Active Learning and the Future of Education

In this interview, Anant Agarwal, founder and CEO of edX and professor of Electrical Engineering and Computer Science at MIT, reflects on active learning, the current value of micro-credentials and the future of modular education.

edX has more than 25 million students from 196 countries, who have taken 80 million courses on the platform.

Transcript

THE BIRTH OF EDX 

Anant Agarwal: You know, MIT has done an open course, almost 20 years ago in 2000, and my course was an MIT course, and so, MIT’s a pioneer in, you know, putting course material online, and then I was seeing technology advance so much and technology being applied to almost every aspect of humanity, but difficult technology wasn’t being applied to education.

And so, to me, the challenge was labs, labs were going to be very hard, how are we going to do labs? So in 2003, I launched one of the first circuit labs at MIT called Websim. Even today you can Google Websim MIT, and you’ll go to my website and a built-in lab myself because, well, before books, it was the first book lab and, on an average day, 300, 400 people would come and do a free, a set of free laboratories. That gave me confidence that you know, this can work. That people from all over the world can not only learn, like, with an open course, with open course materials, but we can also do interactive labs, and this gave us confidence.

Then in late 2011, when MIT and Harvard began together to think about launching an online learning platform, I was so enthusiastic to be part of it and to launch a non-profit approach to online education and to build a platform with simulation and all of these technologies. In fact, the first course on edX was the circuits course my colleges and I taught. And we had a circuits lab inspired by the circuits lab that I had built in 2003, so that’s how I kind of got caught into it, somewhere by accident.

Today, on edX, we have 150 of the world’s best institutions on our platform, including your own Tec de Monterrey. We have Harvard, MIT, Berkeley, Oxford, Cambridge, some of the top universities in the world and top companies like IDB and non-profits, and companies like IBM, Red Hat, Microsoft, and the No Limits Foundation are offering courses.

We have almost 25 million students on the platform, from 196 countries, which is almost every single country in the world, and these 25 million students have taken 80 million courses on the platform. We have more than 3,000 courses on the platform today, we have almost 60 micro-masters programs, we have more than 100 professional certificate programs, we launched a Spanish platform, today we have 300 courses in Spanish on a platform, and we have more than 3 million learners on our Spanish platform.

LEARNING FROM EDX DATA

Anant Agarwal: We think of edX as a particle accelerator for learning, gathering all of this incredibly rich learning data, and we’re learning so much about learning. We make all the data available to partner institutions. We have something called RDX, research data exchange, where universities that opt-in can share the data, research data for research purposes, which is very radical. Just as one example, something radical that we’ve learned: did you realize that in 2012 edX discovered, through this research data, that 6-minute videos were ideal? We had a researcher at edX, Phillip Gould, and he analyzed 5 million video viewing sessions for a number of courses in edX. He plotted student engagement versus video length and found that 6-minute videos were the optimal size. 6 minutes. Today, everybody thinks 6 or 7-minute videos are optimal, and that came from an edX study done in 2012. So that is just one example of what we can do with this big data of learning, and we can look at millions of viewing sessions and, from there, derive insights into how students are learning.

Researchers analyzed data and learning patterns from MIT; this was Lori Breslow at MIT, whom you might know. When you come to campus, you go watch a lecture, and then they give you problems to work on, so you learn something, you hear something, and then they ask you to apply the learning in a problem. Also, on edX, not surprisingly, when we look at all the material available, the problems, the videos, all available for the students all the time, so what we found that, generally, 70% of the students start learning with the video, and then they go and do the problems. 70% of the students start by learning from the video. But if you look at it towards the end of the course, only 30% of the students started learning from the video, towards the end of the course it flips, completely flips. 70% of the students start to answer the homework and problem sets, and depending on where they don’t understand the question, they go and watch the video. So, students, their learning is much more inspired when you give them a problem to solve and then ask them to go and watch a video to learn, it’s a much better way of teaching. When you give them a problem, they get motivated to learn, and then they go and learn.

This was a very interesting result. It showed that when students were given more problems than videos because they’re used to watching videos first, they started watching videos first, then 70% were watching the videos first, but over time, 70% began to answer the problem first. On the edX platform, we gather all the data, so we know what students are learning, when they’re doing it, we know everything.

MOOCs COMPLETION RATE CHALLENGE

Anant Agarwal: It’s very difficult to apply 18th-century metrics to 21st-century technologies. For example, when you go to a movie, if you leave the movie, you pay 10 dollars for a 2-hour movie, and if you leave the movie halfway, that is not good. So you can say that the goodness of a movie is the completion rate. How many people went in? How many people stayed until the end of the movie? That is a good metric for a movie. But today, everybody watches 5 minute YouTube videos, is completion rate a good metric for YouTube videos, where it’s free? It’s easy, you’re watching a YouTube video while on the train, how many people do you know that watch every YouTube video until the end? I’ve must have watched, I don’t know, a hundred videos. I haven’t watched a single YouTube video until the end. So, you cannot apply the same metric that you apply to movies in a theatre where you paid 10 bucks for YouTube videos, the two are different things. It’s like applying 18th-century metrics to 21st-century innovations.

Similarly, I don’t think we can apply the same kind of completion metrics to books, because you can start for free. You can’t apply completion metrics from the 18th century to a product that is free, which just with a click, you can start doing it. Instead of better metrics for books, is the completion rate for people that have signed up for a certificate. So, once you’ve signed up for the certificate, then you’re saying “yes! I want to complete it and am interested in completing it. You’ve skin the game, you’re paying, you have – it could be a small amount of money, it could be 25, 50 dollars certificate and you’ve signaled a seriousness to complete it. Now it’s fair to measure completion rates because when you come to a university, y
ou’ve paid university tuition, you go to IVY League university, you’re paying tuition of 45 thousand dollars a year. So, of course, you’re going to complete the course. The course costs 6 thousand dollars. You pay 6 thousand dollars, you’re not going to walk out of the course in 5 minutes. You complete the course. And so, similarly, with books.

I think the way the completion rate should be measured is the percentage of people that pass of those who have signed up for the verified certificate, and on edX, that percentage is more than 60%. For people that sign up for a verified certificate, 60% of them pass. And these are hard courses, these are not… If you take the circuits course that I teach, it’s MIT-hard courses, probably one of the hardest courses at MIT.

THE VALUE OF MICRO CREDENTIALS

Anant Agarwal: These micro-credentials, micro-masters in particular, have two values: one, it is a valuable stand-alone credential. Imagine you get a micro-masters from EGADE at Tec de Monterrey in Entrepreneurship and might as well be one of the top universities in the world in Entrepreneurship, that means something. You get micro-masters from MIT in manufacturing, from MIT, that is a value in and of itself. At the same time, if you do well in the micro-masters, you can apply for a master’s degree and get admitted into a Masters, and so it gives you a pathway into a Masters, so there are two sets of values.

On edX, 70% of the students are interested only in the micro-masters, but 30% of the students are interested in following up and continuing for a master’s degree. To me, to take surveys of learners who have completed micro-masters and, of the surveys they have done, of learners who have completed micro-masters, we do the surveys 3 months after they’ve completed, and we keep doing surveys all along, but 3 months after they’ve completed, we are finding that learners tell us that 80% of them have a career advancement, 87% of them have had a career advancement in 3 months. What is career advancement? A pay raise, a promotion, a new job, so this is a big deal, this is what learners are telling us.

As companies hire more and more and see more and more progression from micro-masters to learners, we believe that companies will also begin to recognize micro-masters, and we’ve seen this already. A company in India, Tech Mahindra, one of India’s top three software companies, guarantees an interview to anybody from India that completes one of 10 micro-masters, completes one, and they’ve listed 10 micro-masters, they will guarantee an interview. Similarly, in the US, GE will guarantee an interview to anybody from Massachusetts who has completed a micro-master’s in supply-chain management, in AI, in cybersecurity, and on other subjects. So, companies, more and more companies are now asking their employees “do micro-masters,” many of them are beginning to guarantee interviews to micro-masters learners, in hopes that this will keep building up as we get more and more students who are completing micro-masters and go to companies and make their impact known.

ACTIVE LEARNING IN MOOCs

Anant Agarwal: From the very first day, when people say that the pedagogy of edX is a knowledge container, they are not taking any edX courses. The very first edX course, and any course you do on edX, is based on active learning, where we invented this concept called the learning sequence. So, edX, you see a learning sequence, where videos are interleaved with assessments and the learning sequence. You show video, and then you ask a question, maybe you ask the students to go to the discussion forums to discuss something, so transmission of knowledge and questions are interleaved, and that creates, so the basic pedagogy of the platform is active learning.

More recently, it has advanced much more. Three years ago, we introduced peer instruction at edX. Eric Mazur wrote the book on peer instruction, so now we have a peer instruction pedagogical block on edX, where you can teach something, you can ask students to submit answers, you can take a pool of students, you can do a cold call, you can ask a student a quick question, and then you can ask the students to discuss and then you can ask the students to resubmit their answer. It’s peer instruction, so we can do that today, a lot of courses are doing that.

A MODULAR FUTURE OF ONLINE LEARNING

Anant Agarwal: But my dream is to really create a very modular education system, where at an undergraduate level or graduate level, learners can take these modular credentials, like micro-bachelors that can be at an undergraduate level or masters at the graduate level. Learners will be able to take these credentials like LEGO blocks, they will be able to take them and then stack them up to full degrees.

I believe they will also be able to take micro-bachelors and micro-masters from multiple universities and create a truly networked module of education where they will get degrees from universities. Of course, these modules will come from different universities. They will synthesize their own degrees and complete degrees. It will be a very low cost for anybody, and everybody will have access to this kind of education. A truly modular network education environment. It’s completely flexible, you can tune it to your personal leads, and you can also be a lifelong learner. You can say, “Look, I want to earn a bachelor’s degree in the first four years. You know what? I will learn for one and a half years and will get four micro-bachelors, and then, you know what? I want to go and work at that company, and while I’m working, I’ll keep learning.” I think you said just-in-time-learning, “I’ll keep learning just in time, and I’ll keep accumulating micro-bachelors over my career, and maybe 4, 5, 6, 7, 8, 9, 10 years later, I will get a bachelor’s degree.”

You know what? Maybe in 20 or 30 years, maybe bachelor’s degrees won’t be important anymore. People will just look for the competencies you’re getting from these micro-credentials, and that’s it.” Once you’re learning while on the job, the bachelor’s degree comes later, suddenly it’s not important anymore.

ObservatorioIFE

Share

This article from Observatory of the Institute for the Future of Education may be shared under the terms of the license CC BY-NC-SA 4.0