Why the Internet Knows You Better Than You Know Yourself
48:06 Share

This message comes from NPR sponsor Shopify. Start selling with Shopify today. Whether you're a garage entrepreneur or IPO ready, Shopify is the only tool you need to start, run and grow your business without the struggle. Go to shopify.com slash NPR. This is The Pulse, stories about the people and places at the heart of health and science. I'm Maiken Scott.

Sandra Mott is a professor at a prestigious business school in New York City. She's a researcher, a computational social scientist, a psychologist, an author. But in another part of the world, far, far away, she's known for something else. It's a sore point still after so many years. Here's what happened.

Sandra was 15 years old, dating an older boy with a motorcycle. They rode around, Sandra in the back, the boyfriend in the driver's seat. It was super fun until Sandra wanted more. And I, at some point, really wanted to be in the driver's seat. I was like, I've been in the back for long enough. I just want to try this out. She talked her boyfriend into letting her drive while he sat in the back.

She isn't sure exactly what happened, but she must have accelerated way too quickly. And suddenly the bike rises and the front wheel rises up. Boyfriend falls off the back and I just speed away without the slightest idea of how to control the bike. A few seconds in, the bike crashes to the side. Luckily, nobody was hurt, but the bike suffered some damages and they were stuck. Now in the middle of nowhere, and I had to call my dad.

Sandra's father picked them up and they dropped the bike off at the shop. And then... Very quickly, those news spread through the village. A tiny village in Germany with just 500 inhabitants. Which means that the 499 other people knew pretty much everything.

And now they were all abuzz, not just because Sandra illegally drove and crashed a motorcycle, but also because her dad was the local police officer, which made the story even juicier.

The Schadenfreude. And I think for the next months I had to suffer through all of my neighbors not only asking me about the incident, but also judging me based on what I had done. But even when she was not getting into accidents, Sandra felt that the villagers knew her every move. They knew what I was up to on the weekends. They knew who I was dating. They knew what music I was into, what I was hoping to do after school. So they were in my life pretty much daily. ♪

They would also get in her business, suggest things she should do or not do. It was annoying for sure, but also reassuring to have this tight-knit community around her with people who looked out for her. Now that she lives in New York City, Sandra doesn't have to deal with nosy villagers. Her city neighbors don't know her name or her whereabouts.

But there is something else tracking her every move. It's not humans peeking through their blinds. It's the Internet, algorithms that track her and all of us. When you get up in the morning, like the first thing you do, like most people probably, is check your phone. So the fact that your phone has been open gives you a sense that, yeah, OK, this person is up.

now. Maybe you're reading a few news stories, send some text messages. And then you very quickly call Alexa to turn on the music. You get a latte and a bagel at the coffee shop, swipe your credit card. You use your GPS to get to an appointment, post on social media. You search the web. Google searches. I think oftentimes we ask Google questions that are

All of these digital footprints are recorded, sucked in by algorithms. And together, they paint a picture of our lives, our personalities, a psychological profile.

It's a little bit like Sandra's village neighbors. Sometimes it can be helpful, and sometimes it feels like a snoop, keeping up with everything we do and influencing our choices. Sandra has written a book about all of this, Mind Masters, the data-driven science of predicting and changing human behavior. On this episode, we'll talk to Sandra about why algorithms may know us better than we know ourselves. ♪

To get started, let's think for a minute about the mind-boggling amount of data we generate every day. The sheer amount of data that is generated is absolutely insane. So I think the average person generates about six gigabytes of data every hour. And that adds up very quickly. The comparison that I like is that there's almost as many data pieces out there now that there are stars in the universe. And that's just mind-blowing. It's absolutely nuts.

Again, this is not just information that you put out there on your social media profiles. It's your digital footprints, like your GPS location data. So there's all of these small traces that in and by themselves, they're actually not even that intrusive, right? Getting your longitude and latitude from your phone doesn't even feel that intimate. But then if I tell you, yeah, no, but I actually know where you live. I know where you work. I know that you two are dating because...

You know, every Friday night, the two phones show up in the same location. And suddenly these small traces also become big in terms of the insights that they can generate. Sandra says all of this information put together creates a very detailed picture of us, way more comprehensive than what humans can observe about each other.

In the book, you describe how you first met your future husband and how you ended up going to his apartment and you looked around and you kind of realized a bunch of things about this man. What did you see and what did you realize about him?

