Proof That Early Social Media Access Drains Learning?

Transcript

The full word-for-word transcript of this episode. Plain text version

Naomi: If you really want to protect a teenager's high school maths score, you, ah, you shouldn't actually be looking at how many hours they spend scrolling through TikTok or Instagram.

Jules: Right, which is what absolutely everyone does.

Naomi: Exactly. But you really need a time machine to look at the exact day they opened their very first account. I mean, the familiar debate around children and screens usually pitches two very loud sides against one another.

Jules: Right, you have the side insisting social media is completely destroying teenage mental health.

Naomi: Well, yes, and then you have the sceptics who rightly point out that the actual clinical evidence for the mental health side is incredibly thin. It's messy, and it's just full of contradictions.

Jules: It really is.

Naomi: But, so there's this massive longitudinal Italian study that was published in August 2026, and it isolates something that almost the entire debate just skips over entirely.

Jules: Which is fascinating because it doesn't measure screen time at all, does it?

Naomi: No, it measures the exact age of account creation.

Jules: Huh.

Naomi: And, crucially, it tracks learning, not mood.

Jules: So, our mission for this deep dive is to unpack what looks like an unusually strong causal case here. This idea that opening a social media account before ninth grade actively drags down a student's learning capacity.

Naomi: Which is a huge claim.

Jules: It's massive. So, we are going to stress test these findings, look closely at the potential flaws, ah, the human element behind the data, and ultimately deliver a verdict on whether this research actually helps or hinders our understanding of kids and screens.

Naomi: And as you listen to this, I want you to consider your own assumptions about giving a child access to social media. Because if you are sitting there thinking that delaying access just sort of postpones the inevitable harm...

Jules: Like they're just going to crash anyway when they finally get the app.

Naomi: Right, exactly. If you think that, the data we are about to explore is going to challenge that premise completely.

Jules: Because the mental health argument is definitely the one that commands the most public attention. I mean, whenever we talk about smartphones and teenagers, the immediate pivot is always towards anxiety and depression.

Naomi: Always.

Jules: But the data on that front is remarkably murky.

Naomi: It is incredibly messy. And, you know, on the topic of mental health, the sceptics have a very valid point. You can find studies showing slight dips in well-being, sure.

Jules: Yeah, but then others showing absolutely no effect at all.

Naomi: Precisely. And some even demonstrating that digital connection actually helps certain marginalised teenagers find communities. So, my usual move here applies perfectly. The conventional advice tells us we should focus heavily on tracking a child's mood.

Jules: Right.

Naomi: But what this August 2026 research actually found was that the harm of early social media exposure shows up most clearly and most measurably in test scores.

Jules: Which is a brilliant pivot for researchers to make, honestly. I mean, when you stop trying to measure a subjective, highly fluctuating feeling like anxiety...

Naomi: Which is so hard to quantify.

Jules: It's nearly impossible. But when you start looking at hard academic performance over time, the picture really sharpens. But I have to ask, linking a downloaded app to a bad algebra grade sounds like quite a leap.

Naomi: It does.

Jules: How do we know the kids downloading the apps early were not just, well, struggling students to begin with?

Naomi: And that is the most important question to ask. It really is why the methodology of this study is its greatest strength. So, the researchers took a sample of 5,227 students across northern Italy.

Jules: That is a very solid sample size.

Naomi: It is. And to isolate the effect of social media, they used this statistical approach called difference-in-differences.

Jules: Okay, let us build a mental model for that because difference-in-differences can sound like, well, impenetrable academic jargon.

Naomi: Fair enough. So, imagine finding two identical academic twins in grade five.

Jules: Okay, grade five.

Naomi: Right. They are both scoring the exact same marks in maths and Italian. They come from the exact same socioeconomic background, live in the same type of household, and, ah, both even had similar early technology habits.

Jules: Like maybe using a tablet in primary school, that sort of thing.

Naomi: Exactly that. They are perfectly matched. The only difference is that one student downloads their first personal social media account in grade six, and the other student waits until grade nine. So, difference-in-differences lets us watch their academic paths diverge in real time, starting from that perfectly level playing field.

Jules: So, we are essentially establishing a pristine baseline before any personal social media accounts even exist.

Naomi: Yes, exactly. Both groups are performing exactly the same academically. There is absolutely no pre-existing disadvantage clouding the results.

Jules: Precisely.

Naomi: From that grade five baseline, the researchers track their INVALSI test scores, which are the standardised national tests used in Italy, all the way through to grade eight and grade ten.

Jules: And what happened?

