How Will I Know What’s Real?
Helping our kids (and ourselves) figure out what to trust online
Okay, please don’t judge me. My kids watch YouTube Kids on their iPads. Yes, we try to limit screen time, and yes, we limit them to educational games and videos (or at least what we think are educational videos).
The other day, my son Dylan asked me about a video he’d seen. I wish I could remember the details (it was some animal doing something I was pretty sure animals don’t do). The video wasn’t accurate, but it was harmless, a far cry from the health misinformation we see circulating every day. Still, it was our first foray into the world of “you can’t believe everything you see, read, or hear.”
Dylan is super thoughtful and inquisitive, so his next question was, “Well, how will I know what’s real?” My husband and I looked at each other with wide eyes. Yep, our children have reached the age where we need to broach misinformation. I work on this every day, and I still found myself struggling for an answer. Sometimes it’s super obvious (an elephant doing a backflip, for example), and other times it isn’t. Some things that sound completely made up are true (wombats really do poop cubes), while plenty of false claims come with a confident narrator, a chart, and a link to a study. So, how do we help our kids (and ourselves) figure out what deserves our trust? Let’s discuss…
Most of us are worse at spotting false information than we think
We stick to educational content in our house, and I’ve realized how little that word tells me. Anyone can put “learn about animals” in a title, and a video made with the best intentions can still get the facts wrong. AI is adding to the volume, too. In April, more than 200 organizations and experts signed a letter from the children’s advocacy group Fairplay raising concerns about AI-generated videos on YouTube and YouTube Kids (YouTube says it limits this content in the kids’ app).
If you find it hard to tell what’s real and what isn’t, you’re in good company. When Stanford researchers tested 3,446 high school students, 52% thought a grainy video shot in Russia was strong evidence of voter fraud in the US, and two-thirds couldn’t tell news stories from ads. Being highly educated doesn’t make us immune, either. In an earlier comparison, historians with PhDs were easily fooled by websites that professional fact-checkers saw through quickly.
The fact-checkers left the page and looked at what other sources said about it, a technique called lateral reading. The good news is that this skill can be taught. In a randomized study of about 500 high school students, those who got six 50-minute lessons doubled their scores from before the lessons on a test of spotting dubious websites (although they still earned only half of the possible points). I wish this were embedded in our children’s education right now, and well before high school; in fact, many experts argue that media literacy education should start in preschool and continue every year.
Teaching kids to doubt everything can backfire
I don’t want Dylan to walk away thinking nothing is real. I want him to be able to figure out what deserves his trust. Research in adults suggests that warnings about misinformation can lower trust in accurate news. In one large experiment, media literacy tips worked best when they encouraged skepticism of false information and trust in reliable information at the same time. Common Sense Media describes the goal for teens as “investigate, not doubt,” and I think that works at any age.
None of this means becoming a detective every time we scroll past a video (who has the time?). Sometimes it’s enough to recognize that we don’t know enough to believe something, share it, or act on it.
How to talk with younger kids about what’s real online
I think the place to start with kids around Dylan’s age (he’s turning 10 in December) is by asking, “How could we find out?” When they ask if something is real, we can look it up together in a library book or on a trusted kids’ site. It also helps to let them answer with real, pretend, or not sure yet. Pretend can be fun (cartoons, for example), and it’s okay to say “not sure yet.”
Young kids can also understand that every video was made by someone, and that some of those people are trying to teach us, make us laugh, or keep us watching. Most of all, I want my kids to know they can show me anything strange or scary without getting in trouble. Settings can help, too. YouTube Kids has an “Approved content only” setting that limits your child to videos and channels you’ve picked, although that tells you who you’ve approved and doesn’t guarantee every fact is right. Still, this shouldn’t all fall on parents.
How to talk with teens and older kids about what’s real online
Teens are running into this constantly. In a Common Sense Media survey of 1,000 teens, 35% said they’d been deceived by fake content online, 41% had seen content that was real but misleading, and 22% had shared something that turned out to be fake. That middle number is a good conversation starter, because a real video from another year or another country can mislead just as easily as a fake one (a reverse image search can often show where it first appeared).
Older kids are ready for the fact-checkers’ approach. Mike Caulfield’s SIFT method gives them four moves (stop, investigate the source, find better coverage, and trace the claim back to its original context). In practice, that often means opening a new tab and searching for who’s behind a post, since a polished page and a big following tell you very little. The same goes for chatbots. When journalists evaluated thousands of answers about the news from four popular AI assistants in 2025, 45% had at least one significant problem, such as factual errors or sources that didn’t support what the assistant said. So, teens should open the sources and confirm they say what the chatbot claims. If a post makes them furious or thrilled, that’s a good cue to slow down before sharing it.
