The future of education

0:00Susanna Loeb:This is Stanford Engineering's The Future of Everything, and I'm your host, Russ Altman. Since we started this show eight years ago, it's become an archive of amazing and impactful work by my Stanford colleagues. Research is not something that just happens in the lab, and as you'll hear on this show, the research at Stanford can impact areas like health, technology, law, and business, and many other topics that can affect everyday life. We hope you'll tune in to learn more about how research has the potential to help your life and to help the lives of people you care about in your family and your community.

0:32I tend to think about how we can get students the experiences that really help them develop the capacities that they need to thrive. And we know a lot about that already, but we're not doing the greatest job of actually getting them those experiences. So when I think about these new tools, I think, how can we use them so that we can make our education system better so that we can provide students these experiences in ways that we haven't done before.

1:08Susanna Loeb:This is Stanford Engineering's The Future of Everything, and I'm your host, Russ Altman. Thanks for listening to the show. If you're enjoying it, please consider rating and reviewing it. A rating is a number. We like to get a five if we deserve it. And a review is some comments about what you think of the show that'll help us and others understand what it's all about. Today, Susanna Loeb will tell us that for education, it's the experience of the learner that matters. So when you're evaluating new educational opportunities, new classroom styles, new teachers, and especially new technology, it's all about the experience.

1:39Susanna Loeb:It's the future of education. Today, we're continuing our feature, The Future in a Minute. After my interview with Susanna, I'll ask her some rapid-fire questions. She'll give us some rapid-fire answers, and that'll be The Future in a Minute. Before we get started, please remember to rate and review the show. We're very grateful for you listening, and we'd love to hear your thoughts.

2:07Susanna Loeb:There's a field of research called education research. In this field, researchers try to understand how to do education best. How do you set up the schools? How do you choose the teachers? How do you set the curriculum? How do you pay for everything? How do you govern it? And of course, how do you use technology? The questions here are quite relevant because the education of our children and our own continuing education as adults is absolutely critical for maintaining a population and a workforce that's ready for anything. Well, Susanna Loeb is a professor of education and a senior fellow at the Stanford Institute for Economic Policy Research.

2:44Susanna Loeb:She's going to tell us about how education research is grappling with the issues that came up in the pandemic, and also with the explosive issue of the use of AI in schools by both teachers and learners. Susanna, what drew you to a career in education research? Oh, that's such a good question and actually kind of unexpected. I actually started thinking that I would do engineering and work in the environmental field. And then I, working in there for a while, made me realize that even in that field, education was fundamental for the kinds of changes that we wanted to see in the world. And I just deeply believe that it's the basis of the society we live in.

3:30And that's really what drew me to education.

3:32Susanna Loeb:Did you also have a desire to be an educator? Because there's a little bit of a difference, right, of studying education versus being out in the field. And of course, as a professor, you do both. But was this a dual love? Or are you more interested in like the theory of education than the practice? How do you grapple with all those issues? Yeah, I do really love both. I think why I've loved my job so much is that I get to see the joy of the actual education, the relationships that develop, the excitement of learning new things. So I really like the human side of it, but I'm also drawn to trying to figure out how to get students the very best and people the very best experiences that they can have so that they learn and are so excited about the world around them.

4:25And so I really like the scale of the effect that we're able to have as researchers and as working at kind of the systems level, but I'm always motivated by the kind of human interactions that I see in the work.

4:41Susanna Loeb:Great, great. So for people who don't think about education research all the time, can you kind of, as a preface to the rest of our conversation, can you give us a sense of like, how do education researchers approach their jobs? And like, what are the big contemporary issues in education that are like getting your attention and others? Yeah, so of course, there were huge range of education researchers, some, some researchers work on really the neuroscience science of learning. So how does the brain respond to different inputs? What is the very best way that a student can learn how to read? So there are those kinds of researchers.