Naomi: The data reveals sustained, undeniable falls in those scores for the early adopters. For students opening an account in grade six, by the time they hit grade eight, they scored 0.22 standard deviations lower in Italian.

Jules: Wow.

Naomi: And 0.18 standard deviations lower in maths. And then, by grade ten, that maths gap widened even further to 0.27.

Jules: Ah, standard deviations can feel quite abstract to just hear out loud. So, let us translate that into what it means for a student sitting in a classroom on an ordinary Tuesday.

Naomi: Please do.

Jules: A drop of 0.2 standard deviations is an enormous penalty in education research. I mean, governments often spend billions on interventions hoping to move the needle by just 0.05.

Naomi: It's true. It is notoriously hard to move.

Jules: Right. So, the authors of this study interpret a 0.2 standard deviation gap as being roughly equivalent to six months of lost school.

Naomi: Exactly.

Jules: A child who opens an Instagram account in their first year of middle school is essentially missing half a year of education by the time they sit their high school exams.

Naomi: Just compared directly to that academic twin who waited.

Jules: Yes. To contextualise just how massive that is, I mean, think about the social divide in education. The gap between a student with two university-educated parents and a student whose parents do not have a high school diploma is about 0.7 standard deviations.

Naomi: Right.

Jules: So, this social media penalty alone accounts for about 30% of one of the most stubborn inequalities in the entire education system.

Naomi: Which is staggering. And it brings up a fascinating, well, counter-intuitive twist in the findings. Because you might assume that once the late adopters finally open their accounts in grade nine...

Jules: Their grades would just fall off a cliff to match the early adopter.

Naomi: Exactly. That is the assumption parents often make.

Jules: It is the prevailing logic, isn't it? That if you introduce the technology early, they sort of learn to manage it, whereas if you shelter them, they just binge and crash later.

Naomi: Right, the whole forbidden fruit idea.

Jules: Yeah.

Naomi: But the data completely rejects that logic. The harm actually intensifies rather than fades.

Jules: Wait, intensifies?

Naomi: It does. The negative effect is more pronounced the earlier the account was created. And the adverse impact persists and continues to grow by grade ten, particularly in maths, even after the late adopters have become active social media users themselves.

Jules: So, delaying access does not merely postpone the damage.

Naomi: No, not at all. Early exposure actually compounds the academic hit. The earlier the exposure, the deeper and more permanent the cognitive scarring.

Jules: Okay, but, I mean, losing six months of maths education just from creating a profile sounds almost magical.

Naomi: I know it does.

Jules: What is actually happening in the brain? Is the algorithmic content fundamentally altering their ability to process complex numbers? Are we basically talking about brain rot here?

Naomi: No, actually, the mechanism is far more mundane, and I'd argue far more insidious than toxic content. The study suggests the damage is plausibly driven by distraction and displaced time.

Jules: Okay, displaced time.

Naomi: Yeah, the researchers tracked this using what they call a smartphone pervasiveness index.

Jules: A pervasiveness index, which measures what, exactly?

Naomi: It measures how much the phone bleeds into critical daily moments.

Jules: So, they are really looking at the fragmentation of the student's day.

Naomi: Specifically, yes. They measured how often the device intrudes on sleep, the moment of waking up, and the time dedicated to doing homework.

Jules: All the crucial focus periods.

Naomi: Exactly. And they found that this pervasiveness index explained roughly 33 to 47% of the negative effect in maths and about 15 to 34% in Italian.

Jules: Well, if we think about how a teenager solves a complex maths problem, they are relying incredibly heavily on their working memory. Working memory is essentially a mental whiteboard. You have to hold the rule of fractions in your head, carry a number over, and remember the sequence of operations all at the exact same time to reach the solution.

Naomi: And then, the pocket vibrates.

Jules: Exactly. A notification flashes.

Naomi: And the whiteboard is wiped completely clean.

Jules: Cognitive scientists actually call this the switch cost. Every time you pull your attention away to check that notification, you do not just lose the three seconds you spent looking at the screen.

Naomi: You lose the entire flow.

Jules: You lose all the mental scaffolding you had built for that specific equation. You have to start the cognitive process all over again from scratch.

Naomi: Which is exhausting.

Jules: It is. Do that ten times a night, every night, for four years, and it becomes completely obvious why you are six months behind your peers.

Naomi: We should be clear, though, this part of the research regarding the pervasiveness index is a strong correlation.

Jules: Meaning they haven't proved it definitively.

