Skip to main content

Transforming the understanding
and treatment of mental illnesses.

Director’s Innovation Speaker Series Video: Youth-Centered Approaches to Media Research

Transcript

DR. SHELLI AVENEVOLI: Good afternoon, everyone. It's really my pleasure to be here today to introduce our two speakers. I also want to give a shout out to the Innovations Series team. I want to thank them.

When I took over as Acting Director in June, I had a conversation with them and they asked if I wanted to add anything to the Innovation Series, and the thing I asked is that they include some talks on incorporating youth voices and youth mental health into the series. Thank you for helping me achieve that. Thank you for all the work you've done to prepare for today.

Thank you also to Dr. Radesky and Dr. Moreno for joining us today. I'm really looking forward to their talk. I have heard a very early version of this and was very excited by the work they're doing. So pleased that they can be here today.

To tell you a little bit about I'll introduce both of them and then we'll move into Dr. Radesky’s talk first.

So, Dr. Jenny Radesky is an Associate Professor of Pediatrics with tenure at the University of Michigan Medical School and Division Director of Developmental Behavioral Pediatrics.

She earned her BA in natural sciences from Johns Hopkins and her M.D. from Harvard Medical School. Since 2022, she has held leadership roles in her division, including serving as chief and Division Director of Developmental Behavioral Pediatrics.

Dr. Radesky is board certified in both pediatric and developmental behavioral pediatrics and has been appointed as a behavioral expert with the U.S. Federal Trade Commission since 2024.

Her research focuses on the intersection of early childhood development and digital media use, particularly how parental mobile device usage impacts parent/child interactions and child behavioral outcomes.

Through innovative methodologies and collaborations with interdisciplinary researchers, she continues to advance the study of media use in early childhood, striving to translate these findings into clinical practice and policy.

Dr. Megan Moreno is a professor of pediatrics and affiliate professor of educational psychology at the University of Wisconsin Madison.

She's the chief for the Division of General Pediatrics and Adolescent Medicine and serves as vice chair of Academic Affairs for the Department of Pediatrics. Dr. Moreno is an adolescent medicine physician and researcher.

She's the principal investigator of the social media and Adolescent Health Research Team, and her research focuses on the intersection of technology and adolescent health.

Dr. Moreno has authored over 200 research articles as well as written and edited several textbooks. She currently serves as the co medical director for the AAP Center of Excellence on social media and Youth Mental Health.

Welcome to both of you.

DR. JENNY RADESKY: Thank you. Thank you so much for that nice introduction and for inviting us here. Dr. Avenevoli, it's a pleasure to be able to talk with other researchers who think about mental health and child well being in such complex and nuanced ways and to be able to talk a little bit about my research at University of Michigan as well as our translation efforts with the Center of Excellence on social media and Youth Mental Health. It's funded by SAMHSA, run through the American Academy of Pediatrics, and Dr. Moreno and I are the co medical directors. So really interested in making sure that the science we do has impact, and families are benefitting from it.

My goal today is to talk about how in my research approach I've tried to introduce novel both conceptual approaches and measurement approaches because mobile media and technology is just a different beast than television was, and it requires new approaches.

And then I'll also talk about how I've tried to create research questions that have easy translational relevance for clinical work.

I wanted to start by just pointing out how rapidly this topic is changing, and this is just a little bit of my career trajectory here. So, I graduated from med school in 2007 when we had flip phones. Then the iPhone came out, and I did my pediatric residency at Seattle Children's in a tech hub where mobile devices were now popping up in exam rooms.

I found it fascinating, and I really thought, like, I want to study this. I was really interested in early relational health and social determinants of health. When I got to Boston Medical Center, I started trying to figure out ways to measure this and try to study it within the context of early childhood relationships and psychosocial stress.

And then by the time I got my first K award from NICHD in 2017, it was already just a much more ubiquitous exposure in childhood. And tablets and smartphones have very different design affordances compared to TV.

Now, most of the research, when I was starting my K award in the prior decades, had really been focused on these big stationary boxes that sit in our living rooms or our bedrooms that often were providing unidirectional feeds of communication or entertainment and not bidirectional and responding to us.

They weren't portable. Yes, we had portable DVD players and things like that, but these little computers could be taken through any daily routine, especially family routines, that we think are important for child social emotional development.

They're also just used so frequently compared to the way we used to have to start up a laptop and wait for it to boot or to call, dial into dialup.

These, you just grab it right away and can use it. So, it leads to sometime fragmentation. There's more task switching, which means that recall may not be as accurate.

I've also been really fascinated in the fact that with touchscreen, interactive user interfaces, there's a lot more for children to interact with and do. But there's a lot more persuasive design, which I'll talk more in a couple of slides of how we've tried to change our measurement methods to capture the way technologies are trying to interact with our psychology.

And also, it's very personalized. So, when we used to think what did you watch on Friday night TV or Saturday morning cartoons, kids generally had the same exposure compared to now, a lot of their feeds are very personalized and is shaped by both data collection and the marketing that underlies a lot of that data collection.