5:20But then there are researchers like the research that I do that's more like, how do we create education systems that are beneficial for students, that engage and excite them and also help them develop the skills that they need to thrive. So we think, I think very much about how do you recruit and retain excellent teachers? How do you set up classrooms so that students get the individual attention that they need, that they get those interaction opportunities with other students, all of those kinds of things. And how can you do that, not just in one school, but across an entire nation or, you know, globally so that we all have these kinds of opportunities.

6:01And some of the pressing questions right now, I mean, there are these great inequalities in access to the kinds of experiences that students need to thrive. So I think just a question of how you scale quality, how do you make sure everybody gets that is important. And then I think a little bit about what we're going to talk about today, which is new technologies and both the opportunities and in some ways the threats that they may pose for getting students these kinds of opportunities. Yeah.

6:35Susanna Loeb:I had the pleasure of looking at some of your previous work just in preparing for our conversation. It's really impressive that you look at very logistically important things like how is the school financed? How is the school run? How do people make the career choice? And so it's been very broad and very real. I can imagine that teachers in the trenches would look at your research program and say, yeah, that's what needs to be studied because that's what affects their everyday life and their ability to do what they're passionate about. So it was very impressive. OK, so we are definitely going to talk about technology.

7:08Susanna Loeb:And I but one way I want to get to it is I know you've done a lot of stuff on tutoring and effective tutoring. And from my understanding, it had a little bit or a lot to do with the pandemic as kind of a forcing function. So can you tell that story about how your recent work in tutoring was kind of catalyzed by the pandemic? Although, you know, I don't want to give the pandemic too much credit for anything. Yeah, no, I think that's just right. So the pandemic happened. And I think what really characterizes my research, as you can see, is I've worked on finance and teachers and school leadership and all sorts of different projects.

7:44But it's always been grounded in partnerships that I've had with districts around the country, with educators around the country. And from those kinds of partnerships, you could see immediately that this was different, that the pandemic really was different. Suddenly, you had huge numbers of students who were disconnected from school, who had lost some kinds of learning. They knew part of it, but not some of it because they had been out. And so you really had to think, what is the best way of addressing these kinds of needs? And so we went to the research literature. and it's really not surprising because of the kind of clear logic, but intensive tutoring turned out to be clearly the most effective approach for addressing these kinds of issues.

8:31So students who are disengaged, if they have an adult that they work with regularly at the right level, who teaches at the right level and also motivates and engages them, you can make dramatic changes. And the research was really clear on that. And so my work coming out of that was very much a really a question of how do you get this to as many kids as possible? And what was fun about it is that everybody knows that it's the right approach. So all teachers would be like, of course, this is what my students need. All parents felt it was just what they needed, what their children needed. But the struggle was that not everybody believed that they could actually do it.

9:13Is this something that we could really manage as a system at scale. And so we went out to study what kind of programs are effective in different places, and then when districts try to roll these out, try to build these, what are the barriers that get in the way, and how do they figure out how to get over them? And then what we could do is take these proof points of success and really bring them to other places around the country because we worked in so many places that we could really see variation. So this is what worked in a rural place, and here's what works in an urban place, or in the South where the education systems are set up really quite differently than they are in California.

9:59So in both places, we can see that. And so a lot of my work has been doing that, and what's been so exciting is that we really have seen access to this kind of relationship-based personalized instruction grow.

10:14Susanna Loeb:So I wanted to ask you, it is very attractive and it's very intuitive. But of course, the first thing you think of, especially at a systems level is, well, of course, we would all like to get a one-on-one tutor, but we just don't have the person power for the teachers. But one key thing that goes into that calculation, I would guess, is how much actual time does the tutor have to spend? So you could imagine that instead of having a teacher at a school give lectures to 20 students or 25 students or whatever it is, that they could go around the room and that short, brief, little bursts of interaction with the students as they walk around the classroom might be more effective than 20 minutes talking to everybody or some kind of mix.

10:58Susanna Loeb:So what is known about how high touch and how prolonged that tutoring relationship has to be? And will it scale? And do we have the workforce to do all of this? Yeah. So I think that that's it. That is a very good question. And it varies a little bit with age. But the the general findings, if you take all the hundred randomized control trials have been done, is that you need about three times a week. And it should be, you know, half an hour to 45 minutes of tutoring. but if you're a young... I don't mean to interrupt, but that is very reasonable. I can imagine that happening. It's not, we need seven-hour sessions every day.