And then how is that different from what algorithms can gather about us? Yeah, it's funny because I think we're all master snoopers all the time, right? So we constantly make these inferences about who someone is based on their looks, based on their physical space. So the example that you gave with my husband is...

I came into his apartment and it was meticulously organized. So a huge library on the side, the books were all organized and sorted by height and by color. They were perfectly aligned in the front. All of the cutlery in the drawers were just like was perfectly aligned. All of the glasses that he had, no watermarks, nothing. So it was just like this meticulous space.

And which also signal a lot of curiosity because they were like books and many different languages and art. And so one of the things that we know actually from psychological research is that humans are really good at making these inferences. So if you give me access to your office or to your bedroom, I can make inferences about whether you might be more organized, disorganized, more open minded or traditional, more extroverted, introverted based on both.

The signals that you intentionally put out there right in your office, you might have a poster with an inspirational quote, and that tells me something about what you care about. But then also, if I see your waistband overflowing, I probably have a sense that, yeah, maybe you're not the most organized person. So

The same way that we as humans make these judgments based on what we observe, the behavior that we observe and the traces that we observe, computers in a way operate in a similar space. So they observe all of these digital traces. And then over time, they learn with trial and error of like, here's what someone does online. Here's how this relates to their psychology. Now,

The big difference comes in, in just a sheer amount of data that algorithms can process, right? I might observe the behavior of maybe like a few hundred, few thousands people over the course of my life, and I get better at kind of relating, here's what they do and how that relates to who they are.

But computers can observe the behavior of millions of people at the same time. So they just pick up on these very subtle cues and signals that we as like human Sherlock Holmes might actually overlook. So that's the really big difference.

And how do companies get a hold of these data points? Because I'm kind of probably naively assuming that much of this data lives in different universes and is protected by firm borders, but that's probably not true, right? Right.

Unfortunately not. To be honest, a lot of it just lives online with like public access, right? If you think about the stuff that people post on social media, a lot of those traces just sit there publicly available for anyone to scrape. That's true for like social media. That's true for LinkedIn. But then there's also all of these traces that we just oftentimes mindlessly sign away.

We give apps access to our contacts, our microphones, our cameras, our photo galleries, ready to be scraped and used to create a psychological profile. Give us some more examples of, you know, pictures people might pose and then how companies use those pictures. You have a table in the book and some of the things are like beer pong or cheerleading and those kinds of things. What do they translate into?

Yeah. I mean, I think for me, the interesting part as a psychologist is that you both get relationships that are pretty obvious, right? So if you look at what predicts whether you might be more extroverted, for example, those are like people post about social activities, going out on the weekend, beer pong, cheerleading. And you can really see like the social nature of these activities change.

Whereas people who are more introverted, they post about reading, they post about their love for manga and anime, using their computers and browsing the internet. So all of the stuff that you typically do by yourself. The same is true for what people purchase. You see extroverts spending money on going out to bars and clubs. Introverts are more likely to buy books or maybe gardening tools.

So companies use all of this data that emerges, right? It makes a picture of me, of who I am. And it's a shockingly exact picture, sometimes perhaps more exact than what I know about myself or what my friends know about myself.

Yeah. And I think you can think about this picture on different levels. It's been the case for a long time that companies have been using data to understand, well, based on all of these past behaviors that we observe, maybe that's what you've previously bought. Maybe that's the websites that you've visited. There's an understanding that we have now of here's your general preferences. Here's your routines. Here's maybe your needs.

And we can project that into the future and say, well, if you've been interested in like photography in the past, maybe you'd also be interested in this camera that we have to offer right now. I think one of the lessons that we've learned over the last, I would say, really 10 years is that all of these traces, when you combine them, also provide this very rich picture of not just what you do, but also really who you are when it comes to your psychology, anywhere from your personal values, your personality, your

IQ, life satisfaction, mental health. And for me, the reason for why this is so interesting, but also creepy is because we can make sense of it, right? So the moment that I tell you, well, I can take your purchases, and I can learn whether you might be impulsive, neurotic, extroverted, open minded, suddenly, it feels a lot more intimate to people than just saying, well, those are your purchases.

