Matthew Rascoff, Vice Provost of Digital Education at Stanford University: EdTech Bright Spots for Collaborative Learning

Van Ton-Quinlivan: Welcome to WorkforceRx with Futuro Health, where future focused leaders in education, workforce development, and healthcare explore new innovations and approaches. I’m your host, Van Ton-Quinlivan, CEO of Futuro Health.

One of the most dramatic changes caused by the COVID pandemic was the shift to online learning at all levels of education. While it will be years before researchers will be able to render a verdict on how this impacted student performance, it is already clear that in higher education, post-pandemic use of online learning and the number of students and instructors who see it positively have all increased. Research shows that professors and students also expect more use of digital course material and technology in the classroom going forward.

With us today to explore the current and future state of online learning and related topics is Matthew Rascoff, vice provost of Digital Education at Stanford University. Previously, he led the Learning Innovation Team at Duke University and was vice president and founder of the Office of Learning Technology and Innovation for the University of North Carolina System. Early in his career, he worked in education technology at several organizations, including Google.

Thanks so much for joining us today, Matthew.

Matthew Rascoff: Thank you so much for having me. It’s a pleasure to be with you, Van, and I so admire your scholarship and your contributions as an alumna of Stanford University. I’m very glad to be with you and your audience today.

Van: Go Cardinals! Well, I thought we could use some definitions to set the table for those in our audience who may not be familiar with higher education lingo. So, Matthew, when it comes to the use of technology, can you explain the difference between hybrid learning and blended learning?

Matthew: Sure, and I can use that also as a little bit of framing for why we call what we do here digital education rather than online education, and that is because we want to be inclusive of various modalities that mix together different approaches of in-person and online and of synchronous and asynchronous models. I think that is part of the legacy of the pandemic, that there is a plurality of approaches that are now available to instructors. We need to support those that are most effective for our learners.

I think it’s most important to just start with the needs of our learners and work our way backwards to the modality that will meet those needs most effectively. That idea is sometimes called backwards design…rather than entering into the conversation ideologically saying, “I’m hell bent on a fully online model,” and losing sight of what your learners might actually need.

I would describe blended as one of the models of hybrid learning, and hybrids to me are about mixing together different modalities in order to meet the needs of your learners most effectively. In our approach, we will often use combinations of synchronous and asynchronous. So, everything is distributed and distant, but those combinations I think are still a frontier of innovation and I think there are many lessons still to be learned about what is the most effective way we can support different kinds of learners, different kinds of learning and different disciplines at different developmental stages, in different school or non-school contexts. There’s so many different possibilities for these technologies that have been unexplored so far.

Van: How do you think the backwards design experiments or approach would differ between a student population like us, for example, versus your children who are still young, right?

Matthew: It is very different and I think part of the friction that you saw in the pandemic is that we were applying technologies that had been built, like Zoom, for business meeting collaboration and offering them up to kindergartners like my son. I remember very well the moment when he was presented with a dialogue box on a Zoom screen and he didn’t know how to read yet. So, “OK” and “Cancel” — which would have been trivial for somebody to click through — he couldn’t process that and he got totally held up by that. So, that was kind of the moment when I realized that Zoom kindergarten was not a thing… (laughs)

Van: (laughs)

Matthew: …and something that would have been totally straightforward in one context was totally inappropriate in another. I think the frustration that parents felt, including myself, was about the misapplication of one technology in a learning context where that was not the needed thing.

Backwards design has to start with an empathic process of like, who are my students? What do they need? Where are they? Not just in terms of technical sophistication, like I was talking about, but in terms of content level. That is a big pivot, especially in higher education, where we often take a much more “supply side” view of the world, not a “demand side” view. The supply side says, this is what it means to be an expert. This is the knowledge that has been developed over centuries. This is what defines this discipline. You need to have this.  Rather, we should start with where my students are and what will help them get to the next level. That requires a reframing and a different focus on this kind of demand side rather than the supply side of the equation.

Van: So, let me probe this a little bit. Your kindergartner — who had to do Zoom during the pandemic — as he progresses and gets to college, what do you think his learning experience will be like then, do you think?