11:42Susanna Loeb:It's not seven-hour sessions every day. But even there, for young kids, they have shorter attention spans so that you can get it to them that way. The other thing is that I think we do have a workforce that we hadn't leveraged before that now we're being able to. So, for example, work-study students in college have turned out to be wonderful tutors, teacher preparation programs they used to observe for their first time in schools. Instead, they can tutor and get that great experience with students. So there's lots of opportunities to get more adults into kids' lives. That really is exciting and it kind of makes sense.

12:20Susanna Loeb:And I know one of the things you're working on is scaling this. And this goes to a lot of your other previous research about like, how do you effectively disseminate a practice? I know that's been a big focus of yours. Like, okay, we finally learned something that seems to be true and useful. And now we want everybody to do it. So what is the way, what are the knobs you have to turn to try to get an entire nation to adopt a new educational practice that you're sure is a good way to go? Yeah. I mean, it's kind of tricky. So I think we have, I mean, the thing to know about our education system is it's hugely decentralized.

12:56So the biggest decision makers really are at the district level. They make a lot of decisions, but we have more than 12 ,000 districts in the country. And so that's a lot of people to reach. The way that we have tried to do it, nobody's going to pay attention really to a professor. You want to know from someone who's It's your job. Who's going to do it? So really our goal has been to study places that have the potential to do it well, see if they've done it well, and then to have those people bring it out to the world so they can be proof points. And those proof points are really powerful. I do think research on what typical barriers are to getting it done.

13:40and then really clear, we create a lot of materials for districts where we take the research and try to help them recognize where they might run into problems and how they might get over it. And so having those materials, the champions for it, and the proof points are really, really helpful.

14:00Susanna Loeb:Is every time you have to disseminate, do you have to build a dissemination plan and network fresh? Or is there some sort of existing infrastructure of communication that allows you to kind of use that each time there's something useful? There are networks out there, I guess is my answer to that, that we begin to make connections and partnerships with people who have influence. They can get this kind of thing out. The tricky thing is becoming a trusted source of information. And we work really hard to be a trusted source, to make sure that we're really thoughtful about variation that's out there, about the challenges people have in their job, and to provide supports that help them make the decisions that they want to make.

14:50Great.

14:50Susanna Loeb:So I do want to move to technology because many people listening might think, okay, this sounds like something that maybe technology could help with. So what is, I guess, what is your relationship to technology and how as an education researcher do you look at some of the obvious explosive things that have become available in the last couple of years? Yeah. So, you know, in working with all these districts on tutoring, we got a chance to talk with, we regularly get a chance to talk with lots of school leaders and district leaders. And there was just a real demand that developed for them to better understand what their options are, what their worries should be about the new technologies.

15:32So while we were doing all this tutoring, there was suddenly a demand to deal with AI in schools. And so we have something called the AI for Education Hub as part of the scale initiative that I run. And we're really interested in both doing the research to try to understand the best uses of these tools for students and then helping school leaders, state leaders make the best use of these kinds of tools.

16:00Susanna Loeb:Yeah. So maybe can you tell me, like, what's the landscape right now? It strikes me that if you're thinking about AI and, of course, chatbots and stuff like that, it's a little bit of like flying, building the plane while it's flying, because whether or not you recommend various technologies or recommend against them, we know that they're out there. So could you just, as we start this conversation, what's the landscape that you see of usage? And I'm really interested from the very beginning, like the kindergartners up through high school and college. What are you seeing as an educator? Yeah. So schools in the United, in the, in the U.S.

16:37in schools, most of the use is by teachers and administrators and not given directly to students. Certainly students are using Grammarly and some of these tools, but not, it's not really built in as much to their day-to-day lives. But for teachers, they are. They're these tools that are helping them write letters home in different languages. They're helping them create lesson plans that differentiate. All of those kinds of things are already really in the classroom in many, many places. Evidence about effectiveness is really low. We just did a review of the research. We have a repository at the hub for all research on AI in education since ChatGPT came out, basically, so the last few years.

