Matthew Rascoff: But the idea is, if you take a course for credit with a real transcript at the end, there's an immense psychological boost that shows rather than tells that you can do college-level rigorous work. You know, it's courses with the same outcomes as we do on campus, same faculty, but with high-school-level supports. We bring in our own undergraduate students and recent alums to be teaching fellows in these courses. But it becomes a social capital intervention, alongside the psychological intervention of getting a real-deal Stanford transcript at the end, and the academic intervention of taking the course — like calculus or computer science — from Stanford, that sets you up academically for success with the requirements you need. And that gives a lot of confidence to an admissions officer. That's the same course that we offer our own freshmen on campus, so I can confidently admit that student, knowing that they'll thrive in the next course that they take when they show up on campus. But it goes beyond that, because to me it's a real learning experience, and I think it's an assessment model that's embedded in something that is deeply —
Aran: Welcome, friends, to Obviously the Future. The show explores the massive trends that will shape our world, in conversation with trailblazers, non-conformists, and hidden experts who are building tomorrow today. Who do we have today, Caitlyn?
Caitlyn: Today we have Matthew Rascoff. He's the vice provost for digital education at Stanford, where he leads the university's innovation efforts in online and hybrid learning. Matthew is a phenomenal thinker, pioneering this field. Before Stanford, he spent time at Duke, UNC. He previously launched JSTOR's first international office in Berlin, and started his career in edtech at Wireless Generation, where he built their 100% product development team in Durham, North Carolina. Matthew is a leading light in some of the ways that elite institutions in higher ed can build true equity into their agenda, so I'm super excited for this conversation. Welcome to the future, Matthew.
Matthew Rascoff: Thank you so much, glad to be here.
Caitlyn: I want to just dive right in. You have a very cool journey in the world of edtech and innovation in higher ed at some of the nation's best universities, but I'm interested in what sparked this initial journey for you. What got you into the education space?
Matthew Rascoff: I think for me it was a combination of great learning experiences that I had, and then miseducation experiences that I had that I thought were just misdirected. I think I was activated as a reformer by the combination of exposure to true excellence and its opposite, and I think that sparked in me a desire to make change, a recognition of what's possible in education and an effort to democratize that and make that excellence available to many more people. It's remarkable how much variation there is in this critical area of society. We figured out how to do consistent high quality at scale in so many other areas — why is the spread in education? That, to me, is the intellectual framing on it, but for me it was more personal. I had a horrible high school experience, and I think I was just, like, totally bored, just, like, a total underachiever in high school. And then I got to college, and the professor took me aside and said, what it takes to be successful in high school is not what it takes to be successful in college. Forget about that. Like, you're here, you have the world in front of you in terms of the potential to study anything. You have New York City, where you could have exposure to every culture, people from all over the place. Don't worry about what you were. In New York, you can be a new man, you know — Hamilton. That freedom, that sense that you can fashion yourself and expose yourself to difference and broaden your horizon, that was so amazing to me. It was kind of that transition from high school to college — of, like, the intellectually confining space of American high school, which is such a narrow world, and then the unbelievably broad, cosmopolitan space of American higher education. That was cognitive dissonance for me. Why can't we have more of that academic freedom and breadth and the beauty of learning?
Caitlyn: How does a high school underachiever get into Columbia?
Matthew Rascoff: Underachiever — to me, it's not like my grades were bad, Caitlyn. My grades were fine. I also did very well on the SATs, and I think, like, I was one of the cases where standardized tests helped, even though I'm a skeptic about standardized testing overall. It's relative to what my potential was, it's not relative to some abstract number.
Aran: Two things to pick up on in that answer. I can resonate with that point of view, because for me, during high school — even though I wouldn't call myself an underachiever, I would say it's probably the closest I came to — the light in my eyes, like, it was the closest I came to feeling completely disillusioned with what the potential of learning was. And only in higher ed was that fire re-sparked. Yeah, I felt that very much on a personal level. But secondly, when you brought up your skepticism of standardized testing — obviously Dartmouth, Yale have now retracted from going test-optional. There seems to be a trend moving towards bringing standardized testing back into the fold, and, from my very limited look at the data, it does seem like it's kind of like what they say about capitalism: it's the worst form of government except for all the others. I have a similar view of the role of standardized testing. It's got a lot of flaws, but it seems like it has the least flaws compared to so many of the other elements that are part of a college application bundle. So I'm interested in your views.
