Thriving from the start: Early childhood research

Phil Fisher (00:00):

You already have all of the superpowers necessary to build healthy brains and high-quality learning and to improve academic achievement.

Denise Pope (00:14):

Welcome to School's In, your go-to podcast for cutting-edge insights in learning. From early education to lifelong development, we dive into trends, innovations, and challenges facing learners of all ages. I'm Denise Pope, Senior Lecturer at Stanford's Graduate School of Education and co-founder of Challenge Success.

Dan Schwartz (00:37):

And I'm Dan Schwartz. I'm the Dean of the Graduate School of Education and the Faculty Director of the Stanford Accelerator for Learning.

Denise Pope (00:47):

Together, we bring you expert perspectives and conversations to help you stay curious, inspired, and informed. Hi, Dan.

Dan Schwartz (00:57):

Hi, Denise.

Denise Pope (00:58):

I'm going to start us out with a question right away. What do you think policymakers know about how parents are doing with their young kids? What do they know? What do policymakers know?

Dan Schwartz (01:11):

Oh, it's a great question. Because for the economy, we have leading indicators, like consumer sentiment. We have the amount of money that's been spent, traded. I don't think we have anything for early childhood. They would all be economic indicators.

Denise Pope (01:26):

We have nothing. What do you mean we have nothing?

Dan Schwartz (01:28):

Do you think our governor knows something about how it's going on at home with my kid, whether I'm worried about food? Like, how would he know?

Denise Pope (01:41):

I don't know. I guess we're lucky that we have our guest here.

Dan Schwartz (01:43):

Well, funny you should mention that. We do. We have an expert who's tried to solve this problem. So today, we have Phillip Fisher. He's the Diana Chen Professor of Early Childhood Learning in the Graduate School of Education, and he is the Director of the Stanford Center on Early Childhood, which is at the Stanford Accelerator for Learning, which is in the Stanford Graduate School of Education, which is within Stanford. So Phil, welcome.

Phil Fisher (02:09):

Thank you. Pleasure to be here. I like the nested design. We've been talking about nesting and concentric circles a lot in our world these days.

Dan Schwartz (02:17):

Yes. So how'd you solve it? I assume I'm right that policymakers sort of have no way of knowing how early childhood parenting is going on. Is this right? Or –

Phil Fisher (02:28):

I would say no, not necessarily historically that policymakers have had no idea. What I would say for sure is that the information that policymakers have access to is often not particularly actionable. And it's not actionable because number one, it's collected at very infrequent intervals, and also the information that's collected may not map on directly to policy decisions. So there are household surveys that are out there that ask parents about what's going on, but it's often the case that they're done two years apart, five years apart. And so once the information has been compiled and comes in, it’s sort of already old, and it may not relate to actual decisions that policymakers might make.

Dan Schwartz (03:23):

I see. So hence, RAPID turns out to be a very clever acronym because it's a fast survey. So what does RAPID stand for, Phil?

Denise Pope (03:32):

And wait, let's just make it clear where you popped up with the name RAPID. This is one of Phil's major things that he does. Yes?

Phil Fisher (03:40):

Yes. It's actually become perhaps the biggest thing that we're doing at our center right now. RAPID originally was an acronym that stood for Rapid Assessment of Pandemic Impacts on Development. So it started at the outset of the pandemic as kind of an emergency response to what households with young children would be living through in the kind of unknown times that were happening. We originally designed it and were lucky to be able to launch it with some funder support very quickly, and thought it would last for about three months. And April 5th of 2025 marks the fifth-year anniversary for the survey. So we thought it was a very useful tool for understanding and making it known what households would be living through. It's turned into a much more powerful tool that is really oriented to policymakers around what's going on, not just now in households, because there's also a survey of the early care and education workforce that runs in parallel to it. But so it's really about what's happening in the lives of adults and children who are connected in the early childhood space.

Denise Pope (04:54):

Can you give an example of some of the types of questions you're asking?

Phil Fisher (04:56):

Yeah. The survey asks about, every time it goes out, about five specific areas of well-being. So it asks about how children are doing. It also asks how parents are doing, for the household survey. Then we also – and one of our major areas that we've looked at consistently is economic hardship. Do the participants in the survey have enough money each month to pay for basic needs, like food, housing, utilities, child care, health care? It also asks about access to and utilization of healthcare and access to and utilization of child care. So it's sort of a surround sound of what is going on in families around well-being and around how needs are being met or not being met.

Dan Schwartz (05:43):

So just to make it really concrete, dear Dan, how is your well-being on a scale of one to five? What is a question that you might ask?

