The MOOClet Framework: Improving Online Education through Experimentation and Personalization of Modules
11 Pages Posted: 20 Nov 2014 Last revised: 26 Nov 2014
Date Written: November 12, 2014
Abstract
One goal for massive open online courses is that the educational benefits they provide scale as a function of the number and diversity of learners interacting with the platforms, since an increasing amount of data is available about what interactions and content increase engagement and learning, as well as which educational interactions are effective for which learners, particularly since more and more is known about each individual. This paper presents the MOOClet Framework for tackling this goal, which recognizes a key relationship between randomized experiments and personalization of content. The Framework defines MOOClets as modular components of online courses that can be modified to create different versions, which in turn can be iteratively and adaptively improved through experiments and personalized to characteristics of users. We show how the MOOClet Framework provides guidance in identifying MOOClets and augmenting existing platforms with a platform-independent layer that enables experimentation and personalization even when platforms do not provide native support. We present a concrete usage scenario of the framework in an implementation for the EdX platform, showing how the addition of reflection questions and other content to a lecture video could be experimentally evaluated and personalized. A modeling simulation is also presented to show how using the MOOClet Framework allows data-driven decisions about experimentation and personalization to be made using existing machine learning models. Consideration of the MOOClet Framework could help researchers, instructors, and course designers in identifying, implementing, and improving modular components of existing online education platforms through experimentation and personalization.
Keywords: Online education; MOOC; MOOClet; Modularity; Experiments; A/B testing; personalization; adaptive learning
Suggested Citation: Suggested Citation