And you can say that it in a way allows people to understand how intimate their data is. It also offers quite a few opportunities for companies to use it, right? It's not just that I'm purely predicting here's the camera that you might be interested in in the future because you showed an interest in photography before, but I can also now use it to change the way in which I communicate, right? And if you go back to the village, this is such a fundamental thing

human tendency and in a way skill that we oftentimes don't even realize it. But you never would talk to a three-year-old the same way that you talk to your mom or you talk to your boss or you talk to your spouse. We usually understand that people on the other side, our communication partners, they did have different needs, they have different preferences, and we adjust the way that we talk to who's on the other side. And so understanding people's psychology is

based on data actually gives companies a lot of power over just in communication, right? So if you're a beauty retailer, which is one of the projects that we studied at some point, and you try to sell your beauty products to extroverts or introverts,

you might actually highlight different things, right? Extroverts might be buying beauty products 'cause they wanna be the center of attention, they wanna stand out from the crowd when they go out on a Friday night, and that's why they buy it, 'cause that really taps into their psychology. And then introverts, they don't necessarily need the shiny being the center of attention for them, it's much more about making the most of the quiet me time that they have. So the moment that you understand psychology beyond just these individual data points,

The more power you get over adjusting the way that you communicate with whoever is on the other side. There have been studies about how well the Internet, quote, knows somebody as opposed to their spouse. What did they find?

Yeah, I think this was one of the most eye-opening studies out there, which also made it so much more tangible for people who don't live in this world of correlation and statistics. So this was actually an idea that one of my colleagues at the Psychometric Center at Cambridge had when we were both getting our PhD there. And what she said is like, it's so hard to wrap our head around

Like how good are computers at making these predictions, right? If I give a computer access to, say, all of the pages that you follow on Facebook, this is what she was looking into in her study. I give the computer access to all of your Facebook pages and then I have it make predictions about your big five personality traits. So how open-minded are you, how conscientious, how extroverted, how trusting and caring and how neurotic.

how good is the computer? And again, like as psychologists, we live in these world of correlations where it's like, okay, we're going to get a predictive accuracy of a 0.5, but that doesn't mean anything to most people. We have no idea if that's good or bad. So what she says, like, why don't we compare the accuracy of the computer to the accuracy of people who

in our environment who should actually know us pretty well. That includes co-workers, that includes friends, family members, and eventually spouses. So you can imagine if you were a participant in the study, you complete a personality questionnaire. So you tell us how do you think of yourself when it comes to, again, being open-minded, extroverted, and so on. Then we pit you against a computer. So we have the computer try to predict what

how you think of yourself. And we also invite people like your friends, family members, coworkers, and your spouse to fill in the same questionnaire on your behalf. And now we can have this direct comparison of saying, okay, how good are the people in your environment?

compared to the computer. And what she found, and this is now almost 10 years ago now, so the models by now are becoming a lot better and we're collecting a lot more data. But what she showed is back in the day, the computer was already better than co-workers by far, much better than friends.

Also much better than family members. And again, those include siblings, includes parents, includes kids. So people who spend a substantial amount of time with us and yet a computer just by looking at your Facebook pages was already better at predicting your personality than those people.

And it was slightly, slightly worse than the spouse with access to the data that it had back in the day. But again, if you imagine the computer now also getting access to your credit card spending, your browsing histories, your smartphone sensing data, and us having much better predictive models, by now the computer would probably be better than even the spouse. And maybe even better than we ourselves are. Because, you know, I was also thinking every time I'm out in the world,

Even with the people who know me best, I'm presenting myself in a certain way. You know, I'm not completely unfiltered ever. And maybe the only time when I'm mostly unfiltered is when I'm looking around on the Internet. Right.

Yeah, it's amazing, right? It has this, in a way, it's surprising and not surprising because the Internet offers this feeling of anonymity and safety. Google is a prime example. We asked Google questions that, again, you probably wouldn't feel comfortable asking your closest friend or even spouse because we're so worried about their judgment. And yet we feel comfortable typing it into a search bar online.

