaires about learners background. Also we studied completers report on used engagement activities like reading or watching provided learning materials, trying demos and using troubleshooters, submitting mandatory tasks and tests, etc. To proceed collected data we used different statistical methods.

The research revealed four different engagement clusters among MOOC “About programming” completers. We called them “active knowledge collectors”, “minimum knowledge collectors”, “pragmatic knowledge collectors” and “support required knowledge collectors”. The relationship with demographic and social background characteristics within each cluster was analyzed. Based on the research results suggestions on how to design MOOCs and engagement activities are given.

Keywords:

Massive open online course, MOOC, programming education, Python, completers, engagement, clusters, demographics, social background, characteristics.