Phil Fisher (05:52):

So dear Dan, in the past month, since the survey now goes out on a monthly basis, how would you rate how you've been doing in terms of your stress level on a one to five scale? How would you rate how you're doing in terms of worry and anxiety? How would you rate your sense of loneliness? And how would you rate depression? So it's that direct. When we ask about child well-being, we don't ask about those kind of more adult-oriented categories. We ask does your child seem fussy and fearful? Does your child seem to be having more, kind of, acting out problems? So what people often talk about is sort of internalizing, more kind of anxiety, depression, or externalizing, more kinds of acting out. So those are the kinds of things that we ask about. And the survey only takes about 15 minutes. So we're not doing a deep dive questionnaire where we ask about these things 20 or 30 different ways with lots of questions. These are really just very focused kinds of questions that have what people call face validity. That is, we're asking about what it sounds like we're asking about.

Dan Schwartz (06:56):

I’m a parent, this comes to me in an envelope. Does it have a dollar in it? I'm not sure, I'm really stressed, I don't particularly want to answer this. How do you get people to do it?

Phil Fisher (07:06):

Great question. So when we designed the survey originally in the early days of the pandemic, we knew a couple of things. We knew we needed to get data very frequently. As I mentioned, we started by this being a weekly survey, and it's since sort of gone to monthly. We also knew that we needed a very large number of participants that were as close to as possible representative of the US population. What we were really fortunate to have at that moment in time, and continue, is partnerships with really amazing online organizations that have communities of families with young kids connected to them and with whom there is a very strong sense of trust.

(07:47):

So we partner with these organizations and they send out requests to parents and information about what our survey is about. And then if parents are interested, they get a link via email or via text that they can then use to connect to the survey, agree to participate, and then start filling out the survey. And every time people take the survey, they get compensated $5. So we know people are busy. We always feel like if people are contributing time and effort, that they should be compensated. $5 has been kind of the sweet spot in terms of what it takes to get people regularly participating without actually – when we increased the amount, which we tried for a while, we started to get a lot of invalid so-called bot responses. And so $5 has been kind of where we ended up.

Denise Pope (08:36):

So if they need a link, you're already at a level as a family that you have access to the internet, you have a device at home. Are you worried about missing really under-resourced families because they don't have a link?

Phil Fisher (08:53):

So glad you asked, Denise. It's a really good question. You're absolutely right that the way in which we have historically administered the survey requires some kind of access to the internet. It does not require high-speed Wi-Fi. You do need a smartphone or tablet or computer so that you can access the survey. We do and have recognized for a long time that one of the things that that means is that we are potentially missing some of the most marginalized people who are really not able to access that kind of technology. But we're now working in places - for example, we have a partnership with the Delta Health Alliance in Mississippi that provides a lot of supports and services to families that are extremely economically marginalized, in particular, with respect to early childhood. We're doing interview-assisted data collection. And we're testing out those ways to get to people who might be very isolated rurally or just might be so impoverished that they don't have access to internet. And that's allowing us to extend to these, kind of, most vulnerable populations.

Denise Pope (09:59):

So here's a question I have for you, and it gets asked of my own studies as well. When it is a self-report study, how honest do you think people would be? I mean, they're not going to say, "I'm spending this money on drugs."

Phil Fisher (10:10):

Correct.

Denise Pope (10:11):

Right? But I get asked that all the time. I get push back all the time around self-report. So how would you answer that?

Phil Fisher (10:16):

There are some really interesting ways that people have been getting at this. Our survey does not provide us with more than self-report. Again, when you're trying to do a large-scale survey, we can't use these methods. But for example, there are some studies involving basic income or what are called cash transfer payments to families on a monthly basis where they can track – because the cash is being distributed through an electronic benefit card – and so they can track where those EBT expenses are coming from and have been able to regularly identify they're not being used at liquor stores, they're not being used in general to just withdraw cash, in which case, we don't know where they're going. They're being used at grocery stores, they're being used at stores that have household goods and things like that. So I think you're absolutely right. If somebody is using these kinds of extra income for nefarious purposes, it's a little bit hard to track. But I think the idea that that's probably an anomaly and that we have multiple sources of data to suggest that is fairly conclusive at this point.

Dan Schwartz (11:26):

So what is it that your survey's asking that is politically actionable that the other things don't ask? What are some of the political actions you've seen in response to the survey information?

