Sal Khan: Test Prep Is ‘the Last Thing We Want to Be’
For most of us, hearing something just once isn’t nearly enough to commit it to memory. But with today’s crowded curriculum, sometimes one explanation is all kids get.
Ten years ago, Sal Khan set out to change that with his Khan Academy videos, which let kids replay lessons as many times as they want. Here in education, Khan doesn’t need much introduction. On YouTube, Khan Academy’s collection of videos (filmed at first in his bedroom closet) started off covering math but have since branched into science, history and other subjects. Collectively, they have more than a billion views and reach about 18 million learners a month in 40 languages.
A lot has changed for Khan in the last decade. He’s started working with more schools and districts—instead of reaching students mainly at home—and has partnered with assessment groups, such as the College Board and, most recently, the NWEA for personalized test prep materials. The latter group administers the MAP Assessment—a test designed to measure growth in reading and math. But Khan says it’s not all about cramming for a good score, and the ideas behind giving kids more time to master material is still a big part of the organization’s ethos—and that of the Khan Lab School, which is Sal Khan’s small private school in Silicon Valley.
At the 2019 National Charter Schools Conference this month in Las Vegas, EdSurge sat down with Khan to discuss his vision for reinventing schools, the focus on testing and what he thinks about the recent stumbles of AltSchool, a nearby network of tech-driven independent schools.
Listen to the discussion on this week’s EdSurge On Air podcast. You can follow the podcast on the Apple Podcast app, Spotify, Stitcher, Google Play Music or wherever you listen. Or read a portion of the interview below, lightly edited for clarity.
EdSurge: The theme of the conference is reimagining education. What does that mean to you?
Sal Khan: I think there are a lot of levers where one could reimagine education, but the one that we focus most on is on this notion of, in a traditional academic model, kids move ahead at a fixed pace. A teacher will cover a topic for a couple of weeks and then there’ll be an assessment. And on that assessment one student gets a 70 percent another student gets an 80 percent another student gets a 95 percent and even though the assessment identified those gaps, the class will then move on to the next concept and those gaps keep accumulating.
The stat that I’ve been citing a lot lately is that kids get to community college and 70 percent have to take remedial math, which is a euphemism for sixth or seventh grade math. And so the thing we focus most on, is how do we create content tools, software to allow teachers to allow their students to remediate gaps if they have any, and work on something long enough until they’ve gotten to a reasonable level of proficiency or mastery and then move on.
It doesn’t even have to be wholesale change towards that. It could even be 20 percent of class time in that direction. We’re seeing kids grow 20 to 30 percent faster than expected.
And the other 80 percent of class time would be business as usual?
You know, 500 years ago, that is the way that people learned. You would have a private tutor who would flex according to what your needs were. But then with free mass public education, which was a major innovation for the planet in the 18th and 19th centuries, we asked, “How do we do it at scale? We’ll get 30 kids in the room. We’ll move them all together. Some of the kids will keep up. Some of the kids not as well.” And then you get to a point where they hit walls.
So I’ve been preaching that wholesale mastery learning isn’t realistic because you have to really think about your grade system. There are state mandates, district mandates that you have to cover. So the compromise is if we can put an even—let’s call it 45 minutes a week—of students being able to work at their own pace on what’s appropriate for them, we’re seeing that even that dosage is driving gains for them.
Let’s talk about this MAP Accelerator partnership. How does it work?
So students take the MAP. The teacher is going to get notified via email saying, “Your kids’ scores are in. Click here to review suggested placement.” The teachers can then either say it looks good. Or, and this is something we care deeply about and the NWEA folks care deeply about, the teacher can modify them.
They can say, “Well, I don’t think that’s exactly right for Billy.” Once the teacher approves all of the personalized learning plans for each of the students based on what their RIT scores and the various common core strands, then the student gets an email notification saying, “You’re ready to get started on your personalized learning plan.” When they log in, it says here are the four strands you need to work on. And then it goes into the kind of mastery learning framework of Khan Academy. Work on this, get mastery, do it on a unit test, et cetera.
What we’re excited about is that will allow teachers, principals and administrators to understand how that work on Khan Academy is hopefully driving better than expected results on the MAP.
So why this strategic shift to testing? You talked about how Khan Academy was founded on a mastery learning principle. Are you worried Khan Academy will become all about standardized testing?
The last thing we want to be is about test prep. Even with our SAT prep. But you know, both we and the College Board said, look, this cannot be about just, ’How do you game the SAT?’ This has got to be about becoming more college ready. None of these tests are perfect measures, but they can be indicative. There’s actually a very healthy separation between the teams that create the assessment and the teams that create the content.