And that search query goes to a server, right, and it's going to sit there for Google to access. And who knows who at some point will have access to the same data. So I think there's this false sense of security that we have as we interact with technology that in many cases really isn't warranted. Of course, we also benefit from all of the data that's being collected about us.

from all of our apps, knowing our preferences. Those are benefits we may not even notice anymore. It's funny because oftentimes these benefits are so obvious, right? That I think we don't even realize. And for me, the mere fact that there's so much information out there, right? So there's like, if you were to,

watch all of the content on YouTube, it would take you over, I think at some point I calculated, it would take you over 2,000 lifetimes if the average person lives to 80 years. So there's just absolutely no way that we can go through all of the products on Amazon, all of the movies on Netflix, all of the websites that are indexed on the internet, right? That's kind of what Google helps us with. So

we already benefit in a way from those companies understanding what is it that we might be looking for. There's like this funny saying that people get outraged if you make them go to page two on Google searches to find what you're looking for, right? In most of the cases, we actually very much benefit from the fact that someone has a rough understanding of what we're looking for and then gets that stuff to the top of the list. So I think even like we call it the paradox of choice. There's just so much

Sandra Matz is the author of Mindmasters, the data-driven science of predicting and changing human behavior.

Coming up, we'll hear more of my conversation with Sandra and we'll discuss how psychological targeting could help rewrite health messaging. Just the mere fact that I can observe your behavior over time in a very granular way allows me to say, well, actually, maybe I can passively observe that you're not leaving the house as much anymore as you used to. And maybe you're not taking or making as many calls. So there's much less social engagement.

That's next on The Pulse.

This message comes from Thrive Market. The food industry is a multi-billion dollar industry, but not everything on the shelf is made with your health in mind.

At Thrive Market, they go beyond the standards, curating the highest quality products for you and your family while focusing on organic first and restricting more than 1,000 harmful ingredients. All shipped to your door. Shop at a grocery store that actually cares for your health at thrivemarket.com slash podcast for 30% off your first order plus a $60 free gift.

This is The Pulse. I'm Maiken Scott. We're exploring how our personal data is being mined by companies to predict and influence behavior.

My guest is Sandra Matz. She is a computational social scientist at Columbia Business School. Her new book, Mind Masters, examines how all of the digital traces we leave behind, from emails to stuff we buy to things we search for online, can create a very accurate profile of not just what we do, but who we are.

Sandra grew up in a small village in Germany, and sometimes she compares the algorithms that collect and analyze our data to her nosy village neighbors. It's pretty clear how all of this could benefit companies that could market things to me in a much more efficient way. But how could this benefit us?

Yeah, it's one of these questions that I've been thinking about a lot. And it's actually, it's motivated by this way that I related to me growing up in the village, right? Because for a long time, I remember I felt that...

It was purely manipulative. So I remember starting this research and thinking, oh, the moment I can understand someone, that just gives me a means of manipulating their behavior. And then I thought back, it's like, no, there's actually something really beautiful about understanding someone because it can also help me

support them with the goals that they have. So I've been thinking about it a lot in these what-if questions. So if I can use what I call psychological targeting, so taking data, predicting who you are, and then using it to influence your behavior, if I can use that, for example, to get you to buy a certain product,

Can I also use it to accomplish exactly the opposite, which is like helping you save? Saving is like one of these behaviors that is so difficult for us to do because we have to give up something in the here and now for maybe this potential benefit in the future. But we also know that it's

really, really important. So especially in the US, the picture looks really grim. If you look at the population, I think 50% live paycheck to paycheck, 10% couldn't even go a week without being paid. That's disastrous, right? The only thing that needs to happen is your car breaks down, you can't bring it to the shop because you don't have the money, you can't get to work, so you might lose the job. And it's like this very quick spiral downward. And yet still, like it's usually those people who have the hardest time to save.

Sandra wanted to examine what would happen if she created very specific messages based on people's personality traits to nudge them to save money.

If I know that you're agreeable, so this is like one of the personality traits that says whether you are kind of more trusting and caring. So people who really care deeply about their social relationships, for those people, it's not necessarily going to convince them to put an extra dollar in their bank account if I tell them, well, it's just like you're going to accumulate some savings that's going to get you ahead in life.

What really motivates them is to say, well, look, saving today might be hard, but it's the only way that you can actually protect your loved ones in the here and now and also in the future. So it's really playing into their motivation to maintain these positive relationships and protect people that they care about.

The people in her study had less than $100 in savings, and the goal was to encourage them to double that over the course of four weeks. And the personality-based targeted messages really had an impact.

And what we saw is that if we tap into people's psychology, we can actually increase the number of people who managed to hit that goal by about 60%. So like a pretty substantial increase in people, in this case, accomplishing their goals.

And I could see this also coming into play with messaging around health. You know, let's say you're trying to help people who are at risk for diabetes. You could probably figure out how to best message them and how to frame that message to get them to change their behaviors.

