Elite universities have been adapting technology used to predict a student’s final grade to improve their overall results.
Edx, the edtech venture of top US universities MIT and Harvard, is conducting research into how big data can help answer key online learning questions, such as the best ways to teach complex ideas, and which parts of a course are best taught in person instead of online.
By assessing course data — from mouse clicks and time spent on tasks to evaluating how students respond to assessments — the company hopes to shed light on how learners access information and master material.
“We are…Developing best practices to enhance the student experience and improve teaching and learning both on campus and online,” says Nancy Moss, edX director of communications.
Mike Feerick, CEO and founder of online learning company ALISON, says: “There is huge potential for data to improve online learning.”
At Coursera, the Mooc platform with 15 million users hosting courses from top schools including IE, Yale, and Duke Fuqua, the platform offers a dashboard with insights into when students are most likely to stop watching a video, or the percentage of students who are answering assessment questions correctly the first time around.
“If only 10% of learners taking a quiz are answering correctly, teachers can evaluate how they are teaching this particular point or if the wording of the question is ideal,” says Julia Stiglitz, director of business development at Coursera.
Oliver Cameron, VP of engineering and product at Udacity, says there is “incredible potential for online learning to adapt and improve based on student data”.
“Rather than waiting for an end of semester survey to uncover an issue or inefficiency, instructors and course developers can continuously make student data-driven improvements year-round,” he says.
Initially developed by the Open University — the UK distance-learning specialist — these methods are based on an algorithm that is able to predict a student’s final grade within a week of their starting their course.
This works by combining data on how studiously students read online textbooks and how enthusiastically they participate in online forums with data on their social and economic backgrounds.
Simon Nelson, chief executive of FutureLearn, the online learning company with two million users, says big data “helps businesses like ours to create the right experiences and target learners with the right things”.
“Unquestionably there is a huge amount that can be done [with big data]. But we are only scratching the surface,” he says.