A Q&A With Jose Ferreira, Founder And CEO Of Knewton
Join our live discussion on Monday, November 25th at 11:00 a.m. (ET) on what big data and analytics mean for the future of higher education.
Jose Ferreira, founder and CEO of education startup Knewton,will answer your questions live on Monday, November 25th at 11 a.m. ET. Submit your questions now!
Can you imagine 100% of the people on the planet graduating high school nearly for free?
What do personalization, customization, big data and analytics mean for the future of education?
Jose Ferreira, founder and CEO of education startup Knewton, is uniquely positioned to answer these questions. His technology platform, which was named one of Fast Company's Most Innovative Companies in 2012, powers content created by the largest education companies on the planet, from K-12 through college level. The system is capable of analyzing millions of data points on each student as they move through content tagged at the concept level. Courses powered by Knewton's technology have shown higher pass rates, and have students completing material more quickly with better motivation.
If you have questions about the impact of big data on education, come talk with us.
Submit your questions now by using the "Make a comment" box below!-
Jose, I want to start with a quote that appeared in Fast Company in 2009 : "The Internet disrupts any industry whose core product can be reduced to ones and zeros," says Jose Ferreira, founder and CEO of education startup Knewton. Education, he says, "is the biggest virgin forest out there."
-
That sounds like something I would say!
-
Can you expand on what you meant by that, and how far along do you think we are in the process?
-
So -- every "media" or "information" industry is transformed by the Internet
The last two are going now, video and education -
The media component of bricks and mortar industries goes too -- witness online travel, online real estate, etc.
-
Now education won't go entirely, the way the yellow pages did, but it's so big that even a fraction of it going will change the world
-
What does that partial transformation look like, in your opinion?
-
95% of materials (textbooks, software, etc used for classes, tutoring, corp training...) will be purely online in 5-10 years. That's a $200B global industry. And people predict that 50% of higher ed and 25% of K-12 will eventually be purely online classes. If so, that would create a new, $3 trillion or so industry
-
so that is an awfully big virgin forest -- bigger than the rest of the Internet combined
-
In terms of revenues flowing, user hours logged, and bandwidth required, it will be bigger than the rest combined
-
More importantly, as a tool for good, it has awesome potential
-
That's pretty mindblowing to contemplate. What does it mean for learners?
-
Here's my piece (which includes Knewton) from back in 2009.
-
Knewton was also named one of Fast Company’s Most Innovative Companies in 2012.
-
A few things. The Internet really does only two things: distribution and data mining (which, depending on the industry involved, can yield highly personalized products that are much more valuable to users). On the distribution side: expect to see a decade-long movement promulgating tablet/device-based learning all over the world. The best teachers in the world available to EVERYONE for the first time -- it is this idea that has people so excited about Khan Academy (an archived version of this) and MOOCs (a live version of this). In the rich nations these will usually be supplementary and will sometimes be replacement products. In the developing world -- or any places with very bad schools or no schools at all -- these will be mostly replacement products.
On the data-mining side, it turns out that education materials are extremely data-minable -- much more so than any consumer web industry. Which means we're entering an era when every student gets a unique experience each day based on exactly what she knows and how she learns it best. -
Well, and that brings us directly to what Knewton does. : ) Can you describe the technology?
-
While we wait for Jose’s response, here’s a video on the Knewton adaptive learning platform:by Miles Kohrman via YouTube 11/25/2013 4:19:22 PM
-
We are an infrastructure platform that content owners -- whether individual teachers or giant textbook companies or anybody in between -- can use to add very deep data mining and data-driven personalization to their content. For students using Knewton-powered products, we can tell what they know and don't know and where the critical gaps are in their concept profile that will hold them back. So if next Thursday they're going to struggle with a particular concept because they never learned some of the building block concepts from last month or last year at a high enough proficiency, we know that. Rather than let them get to next Thursday, fail, and have no one ever figure it out, we know in advance and we don't let them fail. We seamlessly blend in some of that prior content to get them ready.
-
Or, if a student is advanced, we can get them some more challenging material so they don't get bored and stay engaged. If it happens that you learn something best a particular way -- maybe you learn math better in the mornings, or science best in 24 minute bite sizes, or Civil War concepts best with a bunch of easy practice questions -- if that kind of thing is ever true of you for any concept or subject matter, Knewton knows it and that's how you can get it.
-
Can you talk about the initial math courses you created, for example, and the kind of results you're getting?
-
We work mostly with big publishers as of now -- Pearson, Houghton Mifflin Harcourt, Macmillan, Cambridge University Press, etc. We have just launched a closed beta version of our platform for teachers. That open "anyone can use it" platform that's now in beta will launch next summer with tons of pre-loaded content anyone can use for free, and all the functionality of Knewton
-
sure--back when we were getting started, no one knew what "adaptive learning" was. No one used the term. So we had to help create the industry we now inhabit. We envisioned an industry where everybody would make their own adaptive learning apps. There would be no NON-adaptive apps. And we wanted to build Knewton to power all those apps... by building all the difficult infrastructure to make them work really well, infrastructure that would be way too expensive and difficult to build for any one app.
