Dan Schwartz is the brand new Dean of the Graduate School of Education at Stanford University, but he doesn’t really want to talk about it.
“I don’t think I’ll have too much to say on my vision,” he says, when I first contact him. It’s not because he isn’t excited. He’s thrilled. It’s because he sees his role as a facilitator of the visions of others—“students, teachers, people who care deeply about education in school and out of school.”
Fortunately for us, Dr. Schwartz is more than happy to share his personal perspective on education technology.
He brings a wide variety of experiences to this conversation. He’s been a classroom teacher in wildly different places: L.A., rural Alaska, and Kenya. As an academic, he has played a big role in developing open-ended learning platforms, choice-based assessments, and teachable agents. He heads up the Stanford AAALab (short for Awesomely Adaptive and Advanced Learning and Behavior), which bridges cognitive science, classroom experience, and programming skills to build new technologies and theories of learning. And he’s advised several companies including Pearson, Leapfrog and Kidaptive.
In this Q&A, he candidly discusses the opportunities and pitfalls involved when academics and entrepreneurs work together to improve learning.
Note: This Q&A has been lightly edited and condensed for ease of understanding.
Hamburg: How do you see edtech and academia collaborating?
Schwartz: That’s a big question; I know there's a number of people who are thinking about this.
As an academic: I'm a great starter, and I can prove that something works. But my desire to ever have a 1-800-Call-Dan hotline for people who want to know how to use my inventions is like...zero.
Now, if its software, I could put it up on the Internet and let people use it. But I can't market it, and I can't keep maintaining the code. And I need to get a business plan, but I don't know how to do that. What I want is a partner who's willing to take a discovery or an invention and bring it to scale.
On the other hand, in the edtech industry, they come up with ideas that tend not to incorporate the kind of expertise that academics have about learning. And so they go to scale, but they may not have a product that's particularly effective or fresh, or that differentiates itself from the rest of the market.
So we’ve got to solve this problem!
For example, I can imagine making some sort of industrial affiliates program, possibly at Stanford, where there are people who have entrepreneurial talent and can give guidance to faculty who would like to commercialize their products.
So far, though, I haven't seen a lot of things like that that have been effective. Largely because there's not enough expertise on learning; there's more expertise on business.
Another approach that the industry has tried is to get academics on their advisory boards. It also doesn't really work, from what I've seen, though it could. The reason is usually that the company has already made a bunch of decisions, and then they want the academics to provide advice on that. But most academics are starters. They'll be invested to the degree to which their ideas are being infused in the product. It makes for a tense relationship. The academics come in and they do what they do very well, which is to criticize, and the companies are like, “Ow, ow. Why do I want you here?” And meanwhile the academics are thinking, “You're not listening to what I’m telling you!” A better model is to get the researchers who have expertise to come in during the early design phase.
On that note, can you tell me a little bit about your experience as an advisor to edtech companies like Kidaptive?
Yes, I work with Kidaptive, and I also advised Leapfrog, and at one point, Pearson.
Leapfrog was a large company; they had become toymakers more than educational toymakers. And they wanted to bring in researchers, people with title. They were more interested in the fact that I’m at Stanford than they were interested in me. And that didn't work very well, because they didn't particularly want input on how to make better products. They wanted more of a stamp of approval. Ultimately, the advisory board was disbanded because it wasn’t serving anybody’s purpose very well.
Kidaptive is different. One of the founders was my PhD student, so he already knew much of what I know. And at the beginning, they were going to take a certain direction and I suggested to them, “Everyone is doing that, why don’t you work on this instead?”
And since it was very early, they changed their direction. So that was very good! Because they cared about learning, and we collaborated very early, I could provide some direction on learning, and they could use their strengths in creative design and business planning.
Do you see any particular areas of academic research, perhaps being done at Stanford that you think are under-examined by the industry or might provide really interesting avenues for new edtech ideas?
In a lot of academic spaces, people are studying more open-ended tasks, as opposed to the drill and practice of closed tasks where there’s a single right answer. Many academics are trying to make inquiry modules for students in science, for example, and I think there have been pretty good advances in the conceptualization of those, where you could technologize it quite well. Just making a MOOC platform that organized cycles of inquiry, instead of linear “chapters,” would be a simple but huge improvement.
A lot of people are also trying to figure out how to make it easier to analyze the large data sets that come from open-ended task software, where every click and choice is potentially informative. Additionally, people are trying to come up with messages that explain the patterns in the data that are actionable for and interesting to teachers or parents or kids.
Also, people are working on games. But I think the difficulty of the games is that they still tend to be expensive to build. To make an interesting game is expensive, and the market isn't super.
Another issue is that nobody makes a lot of money in edtech unless they make a complete package. You hear, “I’m a teacher, I want a year's worth of stuff. You're going to give me a program that I have to figure out and I use once? Nope!” Now, I actually think making interventions that students only need once is a good idea; I’ve made lots of programs where we know if we can get kids to use it for an hour, they can be done with it, and it has these lasting benefits down the line. But right now there's not a market for that kind of thing.
Finally, there’s the idea of getting the complete data set on a student's experience at a private university and figuring out how to analyze that. For-profit universities keep very good records, and that data set is really interesting to see.
What about what industry can teach academics?
A lot of the relevant industries are about the world where people self-elect how they use their time. At school, kids don't self-select to play the games they’re given, so I can go into the school and I can make a little computer game that nobody would ever use on the market, and the kids are like, “thank goodness we're doing something besides what we usually do.” And they’ll play it. But the creative sensibilities of industry are much better; they’re excellent at getting people interested and engaged. Just think about the success of video games. They could add a lot of creative horse power.
So that's one place where I think industry is definitely ahead of academics, besides, of course, production and marketing. Scale is also huge. I’ve designed a lot of stuff, but I've never designed it with an idea of going to scale. So it’s a good reminder when someone in business says, “you understand if you do that at scale, you need fifty-seven thousand servers.” Like oh, right. I wish I had known that.
Have you collaborated with anyone in the industry recently?
We just collaborated with Camp Galileo. It’s a science and design camp that’s quite large in the Bay Area; they serve about fifty thousand campers. They asked us - or I maybe asked them - if they would like us to help them find out if they are succeeding in their mission.
So we collaborated very closely with them, and it was a very good education for us to see the kinds of things that are really important to them when they're at that level of scale.
Summer camp just ended. We have all our data - just reams of data. It will take a long time to sort through it all, and watch for different patterns, and figure out which patterns are real. So hopefully this year we will get closer to understanding exactly how to do this again, and next year we’ll be much more effective in doing the research.
That’s the fun thing about a university. I can be wrong and I still have a job. Hey!