Yeah. And it all comes back for me is are we supporting someone in a way that helps them accomplish the goals that they set for themselves, right? For me, that's in a way the distinction between the manipulation part and the no, this is actually helpful support. And you're absolutely right is that I think the health context is one of these contexts where data and technology has a lot to offer, not just in terms of treatment, right? So I think we now know that there's a

a lot of applications that try to at least complement therapy. That way you don't talk to a human therapist anymore, you talk to an artificial therapist. I don't think this is going to replace human therapists anytime soon, but it can certainly be a useful tool for the people that otherwise don't have access to

And that's a lot of people. So if you look worldwide, I think for every 100,000 people looking for therapy or mental health support, there's 13 professionals, 13 human professionals. So there's this huge gap in supply and demand where even though AI therapists might not be perfect, and maybe they're not quite on par with human therapists right now, but if the counterfactual is you don't get any support, then they're certainly better than that. And there's also something to be said about

tracking right so it's not even just a I can help you with the messaging but there's also like just a mere fact that I can observe your behavior over time in a very granular way allows me to say well actually I

Maybe I can passively observe that you're not leaving the house as much anymore as you used to. And maybe you're not taking or making as many calls. So there's much less social engagement. And it could be nothing. Maybe you're just on vacation and you're having some downtime. But it could also be like this early warning system, especially for people who have a history of suffering from depression, for example, that allows them to detect these deviations early before they enter a full depressive episode and then have to

get out of their way to see a therapist to get a real diagnosis. So I think there's a lot of really fundamental opportunities that we have when it comes to data in just terms of better tracking and treating mental health.

Algorithms can collect and evaluate so much more data than humans ever could. And of course, we all generate way more data than ever before. Sandra says this can lead to surprising insights, things we might have overlooked.

And one of my favorite examples in that space is the use of first person pronouns. So that is I, me, myself. And I remember there was this conference, like a room full of psychologists. And one of the presenters, Jamie Pennebaker, who was a pioneer in this field of studying language as a window into psychology, said,

He raised the questions to the room and said, like, hey, there's this finding that we see time and again, which is an association between the use of these first person pronouns and a psychological characteristics that you all know. What do you think it is?

And I remember people at the table talking, and we all were sure that it was going to be narcissism. Because if you talk about yourself, you're probably just so self-focused that you want everybody else to also pay attention to how amazing you are. Turns out that it's actually emotional distress.

So if you talk more about yourself, that's not necessarily a sign of you being a narcissist. That's a sign of you potentially suffering from anxiety, from depression. And when you take a step back and think about the relationship, even though none of us had the intuition that this was going to be going on, it actually makes sense. Because if you think back to the last time that you felt somewhat blue or the

pressed or down, you probably were not thinking about how to solve the world's biggest problems, right? You were not trying to figure out how to fix climate change. What you were thinking about is, why am I feeling so bad? Like, what could I do to get better? Am I ever actually going to get better? So the focus on the self and this inner monologue that we have with ourselves just creeps into the language that we use when we talk to other people, when we post on social media. And so

There's all of these new relationships between someone's online behavior and their psychology that we can learn just because algorithms and computers can observe so much more data at any given point in time than humans could. So it seems like all of this data mining could have some benefits. And I wanted to know, how can we make this system work better for us, where we get more out of our data being collected? Something helpful. ♪

It's a great question. And I should say that I've actually become quite a lot more pessimistic over the years. I think if you had asked me five years ago, I would have said, well, if we only if we increase education and we tell people here's what's happening with their data. So increased transparency through regulation, increased control through regulation and just literacy through through education.

people are going to figure it out themselves, right? I was a lot more optimistic in saying, well, at that point, people, when they download an app, they will know which permissions to untick. They will understand. But just observing my own behavior, right? One of these questions that I oftentimes get is like, well, you live and breathe this data-driven, in a way, kind of manipulation, whatever you want to call it. What do you do to protect yourself and to make it work for you rather than against you?

And I have a hard time. So I even kind of keeping...

track of how technology develops, it's an impossible task, right? Even if I understand the technology, it would be a full-time job to successfully manage it. And I would much rather share a meal with my family or meet friends or even kind of go to work and work on something meaningful research-wise than just going through all of the terms and conditions and managing all of the permissions meticulously. So I think this idea that we can just kind of push the control to consumers, I think, is a tough gamble.