Matthew: I mean, he’s seven now, so I’ve got a little bit of time and technology has a little bit of time. That’s eleven years to figure that out. That’s several generations of hype cycles, ups and downs, and this is an incredible moment of innovation. I think part of it is driven by AI innovations. Part of it is driven by the dissatisfaction that people felt with the technology infrastructure that was available during the pandemic. There’s a huge wave of entrepreneurship that’s happening right now with people who experienced that, were unhappy with that, and they think they can build something better and they’re probably right. Some of them are gonna build the next amazing learning platforms, and I often see them in my classrooms and I often support them. They’re my students here in the Graduate School of Business at Stanford.

They have really amazing ideas for what that future is gonna look like and I hope it is a more inclusive technology environment. I think it focuses on our principles. I think it focuses on accessibility and is available to everyone. I think it’s designed to serve learners who have been underserved in the past…not just making the educated still more educated, but refocuses and reframes learning opportunities on those who might not have the greatest ability to pay, but who are entering into the learning ecosystem from the lower rungs of the ladder and want to climb it. What are we doing to make sure that ladder is climbable and to make sure it’s accessible?

I think even research universities like Stanford, which have typically focused on the upper echelons of that ladder, we need to be focused on all levels of it if we want an inclusive educational system. The technology can either be a force for inequality or it can be a force for opportunity and democratizing knowledge, and I think unless we are intentional about designing learning technologies that support our values and are filtered through the values of non-profit organizations and academic institutions, we will end up with ed tech that makes many of our problems worse and does not actually address the root issues in our society.

Van: Matthew, have you already seen some promising digital learning texts that advance some of these concepts that you’ve talked about?

Matthew: Absolutely. Just this morning, I was meeting with an incredible entrepreneur, Joel Podolny, who’s the founder of an ed tech company called Honor Education that I greatly admire. He’s the former head of Apple University — the chief learning officer who was hired by Steve Jobs to help sustain Apple’s culture of learning and pass it on to the next generation of Apple employees. Now he’s got an ed tech company that is a modern platform for collaborative learning in higher education. It’s built around this social community of people who are reading texts, watching videos and annotating them, and sharing their commentary with one another and with instructors, and building a seminar-style experience in a totally asynchronous learning model.

I’m inspired by that model, and I think there’s huge potential, for example, in bringing modern consumer-level UI, UX — the expectations that we have from phones, from devices — into the world of ed tech, which has typically been a step or a half step behind those modern designs and interfaces.

Van: So, you think a sort of a critical component is fostering that sense of community while learning?

Matthew: Absolutely, and I think that’s what many faculty who are critical of online learning miss the most. That’s something that we’ve learned during the pandemic. We documented this ourselves at Stanford. We published a report on lessons learned during the pandemic — it’s available at pandemiced.stanford.edu — and that sense of isolation and the loss of community was one of the major challenges that students and faculty faced. I think it’s something we need to double down on now that we understand how valuable that was, because we lost it temporarily. Let’s invest much more in that idea of learning communities and build technology that supports it and enhances it, rather than isolates people and takes them away from one another and focuses their attention on the screen, rather than on their colleagues and their fellow learners.

Van: What would be an example of how I could be experiencing community in a digital format?

Matthew: There’s a wave of new online providers that are investing in this idea of cohort-based learning, and I think this is a very important trend that we should follow. One of the leaders in this space is called Maven, and they’ve pioneered this model that creates synchronous small group communities around a topic and kind of recreates the seminar discussion that you might expect in a great classroom. It’s mixing synchronous and asynchronous learning, I think, in ways that are very compelling, and reinvesting in the idea of synchronous, but not exclusively that.  It’s focused on what we can learn from one another just as much as what we can learn from an expert.

There’s another project that I love, it’s called the Catherine Project. It comes out of St. John’s College, and this real educational pioneer in the liberal arts, Zena Hitz, has built this fully free, online model of sharing core texts in the Eastern and Western canons and reading the classics together in small group discussions. There are a few different formats that they’ve pioneered, but all of them depend on reading some classical texts and then creating seminar-style discussions around them with supports, with rubrics, with training for the leaders…but it’s all free and it’s led by volunteers. It kind of recreates this model of a tutorial that you might get at St. John’s College, but democratizes it for people who are not matriculated students there.