Naomi: Right. They cannot definitely prove that the phone buzzing directly caused the bad grade. But it is undeniably the most plausible suspect at the scene of the crime. The slow learning drain is really about time, focus, and working memory being systematically stolen, rather than, you know, the specific videos they're watching.

Jules: So, it is the environment of constant interruption that actually does the damage.

Naomi: Exactly.

Jules: But I understand there is a strange anomaly in the data here. There was one academic subject where they found absolutely zero measurable harm from early social media use.

Naomi: Oh, yes. English competence.

Jules: English?

Naomi: Yeah, there was no drop whatsoever in reading or listening scores for English among these Italian students.

Jules: Are the authors honestly suggesting that scrolling through international TikTok feeds is secretly acting as an accidental English tutoring programme?

Naomi: That is precisely their hypothetical guess, actually.

Jules: You are joking.

Naomi: No, I mean, because an overwhelming amount of online and social media content is naturally produced in English, it provides constant incidental practice. It is ambient exposure.

Jules: So, you are losing six months of structured maths education, but your grasp of English slang is absolutely flawless.

Naomi: Essentially, yes.

Jules: But we have to firmly flag that this is just an unproven theory.

Naomi: Right, of course. It is their best attempt to explain why English was miraculously spared, rather than an established scientific mechanism. But I think it highlights a really important distinction between focused learning and ambient learning.

Jules: Because maths requires deep, uninterrupted focus.

Naomi: Exactly. Whereas language acquisition can often happen passively through immersion.

Jules: Well, knowing that displaced time and constant distraction are the likely culprits here, we really need to put this through a real-world stress test.

Naomi: Let's do it.

Jules: Because we have to ask, who's actually bearing the brunt of this impact? Like, who gets a smartphone and an unsupervised social media account in grade six?

Naomi: It's a great question. And the study itself frames early unsupervised use as far more common in disadvantaged, less educated, or non-intact households.

Jules: So, this is not hitting the population equally at all.

Naomi: No, it's not.

Jules: The parents who have the financial resources, the flexible working hours, and the educational background to police device usage, to navigate complex parental controls, or to simply enforce a strict ban until grade nine, they are actively protecting their children's test scores.

Naomi: And the data reflects that social gradient perfectly.

Jules: Which means the human stakes here are absolutely monumental. I mean, we are talking about a consumer technology that risks drastically widening one of education's biggest social divides.

Naomi: It really does.

Jules: If a student is already starting at a disadvantage due to their socioeconomic background, and then they lose an additional six months of cognitive development because of early, unmanaged social media access, catching up becomes statistically nearly impossible.

Naomi: And sadly, it is actually far worse than the headline numbers suggest.

Jules: How so?

Naomi: When you look at the methodology, the researchers note that their sample inherently excludes dropouts, vocational students, and students who repeat multiple grades.

Jules: Ah, because those students fall out of the standardised testing system and can no longer be tracked.

Naomi: Exactly. And those excluded students overwhelmingly tend to come from the most disadvantaged backgrounds and, historically, have the earliest, least supervised screen use.

Jules: So, the most vulnerable kids are effectively invisible in this data.

Naomi: Yes, which means the real-world effect is likely understated. The authors themselves argue their estimate of six months lost is actually a lower bound.

Jules: So, the true cost is probably much higher.

Naomi: Exactly.

Jules: That is a chilling reality. But before we accept these findings wholesale, I need to introduce a major caveat here.

Naomi: Okay, let's hear it.

Jules: There is something in the disclosures that genuinely warrants scrutiny regarding the entire framing of this research. Two of the co-authors of the study sit on the governance board of Patti Digitali.

Naomi: Right, the foundation advocating for digital wellness.

Jules: Well, it's a foundation actively advocating for a gradual, delayed approach to children's digital access. They are publicly campaigning for exactly the policy outcome that this study conveniently concludes is necessary.

Naomi: I see where you're going with this.

Jules: We have to press hard on this tension. Does this disclosed conflict of interest colour the entire framing of the study? Are they simply hunting for the data required to support their pre-existing advocacy campaign?

Naomi: I hear the scepticism, Jules, and it is entirely fair for any reader to question the motivations behind the research. I will push back slightly, though, on the idea that this sinks the study.

Jules: You don't think it compromises the findings?

Naomi: Well, the difference-in-differences design they utilised is unusually careful and rigorous. If a researcher were just fishing for data to support a predetermined narrative, they would not bind themselves to such a strict pre-treatment baseline.

Jules: Because it limits their wiggle room.