Matthew Rascoff: I think I agree with that. It's imperfect, but when used appropriately with other data about students, it seems to add something that might otherwise be missed. I recommend looking at the analysis that Dartmouth did. They did a very careful study about whether it was effective at predicting performance at Dartmouth, whether it meant they could identify talent that would otherwise not show up, and they answered yes, in their own context. They didn't do a broad-brush analysis of higher education. MIT did something similar. Schmill, the widely respected dean of admissions, published an essay when they reinstituted the requirement for standardized testing at MIT, and basically said, I've looked at MIT's own data and curriculum, which requires that everybody take calculus regardless of what your major is, and I can't responsibly admit students if I'm not confident that they're going to thrive in the curriculum we offer here. This test is predictive of your ability to succeed with our curriculum. So I guess that's my argument for thinking about this more locally, institutionally, rather than saying the whole system should do one or the other. You have to do this analysis with your own students, with your own curriculum, and think about, does it contribute to our mission, does it help us shape the class that we want to build. There's a massive change happening in the way we think about admissions, with the loss of affirmative action and the Supreme Court decision. It's a very fertile time, I would say. There's a lot of innovation, there's a lot of experimentation, a lot of new ideas in that space. It's one of the most creative spaces, I would say, in higher ed right now, because we've been forced to rethink the pathway to college. So, you know, I don't see what Dartmouth and MIT are doing as regressive. I think they're doing this in a way that actually takes into account their desire to preserve socioeconomic diversity in their class, and is looking for all the tools at their disposal in a dispassionate way that's not ideological, that's not, you know, biased in one way or the other. I think that's what institutions should responsibly do — not speaking on behalf of Stanford, by the way, this is not my domain at Stanford, but as somebody who's interested in building more equitable pathways to college and using digital learning to do that.
Caitlyn: So at Avalanche we have an investment theme called new measures of merit, which I think is a really critical part of the move towards skills-based hiring, competency-based learning outcomes, and advocacy embedded through both learning and earning. But it's really hard to underwrite, as a VC, companies that are attacking it, because of the potential institutional change that would have to go on. Have you seen any innovators or companies or technologies that you think might have some special sauce, or be getting it right? Where are you looking?
Matthew Rascoff: I applaud you for investing in a non-sexy but critically important area of education change. The assessment system is the underpinning of the education system. You have to be an insider to get that. It's not something that draws in, like, edtech entrepreneurs who don't have deep experience, who don't see how important it is in shaping the accountability system for the educators, the opportunity system for the students. The governance of the whole system is set by a fairly small number of psychometricians who created this system that we've inherited from the past. Some of it happened here at Stanford University, where the Stanford-Binet IQ model was imported to the United States by Lewis Terman. The system we've inherited is not fit for purpose, not even close. The issues with standardized tests, like for college admissions, are just the beginning of the problem — it goes well beyond that. I do think there is a lot of opportunity right now. I think generative AI may end up being an assessment system. The Jeff Bussgang study at Harvard Business School, in which he trained an AI on the cases that he teaches in his entrepreneurship class, and he used it as feedback — he could see what the students were doing, he could see their interactions, he embedded it in Slack, and then he built himself a dashboard to see what his students were doing with it late at night before class. How are they preparing, what questions did they have? And that gave him insight the next day into what he wanted to teach and what the students needed to know, where were the gaps. That's a formative assessment kind of application. I think what you're asking about is more of a summative assessment question. There's probably demand from companies focusing on skills — don't they need some measures other than credentials? So in theory there's demand from those kinds of employers to do it, but I agree with you, I don't think we know yet what they are. If you ask the psychologist, they would tell you what is predictive of performance, and it's not the things that we measure right now. I don't see a lot of companies that have figured out how to build a product around what we know from growth mindset, for example — building a product that measures it, enhances it, and then hires based on it. Quite the contrary, I think we're in a very fixed-mindset model.
Caitlyn: Yeah, it's really hard to get out of those systems because there's so much infrastructure, expectation, and culture built around them. To hear more about your venture at Stanford of building inroads into high schools for people who maybe don't have access, to get them up to higher levels and attack some of those, like, equity challenges — how do you think about that with the digital education initiative that you've been building for high school students?