So, there’s a whole wave of innovation that’s happening in this space from Maven’s focus on professional skills to Catherine Project’s focus on classical liberal arts and everything in between.

Van: These are great examples. Matthew, what about this issue of all the learning loss and that is about to hit the higher education system?

Matthew: Yeah, there’s a lot of research on this and there’s more coming. The research that I’ve seen is built around state exams that are administered every year and you can see the drop-offs in reading and math and they are very concerning. School districts are investing in interventions to try to mitigate some of these gaps. I don’t think we should pretend it doesn’t exist. I mean, I think there is kind of a movement against the very idea of learning loss and there’s some denial about it to say, “Oh, we couldn’t have done any better.” I think that is a mistake. I think we need to confront this and I think we need to bring all of the science of learning to bear on this generational challenge.

What if we had a grand challenge in education that was focused on closing this gap and every other gap that pre-existed the pandemic as well? Like, maybe that could motivate a wave of researchers and entrepreneurs who want to translate that research into practice.

I would say tutoring is having a moment right now and that is a strategy that many school districts are adopting. There’s a community centered around a faculty member at Stanford, Susanna Loeb, that’s called the National Student Support Accelerator and they are trying to build kind of a community of practice with researchers and practitioners at the table on high impact tutoring…meaning using what we know about dosage and fidelity of implementation to scale up tutoring that will actually be effective for students who are most in need, and in many cases for all students as well. The state of Tennessee has now adopted this as a model and has built it into the state funding formula, for example.

I think there’s a lot of opportunity in other states and also at the school district level to invest in high impact practices that we know will be able to close some of those gaps for the next generation of students.  I encourage your listeners to check this out because even though Susanna has focused on the K-12 system, I think this has broad implications. If you care about learning anywhere, we should be applying what we know about how to make tutoring more effective, and there’s a huge role for technology in this. Some of it is face-to-face, some of it is online, some of the online has a data-driven feedback component that’s helping tutors get better at their jobs. Some of it is machine-driven tutoring, but a lot of it is machine-enabled humans who have more feedback to allow them to improve their practice. So, that’s gonna be a huge role for the AI in giving instructional coaches to these tutors that allows them to be supported, that trains them, that ensures implementation with fidelity at scale.

Van: Sal Khan of Khan Academy is prototyping an AI tutoring tool called Khanmigo. So, lots of experiments going on. Could be good.

Matthew: I’m very excited about Khanmigo. He did a few things with it that I think are extremely compelling. One of them is that it’s constrained and it will only stay on topic with students. I think that gives educators more confidence to assign a tool like that rather than a general-purpose AI which hallucinates and can easily go off topic.

The other thing that Khanmigo did — which I think is very elegant that I hope others replicate — is there’s a teacher interface so the instructor can see what the students are doing with it. It empowers the instructor rather than disintermediating them and it gives them data about what might my students be struggling with. In other words, it’s a formative assessment tool for the instructor, even as it’s a tutoring and debating and learning play tool for the students. And that dual-purpose concept, I think, is a very, very compelling design for AIs in education.

Van: Is there any situation where you would recommend going back to in-person learning without digital tools?

Matthew: Of course. We all need that in our lives. I mean, I myself did a retreat this past weekend in the Santa Cruz mountains, in Aptos, where there was no connectivity and it was an incredible relief to be away for the weekend and not have emails and texts and notifications streaming through. Honestly, those experiences have become precious, I think, to many of us, and especially if you have small kids.

I want to model for them the ability to put down the device and focus my attention on them and on other people and on learning. Print does that and screens don’t, especially for young kids. They have a kind of magnetic force to them that is powerful and dangerous and we need to use that power for good and understand its limits and make sure that we respect them. So, absolutely there’s a role for that.

Van: Right. Now that you’ve survived the device-free retreat, next is the silence retreat for you.