Naomi: Exactly. By matching students who had the exact same grade five test scores before social media was even introduced, they leave very little room to manipulate the raw outcome.

Jules: I can see the statistical methodology is undeniably strong. But when the individuals designing the questions, choosing the metrics, and interpreting the final data are sitting on the board of an advocacy group, you have to account for confirmation bias.

Naomi: Sure, bias is always a risk.

Jules: Did they choose to highlight maths scores over well-being simply because the academic data fit their campaign narrative better?

Naomi: Well, to their credit, the authors are completely transparent about what their data can and cannot prove.

Jules: True, it is in the disclosures.

Naomi: They explicitly state in the paper that they cannot rule out all unobserved confounders, perhaps specific family dynamics or personality traits that only emerge in early adolescence trigger both the early phone use and the drop in grades.

Jules: Right, the correlation versus causation issue.

Naomi: Yes. And they also clearly define the limitations of their timeline, noting these findings reflect a specific window from 2019 to 2023, dealing with specific platforms like TikTok and Instagram. They are not hiding the boundaries of their research.

Jules: Transparency is vital, absolutely. But it leaves the reader with a profound tension. On one side of the scale, you have flawless statistical matching, pointing to a severe, compounding academic penalty.

Naomi: And on the other?

Jules: On the other side, you have potential researcher bias driving the overarching narrative. We really have to leave it to you, the listener, to weigh how much that institutional conflict of interest matters to your interpretation of the facts.

Naomi: It is a very necessary debate to have. We have robust mathematics, but the human element behind who is running those numbers always provides essential context.

Jules: Always.

Naomi: Well, looking at everything we have unpacked today, the difference-in-differences twin matching, the compounding drop of standard deviations equating to six months of lost learning, the working memory destruction caused by device pervasiveness, the ambient English anomaly, and, of course, the conflict of interest, it is time to land on our verdicts.

Jules: Right.

Naomi: Based strictly on the evidence we have discussed, does this study help or hinder our understanding of kids and screens?

Jules: I'll go first. I conclude that it helps, but only if you read the caveats and acknowledge the severe, real-world implications.

Naomi: Okay, that's a fair qualification.

Jules: It helps because it quantifies the academic cost in a metric that is incredibly sobering and easy to grasp. Six months of lost schooling is something parents and educators can instantly understand, far more than a vague warning about anxiety.

Naomi: Yes, it grounds it in reality.

Jules: But it only helps if we acknowledge that the burden of policing this falls incredibly unfairly on disadvantaged parents. If society just uses this study to wag a finger at exhausted, overworked parents who rely on screens as a babysitter, it hinders. We have to look at the systemic, widening educational divide this technology is creating.

Naomi: That makes perfect sense. My verdict is that it strongly helps, precisely because it cuts through the messy, highly subjective mental health debate.

Jules: The mood tracking.

Naomi: Exactly. For years, the conversation has gone in circles, trying to prove whether social media directly causes clinical depression, and the evidence just remains fractured. This study proves that early access has a tangible, objectively measurable cost in the classroom.

Jules: It's indisputable in that regard.

Naomi: It is. It shifts the conversation away from a subjective feeling to an undeniable loss of cognitive development.

Jules: So, we have a consensus that it helps, with some heavy, structural qualifications regarding who actually bears the cost.

Naomi: We do. And for you listening, we want to offer one concrete takeaway based directly on the mechanics of this evidence.

Jules: Because the usual advice handed out by tech companies and parenting blogs is to track screen time and set daily hour limits.

Naomi: Right. But this research suggests a fundamentally different approach. Parents and guardians should focus strictly on delaying the physical age of account creation.

Jules: The actual birthday of the account.

Naomi: Exactly. It is not about whether they scroll for one hour or three hours on a Tuesday. It is about protecting those crucial early years of cognitive development from the ambient pervasiveness of the device.

Jules: Because delaying the account protects a student's working memory.

Naomi: Yes.

Jules: And their focused learning time, in a way that simply monitoring usage after the fact never will.

Naomi: Because once the account is created, the switch cost is already embedded into their daily routine. The mental whiteboard is already being wiped clean.

Jules: Which leaves us with a final thought to mull over. Society tightly regulates what happens inside a physical classroom to ensure children can focus, retain information, and learn in a safe environment.

Naomi: We have endless rules for it.

Jules: Endless. So, if early social media use is actively subtracting six months of public education from young minds, simply by existing in their pockets, should these digital platforms face the exact same strict educational regulations we apply to the physical classroom?

Naomi: It is a question we can no longer afford to ignore.