Design and Implementation of a Learning Analytics Module for the Khan Academy Platform

Abstract

The learning process is changing due to the possibilities offered by the new technologies in education. One of these possibilities is the exhaustive collection of data. Currently, most e-learning platforms are able to collect large data set of students’ interactions as events. However, these low level data are difficult to be interpreted directly by learning stakeholders. A major challenge is how to transform these low level data into intelligent information, and show them to teachers and students in an understandable way. Learning analytics, which is the science that deals with this problem, has emerged strongly in recent years. Khan Academy is one of the pioneering platforms to show relevant information about the learning process. However, Khan Academy’s learning analytics module can be greatly improved to include new intelligent information that is useful to enhance the learning process. In this work, we have designed and implemented a learning analytics module for the Khan Academy platform, which extends the Khan Academy learning analytics support by default. In this way, we propose a set of interesting parameters in order to learn more about the learning process, and we establish a way to process them from the low level data. Furthermore, these parameters have been implemented, as well as individual and class visualizations based on these parameters. We have used technologies like the Python programming language, Google App Engine infrastructure, Datastore based on Big Data, or Google Charts. Finally, we show how this module and the defined parameters can be used to evaluate the learning process. We apply this evaluation to undergraduate remedial courses at Universidad Carlos III de Madrid, where the Khan Academy platform and the module developed in this master thesis have been used.