Phil Fisher (11:37):

The survey data has been impactful from the perspective of policy. In our household survey, what we have found, some of our families just take the survey once or twice, but many of them have taken it multiple times across time. And for those families, when we identify that they report material hardship in a given month, and then we follow up the following month, we find elevated reports of parental distress in those four categories that I'm talking about. And then when we follow them further out in time, we find that in subsequent months, we see increases in child distress. So if you're a policymaker and you're worried about the well-being of children or you're concerned and want to make sure that kids are doing okay, we're able to draw this direct link between how families are doing economically, how that's impacting parents, and then how that's being passed along to children.

(12:33):

It's not rocket science. It's the kind of scientific finding where you say, "Well, your grandmother could have told you that without any data." But the fact that we continue to find it, that it's a continuous drumbeat that we can show in our data at specific windows. And when we've added, and we have done this a few times, measures of child development, not just child well-being, we see that these patterns consistently remain. And so if you're interested in child well-being, then you're interested in parent well-being. And if you're interested in parent well-being, then you're interested in economic stability and security.

Denise Pope (13:06):

It's so important. And although I think grandmother's instincts should lead to policy changes, right, they often don't. And someone needs to collect that data to show the policymakers the sort of, "No duh," right? Then say, "Hey, we need this." Yeah. So Phil, you also mentioned that this doesn't just go to parents, but it goes to caregivers, daycare providers, early childhood educators. Is that right?

Phil Fisher (13:34):

Yes. Correct.

Denise Pope (13:34):

Tell us a little bit about those findings.

Phil Fisher (13:36):

Yeah. We realized after about a year of the survey that we knew that we were missing a big chunk in the ecosystems of young children because if we're just focusing on family, we were missing these care environments where children were learning, were being taken care of so that parents could work. So we started a parallel survey. I think the main finding that I really want to share about that survey is that I've been talking about the economic hardship of families and households with young children. It turns out that the rates of hardship, the proportion of providers that are experiencing hardship is higher than what we see in families. And there's been lots of media coverage about this, about how child care is a bit of a broken market in early childhood education, that parents can't afford to pay. A lot of parents spend more than a third of their paychecks on care, but that providers can't really afford to live on what they're earning.

(14:28):

And this is what we found. So consistently, above 40% of the providers in our survey say that on a monthly basis, they don't have enough money to feed their own families. They don't have enough money to pay for basic healthcare and medical expenses. And I think that, again, has resonated with some policymakers around this idea of, how can we expect people to provide high-quality care to our children if they're hungry, their stomachs are rumbling, or if they're worried about getting evicted? And so I think that's sort of been a big focus of a lot of policy work. And our data really helps to furnish information about these critical issues.

Denise Pope (15:04):

I just have an anecdotal point about that, which is I have now had two Uber drivers who, when they find out I'm in education, have told me that they are preschool teachers during the day and Uber drivers at night. That's two. And I don't take Ubers all that much. So that tells you something. They're not getting enough. And they're wonderful, they seem like lovely people who would be fabulous child care providers. And here they are staying up late driving Uber because they can't live on their money. So we have a problem.

Phil Fisher (15:30):

It's a very good example.

Dan Schwartz (15:32):

So Phil, has it cashed out in political decision-making?

Phil Fisher (15:36):

Yeah. Here's a good example, since we're talking about providers. So our survey has gotten the attention of not only policymakers kind of at very local levels, but also one example of this is in the state of Oregon, we had a meeting with the then-Governor, Kate Brown, who was tracking what we were putting out in the survey. And when she heard about the amount of hardship that caregivers were experiencing, what she did was to make sure that the policies around pandemic relief efforts to child care providers were available not just to pay for things like enhanced cleaning supplies or other ways to make their centers more safe for children, but that they could use them to pay for basic needs. And that was not the original intention.

(16:28):

It was kind of like how do these centers keep their lights on given that it's more expensive to, you know, you're not getting payment from as many people whose kids are in your care because people are staying home. And it really changed the policy to make it more oriented to the more holistic needs of the care and education environment. And we hear lots of stories like that.

Dan Schwartz (16:47):

So Denise, you probably used a baby monitor, right? I did not need to. I lived in a very small apartment in New York City, so there was no getting away from the kid. I was there all the time. It's not like I'd go outside on the first floor and the kid's on the sixth floor.

Denise Pope (17:09):

Well, I will tell you a funny story about a baby monitor, right? So my first kid, you have a baby monitor, we're all outside. My mother-in-law is visiting in town. And she says, "What is that?" And this is a woman who has raised six children, obviously before the time of baby monitors. And I said, "It's a baby monitor." And she said, "Well, what do you use it for?" And I said, "Oh, it's because we're outside, so we want to hear the baby cry." And she looked at me and she said, "Why would you want to hear the baby cry?"