Matthew Rascoff: So in some ways it is kind of a new model of assessment that is meant to track the indicators that we care more about. Our approach is to offer credit-bearing courses to low-income high schools, Title I high schools, and give students a chance to learn and be part of a learning community while they're still in high school. So you can see the connection to my own personal experience — like the desire to build more effective bridges for low-income students. The thesis is that the cracks, the interstices between the bureaucracies, like the K-12 bureaucracy and the higher-ed bureaucracy, are where most of the friction arises, especially for low-income students. Everybody experiences that as very stressful, the college admissions process, the ringer that we put everybody through unnecessarily. But if you're low-income, there's this added factor of undermatching, meaning the majority of the most talented low-income students never apply to any selective college. This is on the order of tens, hundreds of thousands of students in the top 10 percentile points of grades and test scores. The majority of those top 10 percentile low-income students never apply to a selective college. Wealthy students apply, get into, and go to well-matched colleges. If you look at low-income students, they go systematically to institutions that are well below their admissions capacity. This program that we've built, it's a systemic intervention in this undermatching challenge for low-income students who we know are at risk. And the intervention is, like, give them an opportunity to not just get a letter saying you're qualified based on your scores. There's been kind of College Board and other efforts addressing the information gap. There have been other efforts focused on, like, the financial anxiety — so, like, you didn't know you would get full financial aid, so there's kind of financial information and financial literacy intervention that was designed by Michigan and Susan Dynarski.
Caitlyn: Amazing. How effective is that?
Matthew Rascoff: That worked really well, that was very effective. We're trying to build on that. We integrate the financial literacy component into this model. The idea is, if you take a course for credit with a real transcript at the end, there's an immense psychological boost that goes well beyond an information letter, like, to your home, that shows rather than tells that you can do college-level rigorous work. You know, it's courses with the same outcomes as we do on campus, same faculty, but with high-school-level supports. We bring in our own undergraduate students and recent alums to be teaching fellows in these courses, and they're there to be near, enablers of student success. We've seen them become mentors over the course of a quarter, even. You know, the course, it's structured like academic support, but it becomes a social capital intervention alongside the psychological intervention of getting a real-deal Stanford transcript at the end, and the academic intervention of taking the course, like calculus or computer science, from Stanford, that prepares you, that meets a prerequisite, that sets you up academically for success. That gives a lot of confidence to an admissions officer, who can look at that student and say, they didn't just have a good test day, they actually got an A in this course, that's the same course we offer our own freshmen on campus, so I can confidently admit that student knowing that they'll thrive in the next course they take when they show up on campus. So you can think about that — you know, it is an assessment model, Caitlyn, to your question, from the perspective of the admissions officer, but it goes beyond that. To me, it's a real learning experience, an assessment model embedded in something deeply motivating, and motivating for the learners. They get to take a course for free. They don't pay — families, school districts, and states pay a modest amount. I think it has a lot of potential addressing these upstream challenges where we know there's inequities that are baked into the precursor systems. With the face-to-face tools that we've had historically, there's only so much you can do — take the applicant pool as it comes to you. Our thesis is that with these digital tools we can reach upstream and build more effective bridges. They connect the bureaucracies and design more from the learner, design more for what they need, rather than saying, you deal with our forms and processes and systems as they are. So that's the thesis.
Caitlyn: Can you give it some context here? How far along are you since you deployed, and what have been some of the early learnings?