Matthew: Yeah, I mean, it’s not a surprise that these ideas come out of the technology culture in Silicon Valley. I see those who are most steeped in the technology sometimes are the ones who are most skeptical. When they see what its impact might be on their own children, they are sometimes wary of it. I think there’s movements that came out of Google a few years ago that were about kind of algorithmic power, but I think we need to be thoughtful. It’s not about reactionary rejection of screens or screen time — like, that in itself is not hazardous — but we need to be thoughtful about how we bring these technologies into our lives. I’m arguing for moderation, not for wholesale adoption and not for a reactionary kind of Luddite rejection.

Van: Right. Well, Matthew, you founded and you oversee Stanford Digital Education, which was formed just a couple of years ago. What is the mission of your team?

Matthew: The mission of our team is to reimagine digital learning in the context of a great research university like Stanford as a force for our own academic mission. It’s a reclaiming of the technology in concert with people, in support of people to enhance what we do in research, in teaching and in service.

Our pilot demonstration project is a program that offers Stanford courses for credit to Title I high schools across the country. It’s a model called dual enrollment, meaning they get both high school and college credit at the same time. The program is designed as kind of a triple boost for low income, talented high school students: they get an academic boost from a course that’s led by a faculty member, but also supported by a local teacher in their school as well as a cadre of teaching fellows who are Stanford students and alums who run weekly discussion sections with them; they get a psychological boost that comes from having taken and succeeded in a Stanford course with a transcript at the end and they can take that transcript to anywhere else and transfer that credit in; and they get a social capital boost.

The social capital boost comes from the networks that they get to build up amongst their fellow students, but especially with our teaching fellows and with our faculty. We know from research that those networks are really critical. It’s part of what you get at a great institution like this one. Online learning that just focuses on skills and ignores social capital is a highly impoverished shadow of what you get at Stanford. We want to recreate an authentic experience of Stanford that includes both the academics, the psychological insights of mindset and confidence and belonging, as well as the social capital boost that you get from participating in a community like this one.

Van: So, play it forward, Matthew. What happens after the pilot or at the end of the pilot? What are you hoping to do?

Matthew: This is no longer a pilot.

Van: Ooh.

Matthew: We’ve now done eight of these courses across the whole university and we are committed to this model of “pathways digital education” — using the tools to reach upstream into some precursor system where there may be some inequities that are baked in. There are deep inequities that are baked into the K-12 system and we can address those inequities at the root and use the technology, as I was saying before, not just to educate the already educated, but to find those who have talent that might be unrecognized and give them that support, give them those three boosts that set them on the path to well-matched college opportunities.

We have some research now that shows that we are increasing the likelihood that students who participate in this program will move out of state for college. We’re increasing the likelihood that they’ll attend a four-year rather than two-year institution. I know you’re a community college person. I love community colleges, but if you have the ability to get into Berkeley, you should go to Berkeley and not to your community college. That’s undermatching. Undermatching is this phenomenon of the most talented low-income students never even applying to the selective colleges that would admit them and give them more generous financial aid than the undermatched colleges that they attend. That to me is a critical issue in the meritocracy that we’re going after with this program.

I think you could imagine different pathways programs playing out in different contexts around a university like Stanford, where our graduate school might have a similar pathways focus on coming into the graduate studies. I think that idea has broad implications for how we think about technology in the service of a residential learning experience that hybridizes — to your first point at the beginning — and does it in a way that is focused on the funnel, that’s focused on the upstream, that gives people an opportunity at the very beginning and gives them the support along the way to follow through.

Van: For so many of the STEM careers, for example, you need advanced degrees or even just the completion of the bachelor’s and the gating to get onto that pathway tends to be Algebra 2. So, when you talk about these courses for credit in dual enrollment, are you targeting those types of courses where students basically lose momentum or they can’t get the momentum for some of these academic trajectories if they don’t get those courses?

Matthew: Yeah, we are very much focused on key skills. We have not done an algebra course yet. I would love to do something in mathematics. We have not figured that one out yet, but there’s great interest and there’s a curriculum redesign effort that’s happening in the engineering school’s math curriculum that I’m excited about here.