So, I was in my training thinking, how on earth do I measure this? This is very different than just screen time recall. And the dominant pediatric research framework, the articles that I was reading avidly, and the cultural narrative was very unidirectional and simple, straight lines between screen time exposure and a bunch of different negative outcomes.

I once gave a talk this was like five years ago but I gave a talk, and I was trying to capture also the polarized narrative around this topic in online spaces. And so, I Googled child screen time, and I got these two articles within the first couple of search results.

So, I took screen shots because I was, like, this is fascinating. And these are the messages that parents are hearing, too, of, like, this global construct of how much screen time your child is getting and this fear of, I'm somehow harming this, I'm somehow just not being productive enough or perfect enough in my parenting.

And as a developmental behavioral pediatrician I saw a few things that were wrong with this approach. Number one, kids always bring something to the table. It's never just exposure to child. So, there's so much about the transaction between a child's innate characteristics and their environment, including their technologic environment. Number two was the guilt that these messages carried to families, and they were not precise messages of what to do.

I wanted to study this on different time scales and different conceptual frameworks that really got more at, okay, you wake up with your kids, what are you going to do? Where does technology come in and where doesn't it? Are we using technology to calm down or not?

So, I wanted to have more clinically relevant constructs. That's why one of my big approaches in my K award and subsequent grants has been relational frameworks. Not to think about this as child as the unit of what's happening, really thinking of that child in the and in early childhood this is a child's perspective, is that their development is happening within the context of a relationship.

So, I was very influenced by Arnie Sameroff's work on transactional models between the child and caregiver. This was also what I was seeing in my clinics, is that you have certain child or parent characteristics that are shaping and building off one another, sometimes positively, sometimes a little bit butting head and negatively.

But it's this cascade over the short term of a day or an argument or a play interaction or the long term of how things shape through the first couple of years of life that really shape child social, emotional development.

So, I wanted to think of research questions and methods that would help me understand how parents are using technology in their interpersonal spaces with children, how are children using technology in the emotional spaces that they share with their parents as they're learning to regulate their emotions.

I'm going to just show a couple of slides because there's not enough time to kind of go through all the different ways, I've tried to measure this, but just briefly, my very first study was just observing families in fast food restaurants, where we just sat down and took field notes and said, what's going on here? And tried to be really judgment free and just type down like at moment to moment what was happening when parents or kids used a mobile device.

Other approaches we've used have been detailed behavioral coding of videotapes when parents spontaneously took out a mobile device during very boring feeding interactions in some of my mentor studies or during their family meal times at home.

But one of my favorite studies that we published that really fits with this transactional model of child development was a collaboration I did with Brandon McDaniel where he had studied 183 families over the course of six months, getting measures from both parents. And then the children were between one and five. So, they did the CBCL to look at externalizing behavior.

So, he knows how to run models like this that, I'm sorry, is very busy, but I just want to point out that the paths between technology interference, which is a measure he came up with of, on a typical day how much are you using technology during times when you're with your kids?

So not of screen time variable, much more a context of usage. And found that when adjusting for a lot of other parent characteristics, technology interference predicted the development of later more externalizing symptoms in young children.

So, the thought being maybe this is disrupting some of the parent/child play or conversation or nonverbal interaction that we had seen in some of our videotape studies.

But then the converse was also driving some of this model, is that when you're raising a child with more externalizing behaviors, you have more stress, which definitely I see in my clinical world. And that, in most of these paths, was driving or associated in a predictive way technology interference.

So, this tells me, clinically, we can't just say parents put your phones away, you're ignoring your kids. You really have to address some of the underlying stresses, including child rearing stresses that lead to parents wanting a virtual escape through their phone.

We've done a lot of qualitative work with parents to say what's this day like. How can we work with you to get that little escape with you, so you don't yell at your child but then also repair after the difficult moment has passed.

This is a study with an R21 I got from NICHD where we studied kids over six months. These are three to five year olds. What's interesting about this is that we tried to look at those bidirectional associations between, not screen time, but how likely are you to hand your child a mobile device when they're upset and need to calm down, and the child's emotional reactivity, again, on the child behavior checklist.

In the full model of throughout 350 kids, there were not significant paths. But if you stratified by temperament, it was only the high surgency kids, and the boys which I'm not showing here, where there were these bidirectional associations between more uses for mobile devices for calming and higher emotional reactivity downstream and then more emotional reactivity predicting more use of mobile devices for calming.

We found that this was an interesting test of this hypothesis, that if you have a child where some of these emotion regulation skills don't come naturally to them, which are a lot of the kids I see in my clinic, then they may need extra work and effort to learn some emotion regulation skills.

As an aside, this paper was also talked about on SNL's Weekend Update, which was like a highlight of my career. Second to this right now being at NIH.

So that was my relational lens. Then the other thing I have really tried to bring into my research has been a human centered design lens. This is just knowing how much time someone spent with the screen is not as important as what happened during that time and was it a positive or a negative interaction. How is technology nudging us to do different things to meet our needs or meet the company's needs.

The other reason I like taking a design lens on my research is because I think it has greater public health relevance. A lot of the advising we do through the American Academy of Pediatrics is very at the top of Tom Friedman's Health Impact Pyramid where you are one to one talking with a family about here's how to manage your screen time, but that requires increasing individual effort and it's going to have a less of a population, especially an equitable population impact.