Matthew Rascoff: Yeah, we've been doing this since the fall of 2021 with a nonprofit partner called the National Education Equity Lab. And they run a two-sided marketplace where they bring Title I schools to us, and they bring colleges and universities — I think there are 15 in their network — to these low-income schools. And they're kind of the brokers in between, that are making the matches, that set up the faculty and the courses that we offer with the students who would benefit from them. From the very beginning we were doing this nationally, in every time zone, from Hawaii to Brooklyn. We've had forty-three students come to Stanford from this program, which is pretty exciting. The goal is to set them up for success regardless of where they go. Other students have gone on to other great universities, and that's also equally rewarding for us. Sometimes we score the basket, sometimes we get the assist — they're both valuable in winning a team sport. We've done five unique courses, offered nine distinct times. We've done our CS course every fall for the past three years. It's an introduction to computer science, taught by a legendary instructor at Stanford. And we've done a course in academic writing, which is the biggest gap that low-income students show up to college with. Most people think it's STEM — it's not, it's writing. We designed this course to provide a lot more feedback than you typically get, to help scaffold how to write a term paper, which is very challenging if you haven't had support in doing that. And then we've done courses that you would never expect, like bioengineering, which — it's never been offered, as far as I know, in any high schools, but we have this amazing professor who thinks that the next generation need to be bionauts, the way we've had astronauts in the past, who are able to hack DNA and build new mRNA vaccines that solve the next pandemic. So he's teaching high schoolers how to do that. They learn CRISPR in their high school labs with him. It's a way of thinking about an introduction to biology that's much more application-oriented, connected to pandemics and vaccines and these things that everybody is aware of. And then the theory is, once you're motivated by the tools of CRISPR and how powerful they are, you'll fill in the biology later. Start with the thing that gives you the project, the application, the motivation, and then you can learn that later. That's classic Stanford, that's how we teach science here. So it's a much more project, much more application-oriented approach to science, that would never pass the Board of Education of some big state.
Caitlyn: Do you show the efficacy of that program and then try and push high school curriculums to adopt it further?
Matthew Rascoff: Yeah, so there is a study that's looking at the longitudinal impact of the program, every quarter, looking at the performance of students, their completion, grades, local indicators that you would expect. There's a study led by Robert Balfanz at Johns Hopkins, who's looking at all the students who participate in this program, using quasi-experimental methods, to understand the impact on their trajectory. We've seen with our own data it is impactful for the students who've come through. We've had about 1,500 students take our courses so far, so we're tracking them ourselves, especially the ones who come to Stanford, and we survey them. We also, by the way, survey our teaching fellows extensively, who are our own students and recent alums. They say it's the most impactful thing they've done at Stanford in the past year. In no way is this dilutive of the face-to-face residential experience that Stanford offers and will continue to offer. It's additive to those students. They love this program, they want to share their knowledge, they don't think it should be cloistered away and reserved for the, you know, select few. They want opportunities to share it. This program gives them an opportunity to do that, through our platforms, with our faculty, enabled by technology, well beyond what they would be able to do as a tutor, for example. So it's been very rewarding to see a synergistic investment in something that is educationally meaningful for our students, some of whom say it's reshaped what they want to do with their careers, and also impactful for the wider world. And I think that's our job in legacy institutions — we've got to look for sustaining innovations that advance our mission and also address some need.
Aran: Yes, but let me ask one question, which is — from Caitlyn and my experience, anytime you try to do something ambitious, there's always some sort of dark side to how the implementation goes. It can't all be roses. So what are some of the gotchas in this process, what are some of the frustrations that you've learned along the way?
Matthew Rascoff: I would say, you know, we haven't solved Baumol's cost disease. It's like, there are fundamental questions about what is the ratio of a teaching fellow to students. Putting a learning experience online does yield some productivity benefits, but it doesn't solve Baumol's disease. There's still some kind of ratio. Maybe the combination of asynchronous learning and synchronous learning shifts it a little bit. I think we've staffed this program quite generously. We'll assign a teaching fellow to every high school that we work with, that might have 15 or 20 students. We don't see the economies of scale that you might get in a fully self-paced asynchronous environment. That's because I believe the efficient frontier, in terms of learning outcomes relative to investment, requires that we staff this program more generously than you might in a for-profit context. We want our students to be successful, we're willing to invest more financially because we pay our teaching fellows, to give students the best chance of success. That limits our ability to scale. We could do a program that was 10 times bigger if we didn't do 1-to-15 ratios but did 1-to-150 ratios. We've chosen not to do that. We've chosen to get to 1,500 students in three years rather than 15,000. And the way we're thinking about scale is to build a consortium of other universities that also want to do this model. Other institutions have made different bets on, like, what is the appropriate ratio to invest. I think that's an unsolved problem in learning science. In the world of preschool, the ratios are governed by states — they say, if you have three-year-old children, you have to have one teacher relative to every, you know, six or seven kids in the class. Higher ed doesn't have a regulatory system that says here's how much staffing you need to have for how many students, and I think we need more science on this. I think we need more investment. I would love to know, what is the efficient frontier, where are the 80/20 rules in terms of how much human versus technology do you need in order to get the best learning outcomes for students, and where do you see kind of diminishing returns? So right now this program is arguably subscale by design, because we want to figure out these important design challenges before we pour gasoline on it and put a lot of students at risk.