But we have focused on other core skills, and I would say the paradigmatic example of that so far has been our writing class. It’s another key missing skill that many low-income students don’t get enough feedback on to get better as writers and then they show up for college and they see that they’re asked to write a term paper in their first semester. They’re just not prepared, having never done that before, having never scaffolded the skills to go from beyond the five-paragraph essay, which is where a lot of people kind of stop out of the writing skills. The five-paragraph essay does not set you up for college success, not even close. There’s a huge gap between that and what you’re expected to do in the freshman curriculum at any college. So, our course was designed to prepare students for academic writing in college and to close that gap for low-income high school students who just typically don’t get the support that they need to become effective writers.

Van: Are you skimming from a certain population or is it open admission to anyone who’s got the enthusiasm and drive?

Matthew: Our model is that we work with schools and a school network called the National Education Equity Lab and they help us find these students. Students are nominated by their schools, their teachers and principals. The Equity Lab runs the network of Title I schools, so they recruit the schools and then they bring those students to us. We’re not involved in the selection process for students.

Van: I see. Well, that’s good work, good work. So, Matthew, I’m gonna bring you to a different population. The majority of Americans don’t have college degrees and many are finding themselves with limited work opportunities. What can Stanford and other higher ed institutions do to help here?

Matthew: It’s a wonderful question. I think it’s very apropos of ideas that are gaining currency here, and there’s a group of colleagues within Stanford who are actively trying to figure this out. I would not claim to have the answer to that. I think that we’re making progress on different approaches that might get traction and I’ll give you a couple of examples that I think are promising. But, I’m also very interested in your perspective on this one because it’s an audience that we have not served effectively in the past.

I am very concerned. There are new numbers out that show that it’s now more than 40 million people who have some college, but no degree. That to me is a travesty that the higher education system has basically set up people for failure where they have debt but they don’t have the ability to pay it back. I think we, collectively as higher education professionals, bear some responsibility for designing an inclusive system that supports people’s wellbeing, that supports their opportunity.

There’s new data from Raj Chetty that shows that on average, people born after 1984 will be the first American generation to not out-earn their parents. The basic principle of the American dream that you will, in every generation, do better than your parents did has ended with those born in 1984 and thereafter. So, it’s no surprise to me that there’s frustration with higher education and that there’s concern about the debt levels that we have put onto students and the inability of those students to see the return on their investment.

I think we need more flexible models of higher education. I think traditional providers need to figure out what that might mean for themselves, and it can’t just be the skills sector, the bootcamp sector, that is invested in that. I think we need to figure out how we hybridize our traditional offerings with more flexible, shorter-term models.

One project that I’m excited to announce is a collaboration between the Bay Area Community College Consortium and Google and Stanford, where we are providing professional development for faculty in Bay Area community colleges to improve their data science instructional practices. Data science is a very attractive, emerging field. You actually don’t need a bachelor’s degree to do data science in many contexts because the tech industry has been more inclusive in its hiring practices and is willing to look at people with different backgrounds. So, if you’ve got the skills to be an effective data scientist and you’ve got a credential that demonstrates those skills, you can actually get your foot in the door. And then once you’ve got a job at a tech company, they often have a tuition reimbursement program that will accelerate your path to the bachelor’s degree.

I don’t see it as an alternative to traditional degrees. It’s a foot in the door that can be that first rung on the ladder that I was talking about before. But how do we support student success in those models? And how do we integrate these kind of industry certificates into the traditional system? Those are the questions we’re asking in collaboration with these Bay Area community colleges and with support from Google. The “Grow with Google” team, I think, has been at the forefront of this kind of innovation in credentials that are tied to skills that Google knows that it needs and content that they built, but they don’t have the ability to train faculty in community colleges. We do. We actually know a lot about faculty development and we have amazing resources within Stanford.