(17:38):

That was the most ridiculous thing in the world. She's like, "I raised six kids. And we just were trying to get away from all of that," right? So it just shows you just how the parenting has changed from, you know, she once told me if everyone had shoes and socks on, it was a good day. If everybody, like nevermind, everything else could be going to crap, but if everyone had shoes and socks on and got to school, call it a day. So yes, we did use a baby monitor. So was it just a very different time and a very different opinion.

Dan Schwartz (18:09):

Right. It was like, "No, if the kid's busy, I don't want to know."

(18:12):

So you've been working with RAPID on policy levers, but you also do stuff directly in homes, right? So first say what FIND, the acronym, stands for, and then describe it briefly.

Phil Fisher (18:30):

Yeah. FIND is an approach. It's a very strength-based approach that uses video in natural environments. It stands for Filming Interactions to Nurture Development. And it's an approach that is part of a large family in the early childhood field of coaching to help improve the, kind of, quality and sensitivity of caregiving. I will correct you just in one small regard, which is that we started this as a home-based model to coach parents. We quickly got interest and demand from the early care and education field as well. And so we've actually implemented FIND not only in family context, but also in the context of child care and early education environments. In fact, we've actually done a statewide implementation over the last many years in Washington state of a system called Early Achievers, which is a coaching system for both home and center-based providers in Washington state.

Dan Schwartz (19:27):

So strength-based means you have to observe their strengths, and then you point it out to them? Is that how you do it?

Phil Fisher (19:32):

The basic premise is that people are actually naturally engaged in behavior and interactions with children. We call them serve and return interactions that science, decades of science, shows are the basic building blocks of learning and development. And the issue is that these instances where these things are happening occur on a very quick timescale and are often sort of below the level of awareness that an adult might have of these things being really critical or that they're doing them. But in collecting video from thousands of families and care and education environments, I can say that these basic building blocks are fairly ubiquitous. That is, they happen almost all the time, even in very high adversity, low resource contexts. And it's really a matter of identifying them and being able to make them salient or noticeable to the adult.

Dan Schwartz (20:28):

So I'm just trying to make this concrete. You've got a video camera watching an adult interacting with the child. You then take the video and somebody looks through it to find really good moments of interaction. And then you come back and you show the adult, "What you did here is great. Do more of that"?

Phil Fisher (20:47):

Bingo. That's exactly right. I think the trick is that identifying these moments requires kind of a measured eye. It's not rocket science. Again, it's pattern recognition. It's can you identify the signal amidst the noise? But if you just try to say, "These are good things. You should do them," or try to show somebody a long sequence of video, "Let's just watch 10 minutes of you with your child or with the kids in your care," people don't see it.

(21:15):

And so the thing that we've had to do is develop a system for editing and then for kind of doing a highlights reel, just like you might see in sports where we're showing something in real time, but then we break it down into much smaller pieces and then reconstitute back into real time. And by doing that, we've done lots of research on this, lots of what people call randomized clinical trials and we see very positive impacts by doing this. We show people things that they then dramatically increase, and then we get better child outcomes. But the challenge has been that it's very laborious and resource-intensive to edit these films in order to make them into the kinds of tapes that we want to be able to show people.

Dan Schwartz (21:59):

Hang on, Phil. So Denise, watch this. It's all AI all the time.

Denise Pope (22:04):

I knew this was coming. Because I was going to say, I think we have something to help you, Phil. And then you go into it. Can you just give us a quick example of a serve and return? What does that look like?

Phil Fisher (22:15):

When we developed the model, we began with the idea that serve and return could be any interaction back and forth. As we began to really examine much more closely what it was that was developmentally supportive, that has helped kids learn, helped brain development, what we realized was that the serve has to start with the child. So a child serve is them exploring the environment, looking at things, vocalizing, gesturing at things. It's anything in which the child is exploring the world around them. It can be directed towards the adult, but it does not need to be. And a return is when the adult notices what the child's doing, acknowledges it, and responds in some way that makes the child understand, "I see you. I hear you. I understand what's going on." That gives you language, which helps your own language development and describes and narrates what you're doing.

(23:08):

And again, it may be that an adult does that on a very infrequent basis, but in 10 minutes of video, which is about what we usually collect, we always find instances where the child is doing something and the adult responds in a meaningful way. We think that this is something that's kind of conserved through evolution. That is, most people find kids pretty cute and tend to notice when they're doing things and tend, in some ways, to acknowledge and respond. So it's why it's sort of fundamentally strength-based, because these are skills that people already have. And we're increasing the likelihood that they will use them more often rather than trying to stop them for doing things that are not productive or kind of a deficit model that they don't know what to do and we have to teach them.