But Sandra says there are some big picture changes that could happen. So for me, what I think we can actually do is we can ask for more. And that is true from policymakers. I think what I want policymakers to focus on is like make it easier for us to do the right thing. If we're so constrained in time and cognitive mental energy,

I can't be expected to have to opt out of tracking from all of the products and services that I'm using. Change the default to opt in. And now I can decide whether I actually believe that the value that the companies create for me is so good and is so powerful that I'm happy to share my data. So that is one thing. And the other thing is that

I think we should be pushing a lot more collectively for these new types of technologies that in a way eliminate the trade-off for us. So because if the trade-off for users like ourselves is I can either protect my privacy, I can protect my data, I can protect my ability to make my own choices, as opposed to I get all of the service and convenience that comes from companies using my data, let's say Google Maps getting me from A to B, my phone recognizing my speech.

then I think that the human brain is always going to gravitate towards this immediate reward and personalization and benefits and convenience. But there are technologies where you can actually get the same intelligence, the same service without having to give away and sign away your data. So what I'm thinking about specifically is something that's called federated learning, and it's how Apple actually trains Siri on an iPhone. So instead of

Apple, grabbing all of your data, sending it to a central server where they process it and they train their models, they get better at understanding language. What they do is they send their intelligence to your phone. So they send the model that they've trained, they send it to your phone because your phone is a supercomputer, right? Your phone is so much more powerful than the computers that we used to launch rockets to space with a few decades ago.

So you don't need to send all of your data. They send the model. It locally updates. So it still learns how you speak and how to respond, sends back the intelligence. But the data itself, so what you talk about, the actual conversation, they never leave your phone. So they never leave the safe harbor. And for me, that's a complete game changer, right? Because now suddenly we're

placing you having to trust a company that they're doing the right thing with a somewhat much more trustworthy system where even if they were to exploit it, they just don't have the data. So again, long answer, but I just feel like

Like, I don't want people to have the feeling that they should be the ones managing everything by themselves. I think what we should do is we should ask both policymakers to do a better job and we should expect companies to do so and reward the ones who say, no, we can give you the same service and convenience. But we also care fundamentally about protecting your privacy. What do you want people to take away from this book?

Yeah, a couple of things. But I think the main one is that it was really meant to tell this human story of data and say, look, data isn't just these individual pieces of what you post on social media or what you buy with your credit card or your GPS location. It's really this individual.

very, very intimate window into who you are, like your motivations, your needs, your preferences, hopes, dreams, and so on. And that can be extremely powerful in shaping your behavior for better and worse. And in a way, I think it's up to us to say, we need to ask more, again, from policymakers, from companies,

in how we're using the data. Because right now, I think that it's a very tilted playing field, right? There's like companies and big players that very much benefit from the data that we generate, but most of us don't really. And we talked about some of these really amazing use cases where we could benefit with better mental health care. But for that, we first of all have to understand it. And we also just have to

Sandra Matz is an associate professor at Columbia Business School in New York City and the author of Mind Masters, the data-driven science of predicting and changing human behavior.

You're listening to The Pulse. I'm Maiken Scott. You can find us wherever you get your podcasts. Also, subscribe to our newsletter to stay in touch with us and to find out what's happening on the show. Every week, I'll send you a recap of favorite moments from the latest episode, a preview of what's ahead, and there'll be ways to participate in upcoming episodes. To sign up, go to whyy.org slash The Pulse Newsletter. ♪

Coming up, how all of this internet tracking and sharing has affected our emotional well-being. It was where people started to kind of have that aspect of like, oh, this happened and I wasn't invited. Or I saw that my friend said that they were busy this weekend or said that they couldn't do something this weekend. And then I saw pictures of them posted that they were at a party. That's next on The Pulse. ♪

This message comes from Messina Touring Group, presenting The Lumineers in Washington, D.C., with a live performance at Nationals Park on Saturday, September 6th. Tickets available now at thelumineers.com. This is The Pulse. I'm Maiken Scott. We're talking about psychological targeting that happens online and how the Internet knows more about us than even the closest people in our lives.

So much of this feels totally commonplace and normal now, but it's all still pretty new. And in the early days of social media and our new online existence, we were much less used to sharing so much about ourselves and so much info about us being out there for the whole world to see. None of it felt right, and we could tell that it had an impact on our emotional well-being.

Liz Tang takes us back to the early days of Facebook, which started over 20 years ago, and how it changed the college experience.