So, we’re inviting in our colleagues from the community college sector to participate in some of the professional development opportunities around data science and around the networks that are building at Stanford in support of the growth of data science. We have new majors in that space, we have a new building, we have an incredible network called Women in Data Science. There’s a whole inclusive community that is invested in making the data science skills more widely available and democratizing them.  I think doing it through community colleges in partnership with our regional allies is a way that Stanford can enter into a space that is unfamiliar to us where we know we want to make an impact.

Van:  Of course, I have to ask you the question of generative AI’s impact on teaching and learning at the college level. What’s your prediction on where this will be maybe in two to five years?

Matthew: I’m an optimist, I would say. I mean, Kahnmigo is a good example of what it might look like to reimagine these tools in an educational context and give educators some confidence that they can assign them. I think what we’ve seen is something of a kind of moral panic around it and a backlash. That was the initial reaction that you got from many educators. Like, the New York City Public School System originally banned ChatGPT. Just this week, they said, actually, no, we’re rolling back the ban and we’re gonna integrate it, but we’re gonna do it in a way with support, with professional development and think about how can we harness this technology rather than pretending that it doesn’t exist. That is a natural process in education.

I think we’ll see roles for this in more vertical support. Plain vanilla GPT tries to give you the answer. Good tutors don’t give you the answer. Good tutors hold back the answer and help you figure out what are the steps that you need to take to overcome whatever’s holding you back from being able to do this on your own, and they don’t give you any more than you need. So that is a role, I would say, for an education-specific approach to generative AI that is almost the opposite of what a search engine company or a general-purpose AI company might give you. It’s intentionally obfuscating the answer rather than giving it to you as efficiently as possible. That’s just one small example of kind of a re-imagination of this technology in the context of tutoring that might be necessary in order to make it effective for learning.

Van: Well, I bet our listeners have been extraordinarily impressed with the range and depth of your knowledge. I mean, every time I speak to you, the richness of solutions that you see within your radar is just so impressive.

Alright, so let me close by asking you this, Matthew. What excites you about what is to come that we haven’t talked about yet?

Matthew: I would say my students and what they’re interested in, to me, are the bellwether for the future. They are making very careful opportunity cost decisions about how they spend their time and what jobs are attractive for them, what they think is worthwhile. The focus and discipline that they bring to that, to me, is a kind of heuristic of what is worthwhile. What’s worth my time, and where should I be focused on for the future? In some cases, they’re designing the companies that will become part of the firmament of the future. That’s why I teach, because I get that window into what’s coming and I get a huge morale boost from it.

I would say my students are very drawn to your audiences, Van. They care deeply about the underserved learners who are adults, who’ve been left behind by the traditional higher education system. I think there’s been kind of a move. Many of them are former K-12 reformers and have evolved to be more focused on adult learners because they see that, you know, the college-for-all idea did not play out as they hoped and it left many people cold and left many of them economically behind and now we are dealing with the consequences of that.

I’m thinking of entrepreneurs like Nitzan Pelman from Climb Hire — somebody I greatly respect and somebody my students are drawn to — who’s building a social capital-driven nonprofit bootcamp that helps people gain critical, job-ready skills in Salesforce administration. This is something that not just tech companies, but most companies need, and they need people to manage the CRM. It’s a fantastic entry-level job. Nitzan’s program prepares you for that, and it plugs you into the networks that actually help you get the job, not just learn the skills. She is invested in applying research into the design of this program, and thinking about what a program might look like that doesn’t just center skills: it’s skills plus networks that set people up for career success.

My students get that. If you’re at the GSB, you don’t need to explain the value of networks. That’s obvious. But who is thinking about how to democratize those? Who is making those more widely available? It’s social entrepreneurs like her.

Van: Well, we want you to keep spinning up those talents who can help solve these issues of education, learning, and social inequity. It’s been such a pleasure, Matthew, to have you here today.

Matthew: Thank you so much for having me, and I am excited to listen to more episodes of this and to read more from you. I’ve so admired your career and your contributions to the Stanford community, and I’m really grateful for your friendship, Van.

Van: Likewise, Matthew. I’m Van Ton-Quinlivan with Futuro Health. Thanks for checking out this episode of WorkforceRx. I hope you will join us again as we continue to explore how to create a future-focused workforce in America. Thank you.