Caitlyn: How do you think about this program vis-à-vis Coursera, which is a Stanford disruptive innovation started by Stanford professors, that is for-profit, now publicly traded? Like, did you take the lessons from MOOCs and their low completion rates but still wide aperture, and learn from it? Do you think of it as the next extension, with maybe more impact?
Matthew Rascoff: I think there's a kind of calibration process that we're doing. So, like, to me, Coursera — at least the MOOC model of Coursera, which they've kind of moved away from, but the original MOOC model of Coursera — was very focused on asynchronous, self-paced learning. Originally there were cohorts in the first couple of years, then they moved away from that completely, so you didn't have a learning community that you were part of, you're very much on your own. But the benefit was massive scale — they've got 130 million users at this point, and that's very impressive. The other, almost opposite, approach that we pursue in face-to-face learning is tiny seminars that are very boutique. I taught a first-quarter class at Stanford last fall which was unbelievably rewarding, amazing, but it was two instructors and 13 students. So it's the opposite of a 1-to-130-million ratio. Basically what I'm arguing for is, let's find the middle grounds that achieve some kind of balance between what we know about students' need for connectedness, for community, for social relations, for accountability — all that happens in models that don't scale to the millions. We need to really understand, like, what are the best middle grounds in between those two extremes, where are those trade-offs, and where do you get the most ROI in terms of learning?
Caitlyn: Totally. I think that's a wonderful framing for thinking about it. I think community is one of the most loosely thrown-around words when it comes to a lot of digital experiences, and it's nice to hear you thinking about that so intentionally. And I'm wondering what you think for that high school age particularly. I think there's a lot of evidence right now in terms of the time they're spending online and what that's doing from a mental health standpoint. So how do you think about the community formation, how the depth of connectedness becomes an asset and doesn't contribute to a growing problem for this age group?
Matthew Rascoff: Yeah, I share that concern. I would say, in our high school programs, it's more of a — they're with other students in their school during the school day, during the school year. There's a teacher who's there with them physically, so there is a face-to-face component. Students don't experience this as fully online. Our teaching fellows are remote, but there's also a local community for them in their school. High school students don't arrive in fully online. There is this program called Apex that tried to do online AP courses in the 2000s. They're the most motivated students — there's no AP Chemistry offered in my school, so I'll do it online, sounds like a great idea. The brightest, most motivated students didn't finish the course, they didn't go on to take the AP exam. It turns out, like, high schoolers need to have community, they're highly social — that's developmentally what they need the most. So I like that the Equity Lab has this kind of hybrid model. So that's the hybrid trade-off this model has. And it's another non-dogmatic approach. Most people who build online programs say it's got to be online, they're kind of hell-bent on that, and, like, that to me is like a hammer in search of nails. Like, we're hell-bent on, like, good learning experiences and opportunities and outcomes for low-income students — whatever model it takes, we'll do this.
Caitlyn: It's awesome to hear you say that, and Aran's trying to hold in his excitement, because Aran is an entrepreneur in residence at Avalanche building the high school of the future, and we've talked a lot about, do you do it online, do you do it offline, and our Aran is very adamant that it's mostly an in-person experience and the social elements of high school cannot be abstracted away or even put online without massive detriments.
Matthew Rascoff: Yes.
Caitlyn: We could talk about what you're doing here for hours, because it is really fascinating, but I want to shift gears here, because part of the benefit of getting experts like you that are so deeply building the future is people want to know, what is the future. So from your standpoint, when you think about the future of education, what do you see as the future that may not be so obvious to other people?