21-year-old Mark Zuckerberg turned his little web experiment into a very big business. It's called thefacebook.com, and it's like no college directory you've ever seen. For example, you can take a person and look them up, and you can get a picture of how they'd like to portray themselves, in addition to, say, the courses that they're taking, sports teams, clubs that they're a part of, and who their friends are at the school. But I think that there's a lot of times where people just go out and meet

I remember the first time I heard about Facebook. It was the fall of 2004, and my high school friends were home from college for Thanksgiving. They told me there was this new social networking site even better than MySpace, and absolutely everyone was on it. I unfortunately couldn't yet join because I was taking a year off and didn't have a .edu email address. Back then, only college students could join.

But that just added to Facebook's mystique. It was like this secret club just for people our age. No adults allowed. So do you remember when Facebook first came on the scene? Oh, yeah. No, I distinctly remember it. This is Marcus Hotaling, director of the Counseling Center at Union College in upstate New York and president of the Association for University and College Counseling Center Directors.

And prior to that, it was MySpace. But it was really interesting because MySpace really wasn't... It was just kind of a space... It was MySpace. People would put out music videos and music and...

It wasn't, as I would argue, there wasn't the emotional connection that Facebook and now other apps have had. I remember those early Facebook days once I was able to join. It was exciting being able to connect with people outside your immediate circle, getting the chance to secretly stalk your crushes, feeling like you were part of something. But as time went on, Facebook began to morph. The first big change came in 2006 with the news feed.

The ability to see what other people were doing, what pictures they'd posted, what events they attended, whose profile they'd posted on. It was so controversial in the beginning that hundreds of thousands of students across the country signed a petition against it, calling it, quote, creepy and, quote, stalker-esque.

Here's how Danigal Young, a psychology professor at the University of Delaware, recalls the uproar. I remember the first day that I walked into my college class in 2006.

And my students were all, it was a small class, 20 students, and they were all freaking out. I'm like, what is the matter? And they said, this is so weird. You know, we're on the Facebook and it's like sharing our online behaviors with everyone that we're connected to. It was the introduction of the newsfeed. And they were like, this is creepy and weird. It's telling everyone that I broke up with my boyfriend or whatever. So there were these little moments when Facebook was sort of

Just experimenting with what level of disclosure are people comfortable with? And every time people were kind of adapted. They're like, "Oh, I guess this is cool. This is what makes it interesting."

It's around that time that Marcus Hotaling remembers starting to hear different kinds of complaints from students during counseling sessions. It was where people started to kind of have that aspect of like, oh, this happened and I wasn't invited. Or I saw that my friend said that they were busy this weekend or said that they couldn't do something this weekend. And then I saw pictures of them posted that they were at a party. And that was just the beginning.

Richard Shattuck, who directs the Counseling Center at Pace University in New York City, says things only got worse with the advent of smartphones. Then it became a real problem.

And I think one of the interesting things is although FOMO was coined before social media, it became a very popular term because of social media. And that led to students spending a lot of time online. In 2009, Facebook introduced one of its biggest and maybe its most addictive features, the Like button.

And then when the economy of likes came about, you know, the currency, it started to take a turn, and many might say for the worse, because it became a currency that you, as a student, wanted lots of.

The more likes, the more looks, if you will, the more attention one's pictures get, seemingly the more popular you are. And I think that led to this phenomena that we're seeing in which college students are spending, not all, but some are spending an inordinate amount of time

on social media, not just scrolling, but also posting and providing a lot of fresh content to keep the attention up.

It seemed like each new feature brought new problems. I did an informal poll of some friends, and they reminded me of all kinds of issues I'd forgotten, like how Facebook Messenger led to communiques from creepy strangers. Like, who is this DMing me? So then I get to campus, and he started, like...

semi like stalking me. So I feel like in the beginning, people used that as like a way to kind of like show face and like make a connection, but it definitely got creepy quickly, at least with a few people. - There was the close friends feature rolled out in 2011.