17:29And only 20 of those studies really are believable causal effect studies of how these tools are affecting either students or teachers. And what you see is that they are time-saving for teachers. And so there's some potential there. But there are very few studies on K-12 students just because they don't use them very much. That said, there's a lot of use outside of school. so that we do have to think about changing homework assignments and things like that to meet them. But I think the big takeaway is that we are working in a space where we know very little, where things are happening really quickly, and that it's going to be important to set up ways of learning and adjusting quickly in real time because of these changes in a way that we've never had to do before.

18:25And so that is kind of our goal in what we're doing is to figure out how can we become a learning system and not let what's happened in social media and the negative effects that we've seen there be repeated with these AI tools for education.

18:42Susanna Loeb:This is The Future of Everything. I'm Russ Altman, and we'll have more with Susanna Loeb next.

19:01Susanna Loeb:Welcome back to The Future of Everything. I'm Russ Altman. I'm speaking with Susanna Loeb from Stanford University. In the last segment, we had an introduction to research and education, and we heard about how tutoring had a moment during the pandemic, which now has ongoing good effects on our understanding about how to best do tutoring. In this segment, I want to ask Susanna more about how AI is being implemented, how she thinks about the use of technologies like AI and some of the work she's doing with the AI companies to really understand how it can best be used to advance the education of both youth and adult learners.

19:36Susanna Loeb:So Susanna, as you approach AI, like what's the way you think about it with respect to its impacts on education? Because it's multifactorial. As you said, the teachers are using it. We know the students are increasingly thinking about it. How do you approach it? Yeah, so I tend to think about how we can get students the experiences that really help them develop the capacities that they need to thrive. And we know a lot about that already, but we're not doing the greatest job of actually getting them those experiences. So when I think about these new tools, I think, how can we use them so that we can make our education system better, so that we can provide students these experiences in ways that we haven't done before?

20:23So, for example, while we know that students learn much better if they're taught at the right level, so that we're actually teaching them right above what they already know, we have a very hard time doing that. And maybe AI can help us do that because it can differentiate. But similarly, we know that things like apprenticeships can be really motivating for students, but we often have a hard time making those happen well because there are lots of logistical factors involved. And we don't also do a very good job of measuring whether students are learning in those situations. So we can't scaffold them and push them forward in that learning.

21:03And so I think how can we change these kinds of opportunities for students by using these kind of new tools that have this great ability to do things like differentiate and deal with really complex organizational matters. Yeah.

21:23Susanna Loeb:And I could imagine if you're doing research on AI and education, like with chatbots, like chat GPT or something, and its competitors, one thing you could do is control trials. Like in medicine, we do this all the time. You know, there's an intervention group that gets it and another group that doesn't. And you could look at differences. But you can also, we do this in medicine as well, do what we call like observational things where you just let stuff happen. And you look at the people who did whatever and the people who did whatever else and compare them. what's happening in terms of trying to gather this information?

21:57Susanna Loeb:And is there appropriate support for doing these studies? Yeah. So in education, it's really hard to compare students who use a tool more to those who use it less because those students are so fundamentally different that the ones who use it more almost invariably would have done better anyway. So it is important to try to make a fair comparison, and randomization is the best way. Historically, what we've done is prepared far in advance for this kind of randomization and then followed students for many years after. We don't have that kind of time frame right now because the tools are changing so fast, which is both great because they could get better, but also scary because we want to know if something is going to have a negative effect.

22:48So I think the trick now is to treat them more as A-B tests still randomizing. So we still get around that problem of the students who take it up being really different, but we do it in a fast way. In order to do that, we really need to set ourselves up to do it. So it's important to do it kind of in partnership with both the industry who has these tools and the schools and districts who have the ability to work with students with the tools and then, you know, make the test, see if you can get it better, do a test again and keep that learning process going. Yeah.