Posted Sep 11, 2014 by Dan Friedman
[original post found here.]
Three years ago this week, Sebastian Thrun recorded his Stanford class on Artificial Intelligence, released it online to a staggering 180,000 students, and started a “revolution in higher education.” Soon after, Coursera, Udacity and others promised free access to valuable content, supposedly delivering a disruptive solution that would solve massive student debt and a struggling economy. Since then, over 8 million students have enrolled in their courses.
This year, that revolution fizzled. Only half of those who signed up watched even one lecture, and only 4 percent stayed long enough to complete a course. Further, the audience for MOOCs already had college degrees so the promise of disrupting higher education failed to materialize. The MOOC providers argue that completion of free courses is the wrong measure of success, but even a controlled experiment run by San Jose State with paying students found the courses less effective than their old-school counterparts.
This shouldn’t have come as a surprise. Online learning long ago solved the access problem: Between the 8 million people who have signed up for MOOCs, and the more than 1 billion downloads from Apple’s iTunesU (Apple is quietly a larger force in online education than any upstart), we already know people want to take online courses. What we don’t know is whether they can be as effective, or more effective, than sitting in a classroom. It’s time to focus on that harder problem: engagement.
As an online learner myself it’s hard to stay engaged: when I get home after a full day at work, the noble goal of learning new skills often is put aside with Netflix only a click away. Online learning needs to pass that test: it needs to be not only good for you, but enjoyable in its own right. Startups, big companies, and universities are finally focused on the true promise of online education: driving substantial learning at an affordable price. There are three areas that have shown promise towards that goal: mentorship, retention marketing, and new forms of learn-by-doing.
One-on-one mentorship was long ago found to be dramatically more effective than group instruction. Having the full attention of an instructor accelerates an individual’s learning by focusing them on the right problems at the right times, and having a real relationship with one person provides students with accountability. At Thinkful, we see a spike in learning the day before students have sessions with their mentors. Students want to achieve more because of their relationship, and that motivation translates to more efficient learning. We’re now working to apply that same social pressure throughout the week to bring up overall learning time further. Sometimes you just need someone you respect telling you to eat your vegetables.
Second, retention marketing needs to be brought to learning. There’s an entire discipline in marketing to existing customers through email, phone calls and push notifications to keep them engaged with your product.
For example, if a Dropbox user views a page to learn about Dropbox for Teams, Dropbox sends an email saying, “Thanks for checking out Dropbox for Teams! Here’s a few features you might not have known about…” By putting a product in front of the right users at the right time, online learning companies can instill good learning habits. Sometimes all that’s needed is a simple nudge to remind a student to spend more time learning.
Third, new technologies enable methods of “learn by doing” that just weren’t possible before we could deliver immersive experiences to people’s laptops and phones. In the 1960’s, Jerome Bruner expanded an educational theory known as constructivism with the idea that students should learn through inquiry under the guidance of a teacher to grasp complex ideas intuitively. That process of trial, failure, and then being shown the correct path has been proven to drive student motivation and retention of learning.
What we don’t yet know is if that process of trial and failure can become 10x more engaging when delivered through a new medium such as Minecraft or Oculus. There’s been promising early results around using Minecraft for learning, and we’re excited enough about the possibility of using Oculus that we’ve got one shipping to the Thinkful office as we speak. These new immersive worlds promise to hold the attention of students in ways textbooks never could.
Finally, there’s one other trick education startups use that works but hurts everyone in the long-run: making courses easier. Companies see low completion rates and cut important material as a “quick fix” to higher graduation rates.
When Udacity launched, its initial courses dove as deep into subjects as their Stanford counterparts. Now, many Udacity courses take as few as 40 hours to complete; if Stanford were giving course credit for a week of work, it wouldn’t have the reputation. Soon we’ll be hearing about the “dramatic rise” in MOOC completion rates, without a mention of how much easier it is to complete a course.
Those “improvements” hurt the entire industry as the actual outcomes – jobs, promotions, improved productivity – still won’t be delivered. As we’re seeing with thedecline of the for-profit universities, quick fixes that sound good in the press don’t work in the long-term.
The future of online learning isn’t about accessibility: it’s about taking what we already know works offline and combining it with what you can only do online to create the most engaging experience. We’re all still searching for the right formula, but the ingredients will be the same as they’ve always been: Learning through exploration, thoughtfully designed for the right behaviors, with great teachers providing support.