If you look at the content on the MAP accelerator, it is common core aligned content and it’s all from the standards in the common core and not just what may or may not be in the MAP. And it is done independently from looking at MAP assessment items. So we believe that the students, by doing this work in a mastery framework, will learn the common core standards better. And MAP believes that they are a good measure for the common core or mastery of common core standards.
I want to talk about the Khan lab school. How’s that going? Successful?
It’s doing just fine. It’s approaching 200 kids now, K through 11 because the oldest kids are going to be 11th graders. But it’s been great.
I wrote “One World Schoolhouse” back in 2012 just kind of theoretically asking, ’How could education be re-imagined?’ And it’s one thing to write about it and say in theory. It’s a whole other thing to actually try and do it. And I have to say I’ve learned a ton about the reality of real kids coming in every day. Real parents, real teachers. The biggest successes have been with the thesis that you can decouple academic progression from age.
It’s true, although in difficulty there’s a lot of edge cases. Like what if a kid isn’t on track to get to trig or calculus by the time they’re 17 or 18 years old? What do you do? Now what’s cool about this school is that we can flex. They can have more time, more interventions, maybe. I think the mixed age has been very powerful for seeing kids take responsibility for each other. In a lot of traditional environments, kids just worry about themselves. And I don’t know if it’s because it’s a small school, and I don’t want to jinx it either, but there’s not the kind of bullying behavior that you might see other places. They’ll sometimes have to tutor a 7-year-old or an 8-year-old. I think that builds empathy.
The difficult stuff has been having real students, parents who have their own concerns, and a team that is trying to build a school as they’re serving the students and trying to iterate and learn and try new things. It’s not an easy thing. The school was set up from the beginning to make it as easy as possible. It’s still not easy. I definitely appreciate how hard it is if you’re in a more traditional structure to do mass change. No matter how bad the teacher might want it or the principal might want it or the superintendent might want it. That’s actually informed some of what we’re doing at Khan Academy, which is how do we make it as time-saving as possible for teachers? How do we make it as empowering for them as possible and how do we make sure it doesn’t require a wholesale rebooting of everything they know.
Why do you think you’ve seen success with it versus other Silicon valley models like AltSchool, which have been pivoting away from running schools?
AltSchool started in a very different place than KLS [Khan Lab School]. KLS started in a place of: What could a school look like? And it started with some basic ideas around mastery learning, peer-to-peer, full year, full day. It was kind of about these educational ideas first, and then how do we build a school around it? We use tools as necessary. Like, Khan Academy could be useful here or there.
But if you visited the school, most of the kids are not doing Khan Academy most of the time. They’re like interacting and playing and if there’s another tool that’s interesting, they’ll use it. But that’s not the focal point of the school. I think AltSchool—and I don’t think I’m saying anything that their founders would disagree with—started off as a tech company.
They went and got funding from the same people who’d funded Uber and Airbnb and raised a ton of money from those venture capitalists who I don’t think were interested in a model that was going to start one school every four or five years. That’s never gonna give you a billion dollar valuation or multimillion dollar valuation. They were interested in some type of a thesis around starting some type of edtech tool that could be kind of the operating system of [education]. It wasn’t obvious to me how it was going to scale fully as edtech. And it wasn’t fully obvious how it was going to scale as a school. When their Palo Alto campus shut down, a lot of the parents and families and even some of the teachers were interested in ours because we’re only four miles away.
In the Khan Lab School, you mentioned that the kids aren’t doing Khan Academy all day. Is there anything deeper that we can extrapolate about your educational vision from that?
Yeah, I mean I’ve been actually trying to work with the school team on that. It’s about what we can do for the students to give them optimal agency. And then what supports can we give to the teachers so that they can provide that optimal agency for students? It should always be with that vision, and then you think about what tools might be appropriate for it. It might be a spreadsheet, it might be a big whiteboard where kids can ask for help and offer help.
A lot of it is mindset of the faculty around helping that student become independent in whatever they need to do. We see it with our own kids. If they fall, your first temptation is try to pick them up.
And I do that if they fall, but metaphorically, when they’re learning, if they’re struggling a little bit, it’s all of our temptations. Like, “Let me just explain to you how you do it.” But we’ve seen that it really is powerful to ask the kids, “What do you think you should do about that?” No one expects them to take responsibility like that. But if you do it, they do it.