And I just remember being so hurt one time because one of my friends, I remember looking at his page and not seeing myself on his top friends list. And I was like, what the heck? Like, we've been like best friends for so long. And that sort of was the start of a regression in our friendship, which I felt like,

Facebook ultimately drive a wedge through a lot of friendships that I had. Bullying and mean girl stuff, not to mention teenage drama. There are people who would just be going back and forth. And it was also that time where people were posting things you couldn't really tell some of their inflections or the tone of their voices. And that was all just happening on Facebook publicly.

for everybody and everyone's chiming in with their own comments or liking things or like just a ridiculous and that like fueled the fire even more and then these posts exists you know

As Facebook expanded its features and its reach, Richard and Marcus say the social media-related angst students were reporting expanded and deepened. It became a 360-degree immersive experience of all anxiety all the time, making people feel lonely, excluded, and inadequate. Other people always get at least 20 likes on their posts, and I'm only getting 10. Why don't people like me?"

Or, "Everyone's getting amazing internships for the summer and I haven't even started looking yet. I suck." Or simply the steady stream of photos of people looking their best, as the student in question sat clicking through pictures, eating chips in their sweatpants. Of course, these days, Facebook isn't nearly as popular among college students.

But its descendants, Instagram, TikTok, Snapchat, have continued and grown the pressure to always be on and to broadcast your life.

or at least the best parts of it. I think it's kind of morphed into a, you know, a narcissistic, like, hey, look at me all the time. Look at what I'm doing. I had a student once who said to me, and I still laugh at this, she goes, we, it was a Halloween, and she's like, we took probably 250 pictures before all of us decided on a picture that we all agreed on posting. She's like, we were fighting, there were tears, and

There were curse words. And she's like, and when we finally all agreed on one picture, she goes, the statement underneath was BFFs, friends forever. And she's like, and what a crock that that was because we were just fighting about this. So I really think that when it's used for that or when it's used to put on this air of like, look how good my life is.

You know, I've talked to students and I've seen people that I know who post things that I know behind the scenes. Once you kind of like that Wizard of Oz, once you pull the curtain back, it doesn't look as good as it's looking on social media. So if you're putting it up there to put this air of like, look how great things are and look how great my life is. That's where it becomes a problem for me. I asked both Richard and Marcus if their students seem to be aware of the effects social media is having on their lives and mental health.

Richard said a lot of the students he talks to think the problems they're bringing in are interpersonal, not tech-related. After all, he said, social media for them is like water for fish. They've never known life without it.

But Marcus has a different perspective. He teaches a class on social media and mental health that encourages students to take a critical look at how social media affects them. I think they recognize the role that social media plays in their lives. Like sometimes it makes them feel good. Sometimes it makes them feel really cruddy. So I think that they're aware that it plays a role in their lives, in the emotional ups and downs.

And I think that there's the desire to kind of learn about like, you know, the feedback that I get is like, you know, I'm going to try to make some changes in my life. I'm not going to eliminate it. And that's never been my MO is to say you need to get rid of it. It's like you need to look at how you use it.

And like, what role do you want it to be in your life? Because the fact is, Richard and Marcus both told me, social media isn't going away. And college students today are much savvier consumers than my classmates and I were back in 2004.

And for all the concerns we hear in the media about social media, research still hasn't conclusively determined whether social media brings more bad than it does good to our mental health. We have always vilified new things. Like, you know, if we go back to the 80s, I remember when we had to start putting the warning label on music, like, oh, parental advisory and video games make people violent.

I think it's the same thing with this and social media. Like, it isn't just social media that we can't just vilify the tool. We have to, again, look at how the person who's engaging with it. It's like alcohol or any kind of substance. Like, we can't go back to prohibition. It's how you choose to use it. That story was reported by Liz Tongue.

That's our show for this week. The Pulse is a production of WHYY in Philadelphia, made possible with support from our founding sponsor, the Sutherland family, and the Commonwealth Fund. You can follow us wherever you get your podcasts. Our health and science reporters are Alan Yu and Liz Tong. Our intern is Christina Brown. Charlie Kyer is our engineer. Our producers are Nicole Curry and Lindsay Lazarski.

I'm Maiken Scott. Thank you for listening.

This message comes from Sotva. Sotva luxury mattresses are made in America by expert craftsmen using the highest quality materials so that your mattress will provide comfortable sleep for years and years. Sotva mattresses are always delivered to your home and set up in the room of your choice. They're never folded and never squeezed into a small box. Visit s-double-a-t-v-a-dot-com-slash-n-p-r where NPR listeners save an additional $200.

This message comes from Capella University. With Capella's FlexPath learning format, you can set your own deadlines and learn on your schedule. A different future is closer than you think with Capella University. Learn more at capella.edu.