But if you change the context to make individual's default decisions healthy, like removing lead from gasoline or removing trans fats from food, you are more likely to help support opportunities for positive use, including in this case positive opportunities to have positive experiences online.

So, one way we've tried to study design is to measure what are kids actually doing on their phones? Because I wanted to play the games they played. I wanted to watch the YouTube videos they watched and really try to understand what design features occur.

This is again from our R21 where we tracked any child who had a device of their own, we tracked it either with an app that we developed with support from an STTR grant from NICHD where we tracked their Android device and got time stamped output on what they were using when. We categorized it into these big categories.

You could see YouTube. YouTube Kids is like number one in terms of time. For iOS devices, we were using screen shots of their battery or screen time screens. So, what was really interesting to me about this is that through the AAP we recommended a lot of time on PBS Kids, or video chat, which both have really low usage across young children.

And a lot of kids were just watching videos and playing games. So, what did we do? We watched a lot of those videos and played those games. Oh, just to make the point here that we could also, on Android devices, look at time stamped data. So, it was really fascinating to us which sorts of apps tend to run overnight in particular kids. Those being the blue and the yellow is YouTube, YouTube Kids and games.

What we found also is those types of media have more engagement prolonging features that make kids want to watch longer or play longer.

So that's the next slide. We've studied these apps in lots of different ways. How educational is the content compared to academic standards? How much data are they collecting on kids? But one that we published two years ago was how do the characters or the designs in these apps try to nudge the children to spend more money, play longer or come back every day?

So, apologies. This one on the right is not from one of those apps. That's just from our YouTube research where we've studied the algorithm and tried to see how they try to grab kids' attention. But the other things you'll see is like a miraculous ladybug sending a text to the child to come back and help them save Paris.

You get this fabricated time urgency of, like, come save me and you get special, all these different levels of currencies that are shiny objects or you get the character urging the child to come back every day to get different weapons or different virtual currency or things like this.

We wanted to explain to parents, it's not your fault that the kids want to play a lot of time. There are a lot of design tricks that are in these apps that are really encouraging kids to stay on longer.

So, I want to wrap up by putting this all together in terms of, like, where are we going next? Each of these studies I did in isolation. But how could we design studies that really get at what I've talked about today? Individual differences, the relational and family context, the emotion functions or regulatory uses of media, how we're using media to calm down or escape from stressful social interactions.

The shorter time scales of minute to minute, the longer time scales of year to year, and how do we capture design features that shape our media use, or the structural factors like COVID 19 or childcare access that shape how much our kids are using media.

So, this is where our research is going. This is our PO1 group that's also funded through NICHD. My collaborators, Rachel Barr Heather Kirkorian and Sarah Coyne, with the funding from NICHD we did a writing retreat, and we drew out the conceptual framework on a napkin.

It was a lot of fun just to get together and think through how can we capture it all from the individual and contextual factors that shape media use and our motivations for it, media dynamics at varied time scales such as, like I've said, calming, or for joint media engagement, and the sort of responses that we have and how that shapes longer term outcomes.

So, our PO1 is looking at this on longer time scales where we're merging data from our three cohorts, Project emU, Mitten and Media, that should be able to span from one year to seven years of age where we've tried to harmonize our data collection approaches.

And then it's also shorter time scales. This is data from Project emU where they do daily surveys to the parents at the end of the day about their media motivations and their parenting experiences.

These are time lagged by one day. So what's great about this is you see, when parents perceived their child as difficult, they were more exhausted the next day and they used more media to regulate their emotions the next day and that fed them into more feeling that their children were difficult, compared to when they just used media to kind of chill out, take a break, and go back to their kids, there was really more of a synergistic relationship with more positive moments with their kids.

We're trying to get down to this really specific place where we can give families guidance of, it's okay, go check out for a little bit, but then come back and try to have what you see as positive moments with your kids.

So, I'm going to wrap up here and transfer it over to Dr. Moreno. But this is what we're focusing on right now. That relational health matters how media connects us or creates distance, that children's emotional relationship with technology matters. That parent well being matters, especially is it a risk or a resource when families are stressed? Time scale matters, looking at both short term and long term effects, and the other parts of the ecosystem, what structural factors are at play that are decreasing or increasing kids' opportunities for positive well being.

And then what's been really fun is, with our Center of Excellence, when we get questions through our question and answer portal for some of the guidance we've created for pediatricians or parents like the five Cs of media use that Dr. Moreno will talk more about or work trying to advocate for more child centered design, all of this really feeds upon the work we've done, and it's really easy to say, like, oh, we can cite one of these studies we did with NIH funding to then come to say, here, parents, try this strength based approach that we hope makes you feel less guilty and more empowered to try to make some positive changes.

I'll hand it over to Megan now. I wanted to say thanks to my multiple grants and collaborators, and then...

DR. MEGAN MORENO: I'm really excited to be here and to share some of the work we've been doing focused in the area of adolescent technology and digital media use. And what I would like to share is expanding a little more on the five Cs framework that Dr. Radesky framed up and then talk about some structural elements and some of our early findings in the brain behavior and well being study, which is the PO1 project we have right now that's funded by NICHD.

And I'm hoping that I can tie these together and have us think together a little bit about the relevance as we look to future research, as well as, as Jenny said, we're always thinking about translation and dissemination through the Center of Excellence.

Building on the great content that Dr. Radesky shared, frameworks are so important. They really shape our science through things like conceptual models, but they also get translated to the public and they shape the narrative around how youth interact with this topic and how educators and parents and other adults teach the child about that topic.

And a really common narrative for adolescent technology and digital media is substance use or addiction.

I've been at conferences where people have talked about social media as the new tobacco or as this is as devastating as the opioid crisis.

In addition to being somewhat hyperbolic, especially for folks who work in the substance use field, it's an example of a not helpful framework. It really centers the blame on the child; the framework of addiction really places the blame on the child. It ignores all the aspects of digital design.

It really removes agency from youth and removes their own ability to use their strengths in navigating their digital experiences, and it also really frames media as something that's all bad. And it's such a complex topic and kids have such complicated experiences that dichotomizing it as either all bad or all good is not a helpful framework for us as scientists.

And so, this is why, when Dr. Radesky and I really started our work with the other experts at the Center of Excellence, we put a lot of thought into thinking about frameworks that we could use to guide our work. We conducted literature reviews to understand the science across different disciplines.

We did a lot of listening sessions across caregivers and youth. We really wanted to incorporate the perspectives of folks who work in education, other researchers, as well as healthcare providers, both in our home space of pediatrics, but really incorporating the expertise of mental health professionals as well.

We also both brought a couple of decades between us of work of being in clinic and watching these interactions with kids and their parents and wanting to think about frameworks that we could take into those clinic spaces as well.

And that's what really brought us to this framework of the five Cs. Our goal is to have this be a framework that is useful in guiding conversations between kids and their caregivers as well as used in spaces like education or in clinic settings, as well as thinking about how some of those concepts might apply in different research studies.

So, in explaining out the five Cs, I'm going to use the framework of how we've thought about this being used in interactions between kids and their caregivers.

So, the first C is the child. It really grounds the whole framework in understanding that the child is at the center of it, not the technology and not the platform.

So, we really encourage caregivers to think about, who is your adolescent? What are the unique strengths and unique challenges that they bring that might impact their media use?

How does their personality, their resilience, all the pieces of them that make them, how does that interact with the experiences they have?

The second C is content. This gets at the real importance that it's not just how long you're interacting with media, but it's what you're seeing as that content piece.

And encouraging caregivers to make this a focus of conversations. What does your family define as high quality content? What do you learn from? What's really worth your time? In a sea of endless content, what do you want to pick?

And how does negative content impact you? How do you feel after seeing content that really negatively impacts you?

The third C gets at calm. We all recognize that in today's world, kids, as well as adults, we can sometimes use media to calm down after a stressful day or decompress or distract ourselves. And that's okay, but we also want kids to build out a toolbox that includes both online and offline tools so that they have an array of options to choose from when they're having those strong emotions or feeling stressed.

The fourth C is crowding out. That gets at really critical health behaviors that we know are so important during the adolescent development period. Sleep comes to mind. Physical activity comes to mind. Learning to negotiate in person interactions that can sometimes be a little cringy. That's an important part of adolescence, and we want to make sure teens are having all those experiences.

So being able to frame as what kind of activities are getting crowded out by media. What is that impact on you? And how do we crowd those back in?

The final C, which is probably my favorite, which is really encouraging ongoing communication between parents, caregivers, teens, other adults in their lives, and having that communication be really open ended and really interactive, where maybe parents sometimes are sharing challenging situations they're having navigating their own media spaces as well as being able to provide guidance and support for teens as they navigate their own.

So, with that as the framework, I'd like to shift gears and talk a little bit about our current study, which is the brain behavior and well being study. This is a PO1. It is comprised of three distinct projects as well as two cores.

And we're at the beginning of year three. So, I'm excited to share a little bit about the projects and the collaborators that we have in this work.

This project's background is centered in adolescent development, knowing that during that developmental period, adolescents have a lot of growth, a lot of changes across multiple areas, including cognition, behavior, emotional regulation. Some of the really critical unique tasks of adolescents include identity development and peer connection.

If we even just think about those two tasks, identity, development and peer connection, technology and digital media is almost perfectly designed to support both those tasks. It really closely aligns and is so integral to modern adolescents to be able to develop that digital identity, to put it out there, to get feedback on it, to develop networks, use it for communication. It's almost perfectly constructed to align with some of those tasks.

We know quite a bit about adolescents' interactions with social media. This field is a little over 15 years old, and a lot of the work has come at this topic from a very solely risk centered lens.

So, we know a lot about how adolescents represent health risk behaviors on social media. We know a lot about how they represent alcohol, how they represent tanning behaviors, how they represent self harm. How they might represent unhealthy diet trends. And I remember back I think one of my very first studies were looking at MySpace and how often kids talked about getting drunk on MySpace. Spoiler alert, it was a lot.

But that has really been such a central part to this work, and I think in more recent years we've really come to understand that adolescents are so much more than their risk behaviors, and there's so much else that they're doing with media that is not highlighting those risk behaviors.

So, we really understand that there's a lot we don't know. We don't know as much about how youth present themselves as members of communities, as members of sports teams, of church groups, of school based types of extracurriculars.

We don't know much about how they move their bodies and enjoy being active.

We know less about how they support each other with their mental well being and how they communicate and share their own experiences with mental health and well being, which is represented in the middle panel.

We have seen, particularly since COVID, just a lot more content intended to, sometimes satirically or jokingly, really lift up and celebrate a lot of well being aspects. Teams will often share those to each other or share them in group settings.

Another area that I've been interested in as someone who has done work in the substance use piece is we know less about how teens are pushing back on substance use and the narrative of that as normative.

So, one type of post that we look at is posts that we call SAM, which is Sobriety, Abstinence and Moderation. And these types of posts are something, over the last 15 years, we've really seen increase.

I think our study that ended in 2016, we found that less than 1% of posts were like this. But a study that went between 2021 and 2023, we actually found 15% of posts related to alcohol fit in this sobriety/safety moderation category.

So, there's a lot going on out there that we haven't been looking for and that I think is really important for us to understand to get a more holistic view.

So, within our brain behavior and well being project, we have three distinct and synergistic projects. They utilize the same participant pool. So, our participants are recruited between the ages of 13 and 15, and they all enroll in the study for a two year time period.

And over those two years, they have multiple data collection time points that include self reported data via surveys and interviews. We observe the way that they share information about themselves on social media. We're interested in how they're crafting their online presence and what they're showing to other people.

We have functional MRI data, and we also have a component of Ecological Momentary Assessment.

Our three projects are listed here. The first one focused on looking at self generated and other generated content towards understanding mechanisms in adolescent health and behavior. And that's the behavior component of the three Bs.

The second project focuses on functional magnetic resonance imaging to understand how positive and negative experiences within technology and digital media relate to mental and behavioral health.

This project is led by my collaborator, Dr. Chris Cascio. This is our Brain B. And our third project focuses on using mixed methods to understand both self and other generated content as predictors of socioemotional well being both in sexual and gender minority as well as non SGM adolescents. And this is our Cheater B of well being. There's always got to be a cheater when you come up with things based on letters, right? And this project is led by my collaborator, Dr. Ellen Selkie.

The project that I lead is focused on behavior and our three aims are here. We're interested in how adolescents represent their behaviors, both health behaviors and risk behaviors, on social media.

We rely on social media observation and self report data. In our second aim, we're interested in what they're seeing over the course of the day, when they're interacting. In those little, tiny moments between class or within class, what are they seeing? Are they looking at their own content, at other people's content? What sorts of health content are they seeing? And this aim is being achieved through EMA as well as daily diaries.

And then our third aim, we're collaborating with Dr. Cascio to take content that they've generated as well as other generated content to understand what types of content are influential to them and how the way that their brains process that content is related to their self reported behaviors.

So, I'd like to share a little bit about some of what we've seen in our social media observation. So, we've been interested in the frequency as well as the content of their self generated displays of health behaviors and risk behaviors on social media.

For this part of the project, we are following youth on Facebook, Instagram, TikTok and X. We selected these profiles because they have these affordances of being identity centered platforms. You create a profile. It's an expectation you share content. You share it within a network, and so those affordances were really important to us.

The youth profiles are linked to research team profiles after consent and after them understanding why we're following them and that allows us to conduct a monthly coding procedure where we evaluate for these types of content.

Our code books are generally based in clinical criteria that then are built out so that coders are able to look through content and make reliable yes/no, it's there or it's not, decisions. But I'll share a little in a minute about how we brought youth voice into helping us interpret some of more gray area content.

And for our measures in the health behavior category, we're interested in physical activity, sleep, as well as the SAM post, Sobriety, Abstinence and Moderation; and for risk behaviors, we're evaluating for alcohol and substance use.

So, one of the areas I wanted to share some early findings on is around what we have seen in terms of physical activity. So physical activity in these youth profiles, again, they're 13 to 15, so on the younger level of adolescence, it shows up in a couple of different ways.

On the left, you can see it shows up as youth sharing out output from watches and health devices, showing evidence of behavior that's already taken place. So, this is one way that we see physical activity behavior showing up. It also shows up as youth who are in the middle of doing physical activity at the time the image was captured. So, it's really showing them in the moment and doing an activity.

And then we also see it show up as a consequence or an outcome of that activity.

So, in the far-right image you see a youth who is holding up a trophy for presumably a season end trophy that they have won.

Now, one thing I'll share that has been interesting as we have navigated this space is that we've had to learn about some different sports that not all of us all had experience or knowledge of. And we've also learned some new terms.

So, one thing that came up early in our coding was that youth would often post about sports or an activity and then they would post, "Yes, I got my hardware." And we I was trying to figure out how they were merging a technology club and hockey, or what was going on with all of the posts about hardware.

So, a really integral part to this project and many of the project our team does is we have a Youth Advisory Board where we can take this content. We can unpack it with them. We can allow them to laugh at us and our misinterpretations, and they really help us to understand this is the lingo, this is what it means, this is what it doesn't mean.

And so, through that, I learned that hardware really refers to anybody want to trophies, yes, trophies, medals, any sort of bling you get as a result of sports or a competition. So, lesson learned.

In terms of physical activity, we break it down into three categories based on what we see most often. Sports, Exercise and Fitness, our SEF, category is pretty explanatory. We also see a lot of what we have categorized as recreation. So, this study is taking place in Wisconsin. We see a lot of farm chores. We see a lot of hunting. We see a lot of outdoor activities that wouldn't fall into a sport, but clearly representing physical activity and moving your body.

And we also see a lot of posts around nature where the youth maybe is representing themselves at the top of a large hill. You can't necessarily say, gosh, they're hiking in the moment, but something got them up that hill, and we think about this as when we're evaluating posts that have to do with nature: How did they get there? Did they have to get there on their two legs or two wheels?

And of all the posts that we've coded to date, again at the beginning of year three, of the behaviors, we code about 90% are physical activity. So, youth are really using this platform to share these types of behaviors.

This bar graph shows some of the places within the platforms that it shows up. So, the bio is generally the bio that they provide. It's a very high identity focused area of the platform because they're sharing the three or four things that are most important to them.

Similar, the profile photo, we see a lot of content there. They're choosing this one singular photo to represent the identity of that platform, and sometimes it's them in a sports uniform. Sometimes it's them going for a run.

This is not a rare thing for them to choose physical activity as a part of that profile photo. We see a lot less when it comes to the cover photo. That's kind of the banner photo that often is present on platforms. It doesn't show up there as much. It shows up in highlights, which are often the short reels or kind of more ephemeral content.

Then it also shows up a lot in content that their peers share onto their profile. So, it's a real common thing where people within that peer group can share across each other's content. So that's physical activity.

I also wanted to quickly share a little bit on one of the risk behaviors that we've examined, and I'm going to share a little bit about what we've seen in terms of how alcohol shows up on young teens' profiles.

This is rarely coded. Of all the posts we've coded, only about 3.5%, and that's compared to our early work in college students where it was 50 to 75%. So, this shows up rarely.

But what I want to highlight is it shows up in some really unusual ways compared to kids who are older. We see some posts like the two here where you see a group of friends and they're drinking out of a juice glass or a small glass or sometimes it's actually a shot glass.

It's usually in an environment where you're pretty darned sure that it's not a shot. So, you can see the one on the left, it looks like maybe they're at a hamburger place. There are some cups with straws in them, and it doesn't it's clearly not a bar, and yet they'll label them with these taking shots behavior.

So, one of the things I'm wondering about is whether this is some form of pretending, kind of an acting, trying to act more mature.

We have not seen one of these where it actually looked like they were drinking alcohol. Some of them, their parents are in the photo, which again, in Wisconsin, it doesn't guarantee because our laws are very strange. But it's just a behavior we haven't seen before, and what will be really interesting to us is to see how this evolves over this longitudinal design, and how these types of posts might map onto attitudes, intentions or behaviors.

When we think about where this content is showing up, the data I've shown you is from our first full year of coding. So that's June 2023 to May 2024. And what I showed you was over approximately our first 200 participants, and over that year period we coded about 1100 posts, which looks like a lot, but it also represents that some teens have posted once in a year and other teens have posted 15 or 16 times. So, there's a huge amount of variability at this age range.

In terms of where content is showing up most frequently, it may not be that surprising, but TikTok and Instagram is where a lot of content is shown by teens this age group. Less so on Facebook. Some of the Facebook profiles are family based Facebook profiles. And very, very little on Twitter or X.

So, if I bring this back then to the five Cs of healthy media use, child, content, crowding out and communication, we have done some thought about how our project represents those key elements.

For the child, our project really takes a developmental view of TDM use. We're really centered on the child is the center of data. We collect data from platforms that really, at the end of the day, it's all about that child.

We're interested in both the risks and the benefits to youth of their TDM use, as well as their health and risk behaviors displayed. We're interested in the youth as individuals and looking at between subjects’ differences as well as I'm so excited to be able to look at the arc of teens' experiences as a within subjects.

We've been really mindful about including adolescent perspectives like the hardware story.

For content, we really focus on content, both content that they create, content that they observe.

We see this through our EMA, where they're reporting out what they're looking at in the moment, and we're able to integrate that content into our fMRI tasks.

For calm, Dr. Selkie's project, too, really, has a focus on well being and the ways that the interactions and experiences that youth are having can be beneficial, especially for marginalized communities, and thinking about how teens leverage social media to connect to each other.

For crowding out, we are really interested in physical activity and sleep and understanding whether them posting about sleep has anything to do with their self reports about sleep, what it might indicate when they're talking about sleep on their profiles, and also with so much physical activity display out there, what does it mean for their peers that are viewing that content? How does that influence positively or negatively?

And then for communication, I think a big question for us is how we can bring these findings to teens and families and to the communities that help contribute to this project. And I think that's something we think about a lot in the center, is how do we take that evidence and translate it?

So, I hope that both Dr. Radesky and I have illustrated that TDM research is at a really exciting place right now. We're seeing more longitudinal designs, more data collection methods that push past that traditional paradigm of cross sectional screen time studies.

We're continually thinking and better understanding how to apply child development study design as well as to our interpreting what we learned. And frameworks such as the five Cs can guide provider approaches, family conversations and hopefully be a useful springboard for future study design.

And then as both Dr. Radesky and I have shared, our hope, as co medical directors of the Center of Excellence, that the work and the evidence in this space will be translated to provide better resources for youth and families.

So, we're so appreciative to be here, to be able to talk about our work, talk about the center, and we wanted to make sure to leave some time for questions.

So, I think Dr. Radesky is going to join me up here, and then we are excited to take your questions.

QUESTION: Thank you so much. My question you've alluded to it a little bit, but I really like that you highlighted making sure that you include agency of the youth or the parents, particularly with young children.

How do you think this combined with the idea of youth advisory boards and including youth voices has influenced your work, the kinds of questions you ask, not just getting help interpreting the slides, but how has it influenced what you've been focusing on?

DR. MEGAN MORENO: We've been really fortunate on our team; we have had a Youth Advisory Board for about a decade now. And I am continually impressed in how it impacts the study questions we ask, as well as making sure our methods are feasible.

We tend to have high retention and high completion rates, and I think it's because youth tell us what to do to make it easy for them to be in the study. They also tell us what swag to get and what swag not to get. Very important. No more canvas bags.

And they can also help us think outside of our own frameworks and think outside of our own box and what we expected to find to say, hey, we found these results, what is your interpretation? What do you think this means, and that's been so, so important for us.

So, I think for us it's really been a game changer, and I can't imagine us at this point doing a study without it.

DR. JENNY RADESKY: We have a community advisory board for the RO1 that's part of our PO1, and that's been helpful for things like, are the wording of these end of day text prompts, do they make sense to you, are these the sorts of activities you would typically do with your child, to make sure that they are not reflecting some sort of bias of the academics that are, this is what I think is important in terms of a skill building activity. And then they've also been nice to meet with about emerging topics that may not be the study of ours, but we had this great meeting in June about AI and we're, like, we don't have any results to show you right now, so let's just talk about what's in the news and what you think about this, because that might help inform the way we're approaching future subjects.

They also gave us great feedback on some of our YouTube coding, where we're, like, are we being too precious here in thinking that this influencer's behavior is rude or something? They were, like, no, that's rude. You can think of it as negative role modeling. That's been really helpful just to kind of gut check some of our approaches with families who are experiencing it.

THE MODERATOR: One of the questions online says, what is your perspective integrating AI coaching for supporting children's mental health? Another level of complexity.

DR. JENNY RADESKY: A chatbot, that that would be the interface? We talked about that a little bit earlier today, that there would be a couple important safeguards.

It's a good opportunity for just in time interventions that are tailored to that unique person's way of being in the world because one size fits all parenting guidance or mental health guidance may feel alienating if it doesn't work for you. You're like, oh, this is supposed to work for everyone else, taking deep breaths, and it doesn't work for me.

I think that's the opportunity. We were talking about it more in terms of parenting guidance, that getting something tailored and just in time might be really effective. But with a few important safeguards, it could actually be really important industry standards to role model for commercial products that are trying to create chatbots.

Number one is disclosures. Like, I'm a machine. I'm not a human. Don't create a synthetic relationship with me where you have these psychological vulnerabilities, and you might start to think that I'm your best friend rather than the other people I want you to connect with.

Number two would be data privacy. This is incredibly sensitive data that really can't be shared with third parties or marketers.

Another being safety testing, to make sure it doesn't hallucinate or fabricate, confabulate whatever the word is now, to say something that could potentially be harmful.

I think it requires a lot of humans in the loop, as they say, to make sure that it's and a lot of careful testing. But I don't see I don't know. I would love to hear what other people think about, like, absolutely not or whether what I've heard about is that it really could complement an existing care relationship to make sure that there's other humans or medical systems that are involved in that person's care.

DR. MEGAN MORENO: I think the piece I would add that adolescence is a developmental stage where relationships are so important and learning how to form and navigate and maintain relationships. And so, I think on the one hand there's the worry that would that relationship overshadow other relationships with actual humans.

At the same time, I think one interesting thing to think about for this generation of adolescents is they're already very familiar with para social relationships.

One story I think of is, I spoke at a humanities conference a couple years ago with an influencer. And it's a humanities conference. So, half the audience is guys with elbow patches on their tweed jackets and the other half was all 20 to 25 year olds.

I went and talked to some of them after and said, why are you here? They said, I'm here because I have a para social relationship with an influencer. They just named it and called it out and said I'm here because I feel this relationship.

So, I do think this generation, they're just able to navigate thinking about the types of relationships they have and what's healthy for them.

I think this is an interesting time point to be thinking what would this para social relationship look like for them.

DR. JENNY RADESKY: How to make it safe and not a power differential or anything persuasive in a way that could be that could make I think that just one more thing I wanted to add in terms of training data, is the bias that's contained in a lot of training data and just making sure that this is a tool that would work for a wide diversity of people and their psychological states and backgrounds.

DR. MEGAN MORENO: That's such a great point, makes me want to think you don't want a team to say I want to do the cinnamon challenge. Yes, cinnamon is very healthy for you; you should do that. You need to have some.

DR. JENNY RADESKY: Should I go get some hardware.

[LAUGHTER]

It would be a lot of work, also sorry for keep talking, but there's a lot of hype around, like, M health interventions. This will just get everyone to pick up these healthy behaviors and then realizing that not all of them are effective and trying to make sure it's not just a chatbot hype, that it really is done really carefully with a lot of testing.

THE MODERATOR: To let everyone know, there's another microphone in the room, if anybody would like to ask a question. If not, then we will keep asking questions that are online as well. Do you have a question?

QUESTION: Just very briefly, because you've just touched on it. Rachel Barr, one of the other investigators, a long time ago started studying infants and touch screens and how intrinsically motivating they are, even though there's nothing about them that you would think and it's, like, I don't have an island today because I didn't pay any attention to this, but it's something what is really sometimes the intrinsic motivation of these things for good and bad and then maybe touching on people who have tendencies towards problematic usage and maybe even other pre existing conditions, how much you're seeing those types of things and then meaningful ways of maybe screening or interventions.

DR. JENNY RADESKY: I'll start with thinking about young kids is that kind of the intrinsic it's so satisfying to have cause and effect when you're an infant or toddler, "I did that." "I made it happen." So that's part of it.

The other part is, like, in our study doing educational coding of the designs, there's a lot of extra bells and whistles put on apps, stars, fireworks, everything else, to a child, has such a reinforcement and like satisfying from a sensory perspective, from an agency perspective. And we don't have research to know, are certain kids more susceptible to I'm actually advising on a study in Australia where we're looking at individual differences in executive functioning, for example, and, like, which kids have that kind of attentional drive to seek out those visual stimuli or the reinforcement that comes from it.

But, yeah, we need more research like that. Some of it may come from intrinsic differences. Some of it may be experience based. That's what we thought with our study on emotional reactivity is that, well, if this has been reinforced over time, as if like if I cry and fall on the floor and then I've got a tablet that there might be a behavioral or learned component to it also.

DR. MEGAN MORENO: I'm happy to speak to the adolescent side. I think that the field of problematic Internet use, problematic media use, a horse by the same name but different colors, I think it's been hindered by frameworks. I think if you look at the early studies, it's almost like someone took whiteout and whited out substance use and wrote in media use.

So, I think it's taken a while for the field to really recognize this is something different. Some people have compared it more to disordered eating where it's something very fundamental to your day to day, but you just develop a maladaptive relationship with it.

We did develop a screening tool, the Problematic and Risky Internet Screening Skill that gets to components of risky, compulsive, not just framing around addiction or withdrawal, and we have found that to be really helpful in clinical settings because it really points to certain activities that are being avoided or relationships that are suffering. So, it really is meant to both screen as well as give a direction for what that intervention should look like. That's the problem area.

But what's really interesting to me as well, we've also done some longitudinal studies where we've screened people at six month intervals, and for older adolescents, they can come in and out of risk and self they'll describe the activities they do to kind of get themselves back on track or they work with their friends to figure it out.

So, it's a really interesting phenomenon where a lot of youth are self managing and self managing successfully. That doesn't mean they all are. And we definitely need to understand that proportion of youth who are not able to self regulate, but it seems as though, among many youths, it's accepted you can go into a challenging time and then figure out a way to get out of it.

THE MODERATOR: I think we're just past 3:00, but if I could indulge one more question from online. There's a question, do you see or are there any plans to examine network effects or the effects of self generated displays in nudging similar behavior among network members or users of that social media? So probably more for Dr. Moreno.

DR. MEGAN MORENO: Sure. There have been some really fascinating studies looking at social networks both online and offline and looking at both not only do folks gravitate together in the online space certainly gives a great forum for that.

We've done studies looking at networks of participants on Twitter who share about disordered eating, and they gravitate towards these really centralized networks. And it looks very similar to what you see in the offline world when you study lots of behaviors and how they spread through networks.

What's more challenging is to try to generate enough messaging and communication to form networks around healthy behaviors. That's the bigger challenge.

In the adult world, there is some work on that in the recovery field and some really successful studies of building recovery networks in online spaces. One of the aspects that we're hoping to understand with our B3 study is how positive health behaviors and how those are shared might have influence across youth during that critical developmental time period. So hopefully more info to come.

THE MODERATOR: Thank you very much, both, for a wonderful presentation and very informative research. We appreciate you being here today.

Just to let you know, there were over 344 participants online. So you were not in a sparsely populated room, actually. There was a large virtual presence here today. And we really appreciate you being here. Thank you again.

DR. JENNY RADESKY: Thank you.

DR. MEGAN MORENO: Thank you for organizing this.

THE MODERATOR: Thank you, everybody.