Matthew Rascoff: I don't know if this is sufficiently contrarian, but to me, learning is a human pursuit. It's never going to be entirely mediated by technology. What's kind of obvious to me, but maybe not obvious in Silicon Valley, is that we need new combinations of humans and technology in order to build for the future — something that feels humane, something that feels warm, something that feels connected, something that achieves, like, those goals of education, which are also valid goals. Just skills is a pale imitation of what you get from great learning experiences. If it were just about that, I'd be more concerned about disruption — in fact, it would have already happened. YouTube would have done it. YouTube's a much better way to learn how to make sourdough than going to a workshop. But it turns out bread baking, for me, goes beyond just the skill of how to put it together. It's also about doing it with my children, learning with them, and making something beautiful that smells good. And I think we need to massively rethink the way we use people and technology in tandem. We know the fundamentals don't change, and we need to design for those — building stronger connections, helping people feel like they're part of something bigger, helping them make progress that is meaningful for them, helping them figure out what is meaning, how do they create it, how do they construct knowledge themselves. There's this wonderful biography of Maria Montessori that I recommend. It's called The Child Is the Teacher. It's very much about Maria Montessori as an educator and as an entrepreneur. Little known fact, she was an edtech entrepreneur, but the tech that she built were manipulatives. She built these tools that were meant to be used not just in the classrooms that she ran, but she sold them to other Montessori movement schools around the world she helped build. And then she created a market, she created demand for these tools, and in Italy she had manufacturing partnerships that created these tools that she would then ship out. They were very expensive, very high-quality, beautifully built. So, to me, Montessori — we think of this as very progressive, human, connected, playful learning, right, it happens in preschools — even in that case, she was thinking about combinations of technology and people in tandem. It wasn't an anti-technology vision in her case, it was hardware, it wasn't software, right? That was a true educational visionary. You need edtech entrepreneurs.
Aran: I'm a Montessori kid, and my kids have gone to Montessori. I used to work for a private equity fund that was owning and operating K-12 schools; in the process, bought a couple Montessori schools and got to know the founders intimately. The more I've learned about Maria Montessori and her vision, the more I've been deeply attracted to what she stood for as an educator and how she approached things. Misinformation, miseducation, like cultish ideas around Montessori, even within, like, the Montessori community — it's such a fascinating thing that within a hundred years of this person's existence, it's like all of what she stood for has been abstracted away in all these different ways, and yet they're still core to what she's done.
Matthew Rascoff: That's partly why I recommend this biography, because it's written by a non-Montessorian, somebody outside the movement, who wrote a critical external biography of her, but they still got access to the archives. So I think focusing on her life and what she did is a way to cut through some of those accretions. But I also think what you described is endemic in education. It has to do with our failure to translate research into practice effectively. We would never accept a biotech translational system where the original molecule that was built in the lab was massively diluted when it got injected in people's arms, it didn't actually have any efficacy. Like, that's basically what we do in education. We take this germ of an idea and twist it so completely and dilute it so completely, it's no longer effective, it's lost its original meaning. To me, it's a much bigger challenge, not just in the Montessori area — the learning science rarely makes its way into teaching practice, textbooks, curriculum. There's this huge translational gap.
Caitlyn: That was a really good tangent and a really important point. I'm glad that you mentioned Maria Montessori's book, because our other question that we ask all of our guests is, what's a book that you would recommend to your younger self, for our audience?
Matthew Rascoff: I wasn't prepared for this one. I — a little bit. I would say the essays of George Orwell. Reading that was the critical kind of discovery for me in my own education, and I think they're even more relevant now than they ever have been. He was writing, you know, in the 1930s, and trying to navigate a middle ground between communism and fascism, looking for kind of democratic futures in the United Kingdom that seemed to be fading, and trying to build an intellectual community around ideas that were massively under threat. He's also one of the most beautiful English-language stylists, I think — totally. So, like, Shooting an Elephant, Politics and the English Language, those essays, I think, really shaped how I think about the value of democracy, free expression, and the role of writing in that process, the power that ideas can have and the consequences that they could have, the negative consequences but also, you know, the opportunity that we have to share and to publish. So that's definitely my intellectual role model. I wish I had discovered him earlier, because his writing is so lucid — you can read it as a teenager, and then you can come back to it in your 20s and your 30s. Now I'm in my 40s and it feels, like I said, like, super relevant in our political climate. So that's my recommendation.
Caitlyn: Thank you. That's nice. The slow golf clap for that — that was an excellent choice. Thank you so much, this was an amazing discussion, really appreciate it. Thank you so much for your time and being so open and honest with us. It was a lot of fun.