But there was a problem -- no one new what a Knewton style adaptive learning app looked like or what it could do. So we built a couple on our own -- even though we never wanted to be a content creator or app creator. We built a test prep course, and we built a college math readiness course. The test prep course we built for GMAT was effective enough that we could offer everyone a 50 point money back guarantee. That was 95% successful -- hardly anyone didn't get at least 50 points. Kaplan only offers a 10 point mbg, for instance... Our college math readiness course improved pass rates at big US universities by 10-20 points consistently. At one of our flagship partners, ASU, we increased pass rates from 66% to 78% -- reducing drop outs by around a third. When you do the math on 3.5 years of additional tuition X hundreds of extra students X $15k average tuition, you can see why these schools really like that course. -
that's a Pearson application we power called MyLabs
-
Now, adaptive learning doesn't work for EVERY subject, does it? What are the subjects it's currently being used for by the different publishers?
-
No it doesn't work for everything but the good news is that it still does a whole lot. Basically, it works well wherever teachers would agree that there is a "right answer" -- so we can power nearly all of math and science, but even a surprisingly high amount, say 70-80% of history, etc. We can't power creative writing courses but we can do a lot with essay writing -- since nearly everyone agrees on things like grammar, syntax, topic sentences, one idea per paragraph, logical flow, etc,. The only Pearson subject we don't power -- out of all their titles -- is Dance.
-
Love it. We need to get to some of these great questions from the audience. Why don't we start by talking about the role of teachers:
-
Question: one of the big struggles I have observed is getting instructors of varying ages and tech literacy to adopt, implement and effectively use learning platforms for their content. How does Knewton plan to teach the teachers when they haven't had access to the insight data mining can deliver?
-
And you might also address the reactions of faculty members at universities:
-
How do you think Big Data will appeal to university administrators and faculty members?
-
Great question from Kate -- I've always believed you can't force already busy teachers to do extra stuff if you want to have a positive impact. If teachers don't want to dive in and figure it out yet, the magic can still happen at home while students do individualized homework. They should come to class better prepared than if they all read the exact same homework in the same order, the same amount, the same media format, the same practice questions, etc. So that will be some help...
we do offer teachers a lot when they are ready to engage -- incredible data and predicative analytics on their students, a wealth of free open content that has performance data attached to it (this piece of content has been proven to work well for this type of student, etc), the ability to group students into pods and bring personalization into the classroom... -
it almost sounds like you're trying to make a technology that's "teacher-proof."
-
We've got some questions about reactions on campus: it's been generally very positive, although there have been hiccups for sure. Pretty typical stuff when you roll out a new technology. Some faculty have been worried about the logical extreme of Anya's comment about "teacher-proofing" -- that is the opposite of what we want to do. Knewton makes the teacher MORE important, not less. All we do is take out the dirty work of education; Teaching basic concepts shouldn't happen in class and good teachers don't want to do it. We liberate faculty to get away from teaching to objective standards, since "the robot" can do that well, and focusing on the most important stuff -- the area where there is no right or wrong answer but where critical thinking and judgment exist. We do the concepts you need to learn to be able to get to the advanced ideas, allowing the faculty to focus on those advanced ideas.
What we've found is that on campuses where there is initially some pockets of faculty resistance, because it hasn't been explained right, etc., that that resistance goes away after the faculty use it and can see the above in action for themselves, And then faculty get addicted to the data -- seeing which students know which concepts, at what level, how the class is progressing, etc. -
Related: here's an article about how Knewton's adaptive learning Technology is being used by tens of Thousands of students at ASU, Penn State, SUNY, and more.
-
How does Knewton integrate with, or reduce the need for the standardized testing that seems to rule to day in some states, in many places thanks to your partner Pearson?
-
While we wait for Jose’s response, here’s an infographic that Knewton produced on the state of digital education. (Link)
-
I think we will reduce the overall amount of it in general, as well as reduce the focus on it. Standardized tests are also "big data" in education -- they're just bricks and mortar big data with extremely small data sets. Rather than having so many very high stakes summative assessments based on small data sets, Knewton allows for much more gentle, passive data collection via ongoing formative assessment in homework/classwork. We can predict your score on a bunch of these high stakes tests anyway, lessening the need for so many of them. But, more importantly, we can do the dirty work of "teaching to the test" passively in homework so that teachers don't have to waste class time doing it.
-
I like that--- "gentle" data collection. Here's another question about what Knewton supports in the classroom:
