Schools out of reach
Daphne Koller is a third-generation PhD, and in her own words, she is certainly one of the lucky people. Most people, of course, are not.
In some parts of the world, quality education is simply not available. In South Africa, for example, the higher education system was designed during Apartheid, and there are far from enough positions available. Earlier this year, thousands of people lined up to get a position at university, resulting in a stampede. 20 were injured, and a woman died trying to get a place at university for her son.
Even in the US and similar countries, higher education is not always within reach. Tuition is going up at twice the rate of even the spiraling costs of health care. And yet only a little over half of recent grads are working in jobs where higher education is required.
The opening of education
Koller thinks now is the time when major change can come to education. Quoting Thomas Friedman, “Big breakthoughs happen when what is suddenly possible meets what is desperately necessary.”
She has found what is suddenly possible. At Stanford, professor Andrew Ng offered a machine learning class for free online. He found that 100,000 people enrolled. To get the same number at Stanford, he would have had to teach the class for 250 years. Inspired by that, Ng and Koller founded Coursera, an site for free, online education. Currently, there are 43 courses from four universities with more to come.
In its short time, it’s been successful. ”Unsurprisingly,” Koller says, “students like getting the best content from the best universities for free.” To date they have had 640,000 students from 190 countries, who have view 14 million videos and 6 million quizzes. When she submitted her TEDTalk slides to our backstage techs a week ago there were 1.4 million enrollments, but she had to update a slide to 1.5 million as of the date of the talk.
Of course, the numbers only tell part of the story. Koller points to some remarkable stories, such as a father whose daughter is immune-deficient, and he couldn’t leave the house for fear of infecting her. The daughter has since gotten better, and the father has a job thanks to taking the courses.
What is different?
There have been many attempts at novel types of education. What is different about this one? According to Koller, the key is that “this was a real course experience. It started on a particular day, and the students did real homework for real grades with real deadlines.” At the end of the course, they receive a certificate.
As with many activities that have moved online, one of the main advantages is that the professors can move away from constraints imposed by traditional methods. Instead of a 50-minute “hour,” the material can be broken up into modular chunks. Students can traverse this in different ways. Different students might need background material, or some might want to supplement it because of their interests.
“Of course, we know that students don’t learn by just passively watching videos.” Koller has found ways to engage students using techniques that are proven to produce results. Simple retrieval practice is an enormous help — just asking them to retrieve something they learned increased the learning. So they built that into the video itself.
Additionally, the video doesn’t just run, it pauses and asks a question, so the student has to show they understand to be able to continue. These are the kinds of questions the instructor might ask in a classroom, but in a typical lecture only one student answers, and most don’t even realize a question has been asked.
Are the students learning?
This isn’t the final answer, you need deeper exercises. But that brings up one of the most difficult questions, how do you grade 100,000 students? Some types of material can be graded automatically, and Koller shows some beautiful examples that surpass simple multiple choice.
Of course, they can’t grade all types of work that way. For other disciplines, such as in the humanities or business they have had to build in a different paradigm. Surprisingly, “It turns out that peer grading is a surprisingly effective strategy.” Even better than that, shockingly, are self grades — if incentivized differently. Both of those also turn out to be good learning exercises.
The students have also become strongly involved in their own education. “This is not just about sitting alone in living room,” says Koller. A global community has developed. For example, they’ve developed forums for students to ask each other questions. And in some classes, the median answer time was 22 minutes, which is, Koller frankly admits, “not a level of service I’ve ever offered my students.” There are virtual study groups and physical study groups, some multilingual.
Learning from the learning
The potential for transforming education goes beyond the ability to teach lessons online — the data collected is unprecedented. They can collect every click, homework submission and forum from tens of thousands of students. “That can transform the study of human learning from hypothesis-driven to data-driven mode.” They can understand which methods are the best, and can understand specific misunderstanding. Koller shows an example where a large number of students — 2,000 — gave the exact same wrong answer to a quiz question. Why? They used that to improve the teaching method to address that specific misconception — and offer a targeted error message to students who submitted the answer, to steer them back on the right path.
Personalization is one of the biggest promises of this method. (The exact opposite of what many fear from massive online education.) Koller describes the 2-sigma problem — which is that an individual tutor is better than lecture based by what is called 2-sigma, which means that 98% of the individually tutored students were better than the mean of the lecture based class. That is an extraordinary difference. “Imagine if we could teach so that 98% of our students are above average. We can’t give each student an individual tutor, but maybe we can afford to give them a smart phone or a laptop.” And the computers, she notes, don’t get tired of grading the same work, or repeating lessons if the students need it.
An extraordinary future of education
Are universities obsolete? Mark Twain thought so, writing, “College is a place where a professor’s lecture notes go straight to the students’ lecture notes, without passing through the brains of either.”
Koller disagrees, and thinks a better quote is one from Plutarch: “The mind is not a vessel that needs filling, but wood that needs igniting.” The question, of course, is how do we light it up?
She wonder what would happen if we could offer the best education to everyone around the world. She thinks:
- It would establish education as a fundamental right.
- It would enable lifelong learning.
- It would open the door to wave of innovation, “…because amazing talent could be found anywhere.”
The audience agrees, rising to a massive standing ovation.
Photos: James Duncan Davidson