Denise Pope (23:56):

I love it. It’s kind of innate too.

Phil Fisher (23:58):

It is, in fact.

Denise Pope (24:00):

That's so cool.

Dan Schwartz (24:01):

Yeah. I don't know. It's very different than when I taught middle school, right? It's like, "Stop that."

Denise Pope (24:05):

"Sit down."

Dan Schwartz (24:06):

"Quit it."

Denise Pope (24:07):

"Don't throw the-"

Phil Fisher (24:08):

You were returning a serve still, Dan.

Dan Schwartz (24:10):

Yeah.

Denise Pope (24:11):

Yeah. He does that with us. Super, super fascinating.

Dan Schwartz (24:16):

So I have to go all AI all the time for a second. So coding video, if you're a researcher and you have these videotapes, coding is like an hour for every five minutes. So this must be... It seems very time-consuming, expensive to look at the videotapes. So AI, are you using it?

Phil Fisher (24:37):

Yes. It's not only time-consuming and expensive, it's also sort of an ethical conundrum because if you're a researcher, you might be able to get more research assistance, but if you're a community-based organization, you're trying to help families or teachers and you're spending your money having people edit videos, people might object and say, "You should be spending this time with the people who are working with the kids." Stanford has incredible resources around AI and machine learning. And since I've gotten to Stanford, together with a colleague, Lauren Klein, we've been working with faculty member Nick Haber. And what we're doing is to train machine learning models to be able to identify these patterns automatically. Not only to be able to say, "Here's where it is," but to be able to then, using computer vision, actually create transcripts that describe what's happening, which is what our editors have had to do previously by typing stuff in, and to transcribe the words that are being spoken in that instance.

(25:38):

And so what we're on the cusp of being able to do is really take the human out of the part of this program that's been our major barrier to scale, and only have the human efforts be oriented to reviewing what the machine learning models find, and then essentially be able to fine tune them for use in these kind of coaching contexts. So it's really been remarkable. We've been at this for now for just over two years. And the fact that we're ready to start trialing this with industry partners who are interested in using these techniques on parenting apps and things like that, to me, is really remarkable. It just shows, number one, the resources we have here at Stanford, but also how quickly this field of AI and machine learning is evolving.

Denise Pope (26:25):

Phil, if you had to give one piece of super impactful advice to early childhood parents or caregivers based on these findings, what would it be?

Phil Fisher (26:40):

It would be tune into your child, to the children in your care, notice what they're doing and use language to interact with them. It would also be that you already have all of the superpowers necessary to build healthy brains and high-quality learning and to improve academic achievement. You just may not know that you already have them, but they're there. And if you use them, you're just going to help your child be smarter, be more successful in all domains of life.

Denise Pope (27:11):

I love that. Dan, what did you learn from this episode? That's my favorite question for Dan. See if he's really been listening.

Dan Schwartz (27:19):

Well, I learned a lot. I learned a lot. But I think my task is to do an abstract summary.

Denise Pope (27:23):

Okay. Go for it.

Dan Schwartz (27:25):

The thing that stands out is how much Phil's endeavor is to bring information to light, that people don't see it, they don't know it, whether it's the RAPID survey going up to politics or decision-makers, or it's parents at home. So the information's there, but we need to find how to get it out in a way that people can respond to.

Denise Pope (27:48):

And Phil's doing it. That's what-

Dan Schwartz (27:50):

Yeah. So it's a very university-based approach to improving early childhood.

Denise Pope (27:54):

But not just ivory tower. Really making a difference.

Dan Schwartz (27:58):

Yeah, yeah. But making the difference by bringing information to the front so that people can use it.

Denise Pope (28:04):

I love it. You know what my favorite part was? It's already in us. So many people are so afraid that, you know, you don't have a degree in parenting. So many people just parent the way they were parented. And I love the asset-based approach where you're not telling them, "Hey, you're doing this wrong." You're saying, "You already know how to do this. And you're doing these things right. And we're going to replay it for you so you could see it in real time and learn to do it again." So that is super cool. Phil, thank you so much. I know everyone learned so much from this. And I want to thank you for coming. And I want to thank all of you for joining this episode of School's In. Be sure to subscribe to the show on Spotify, Apple Podcasts, or wherever you tune in. I'm Denise Pope.

Dan Schwartz (28:43):

That was fabulous, Phil. Thank you. I'm Dan.