What we can do as adults
This applies to me, too, because it is getting harder to discern what’s real. We can start by retiring “I can always tell,” since that confidence gets us into trouble. It’s easy to scrutinize a claim that annoys us and share one that confirms what we already think, whatever our politics, so I try to be extra careful with the things I want to be true (aka confirmation bias). It also helps to read past the headline and click through to the original study or report. If a message or phone call feels urgent and asks for money or personal information, hang up and contact the person or company through a number you already know. (Mom, are you reading this?)
Get Dr. Jess Steier’s stories in your inbox
Join Medium for free to get updates from this writer.
The most useful thing we can do might be saying “I don’t know, let’s look” out loud. Our kids are watching how we handle this, and it’s good for them to see an adult who doesn’t have the answer go find it.
Red flags in science and health claims
A link to a study doesn’t make a claim true. These are patterns we see all the time at Unbiased Science, and while none of them proves a claim is false, each is a good reason to look closer.
- An animal study presented as proof about people. Animal research is important, but findings don’t automatically translate to humans. Check whether the doses, exposures, and outcomes are relevant to people.
- A scary claim that never mentions dose. Anything can be harmful at a high enough dose (even water), so ask how much it would take and whether anyone is realistically exposed to that much.
- One dissenting expert against everyone else. The dissenter could turn out to be right. Scientific consensus isn’t infallible, but it reflects a much larger body of evidence than one expert’s opinion.
- Data that looks hand-picked. Test enough comparisons, and some will appear statistically significant by chance alone (researchers call this data dredging or p-hacking). Be skeptical when a claim rests on one surprising subgroup or one study pulled from a large body of research pointing the other way.
- A correlation presented as a cause. Two things can rise together without one causing the other, so ask what else could explain the pattern.
- Someone who stands to gain. Check whether the person sharing the claim is selling a supplement, a test, or a course through the link in their bio. A financial stake doesn’t automatically make anyone wrong (and this applies to companies and institutions, too), but it’s a reason to look for independent confirmation.
These questions can also help us recognize claims that deserve our trust. When multiple independent studies, experts in the relevant field, and a broader body of evidence all line up, that’s a good reason to have confidence in a claim.
Tools for checking what you see online
When a claim is already circulating widely, FactCheck.org and Snopes are good places to start. Google Lens and TinEye can show where an image has appeared before, and PubMed lets you look up a study. If you suspect something was made with AI, the Gemini app can look for Google’s watermark and Content Credentials, but those signals are limited, and finding nothing doesn’t mean the content is real. If you’d like to practice these skills, Stanford’s Social Media Lab offers free, self-directed modules through its Digital Strength Initiative, covering lateral reading, reverse image search, and more.
Why we keep sharing science online
The irony, of course, is that you’re reading this online. Unbiased Science exists on the same internet as all those misleading videos and false claims, alongside many other credible organizations using social media to share good information. That’s where many people are looking. In a Pew survey this year, 58% of US adults said social media is an important way they keep up with the news, up from 43% in 2022 (among adults under 30, it’s 76%). So, what can we do about all of the bad information out there? Part of the answer is to fill the internet with good information, which is why I encourage people to share what they know if they’re willing. When a parent or a curious kid goes looking for an answer, I want them to find something they can trust. That doesn’t mean you should take our word for it, either. The same checks apply to us, so look into who’s behind a page, and click through to the studies we link.
What I want Dylan to know
The answer to Dylan’s question is that he won’t always be able to tell what’s real just by looking, and neither will I. I don’t want him to stop being curious or to decide that everything he sees is fake. I want him to keep doing what he did with that video, which means noticing when something doesn’t add up, asking about it, and being willing to find out (and sometimes being okay with not knowing just yet).
Stay Curious,
P.S. Today is Election Day, and I’m encouraging you to vote like your life depends on it, and your children’s lives, and your grandchildren’s lives. Voting matters for many reasons, but as a public health cheerleader, I am here to remind you that science, medicine, and public health are on the ballot. The people we elect make decisions that shape whether families have health insurance, whether rural hospitals stay open, access to vaccines, how quickly local health departments can respond to an outbreak, and how much funding goes to the research behind tomorrow’s treatments. They also make decisions about the safety of our food, air, and water.
If you’re not sure where candidates stand, the Scientist Network for Advancing Policy (a nonpartisan group that doesn’t endorse candidates) asked them. Its Stance on Science project has published 5,532 responses from 1,026 candidates across all 50 states and DC. Depending on your state, you can see what candidates said about science education, energy, the environment, jobs, and public health. Not every candidate responded, but it’s a good place to start. If you need to find your polling place or check your registration, start at vote.gov. VOTE!
If you’d like to support this work, please consider a tax-deductible donation to The Center for Unbiased Science and Health (CUSH), our nonprofit that funds educational initiatives like this one.
