In 1998, two researchers - Carol Dweck and Claudia Mueller - sat fifth graders (their words, not ours - it's an American study) down with a set of puzzles. They were the kind that look like little IQ tests: shapes, patterns, find the missing piece. The kids did the first easy round fine.
Then the researchers split them in half. To one group they said, "you must be smart at these problems." To the other group, "you must have worked hard at these problems." Same kids. Same room. One sentence different.
Next, they offered every kid a choice. You can try a harder set of puzzles - you might struggle, but you'll learn something. Or you can try another easy set, the kind you already aced.
The kids who got told they were smart overwhelmingly picked the easy ones. The kids who got told they worked hard overwhelmingly picked the hard ones.
That's the whole post, really. Everything that follows is just unpacking why.
What does "smart" actually do to a kid?
Two things, both quiet, both compounding.
The first: a kid who gets called smart starts to fold "smart" into how they see themselves. It becomes part of their identity. And once something is part of who you are, you don't want to risk losing it.
The second: from that point, every task becomes a test of the label rather than a chance to learn. Hard tasks start feeling threatening. Mistakes feel like proof you're not actually smart. The growth mindset Dweck spent her career on is, at its core, about restoring a kid's ability to see hard things as interesting rather than dangerous. Her book Mindset: The New Psychology of Success is the parent-friendly read if you want to go deeper.
Where does this start? At home and in the classroom.
The language parents use is mega important - more on that in a minute.
But there's a second front, and it's worth being honest about it. A lot of primary schools are working hard to design rich, open tasks. We see it in classrooms across Australia and we love it.
There are still classrooms - and whole schools - leaning heavily on worksheets and ability streaming. And in those rooms, deep learning isn't really happening. The kids who get the worksheet right see themselves as smart. They get the tick. They get told they're smart. But getting something right isn't the same as learning it.
The kids who get it wrong figure out something else: they get good at copying. They learn the social skill of looking like they know. Not because they're lazy - because the system is asking them for the answer, not for the thinking. So they find the answer the fastest way they can.
Either way, nobody's being driven to learn. They're being driven to get the answer right. And the kids who do get it right pick up a label they're going to carry for years.
Why does this hit hardest at high school?
It's not that the tasks suddenly get more open. If anything, high school work often gets more closed - textbooks, one thing for everyone, larger classes, less room for the teacher to differentiate even when they want to.
What changes is the kid. The work gets genuinely harder. And for the first time, a kid who's always seen themselves as smart starts to doubt that. The label they've carried since Grade 2 is suddenly up for review, and they can feel it.
This was me.
I saw myself as a smart kid all the way through primary school, because I'd been told that's what I was. Then I hit high school and realised this stuff was actually hard. And the funny thing is, I still saw myself as smart - I just figured out, completely subconsciously, that I could preserve that by not doing my best.
If I wasn't trying my hardest and I was still passing, that made me smart, right? It also meant I didn't have to compare myself to the kids who were grinding. They were trying their best and getting better marks than me, sure - but I wasn't trying. So in the made-up scoreboard in my head, I was still ahead. The smart kid who could do it if he wanted to.
This is a pattern a lot of teenagers fall into, and it's almost always invisible from the outside. The kid looks lazy. The kid is actually protecting something.
Jo Boaler at Stanford has written about this exact moment - the cost showing up in adolescence. Her book Mathematical Mindsets is the one I'd hand a teacher. Limitless Mind is the one I'd hand a parent.
What should we say instead?
Three small replacements. None of them are hard.
One. Replace "you're so smart" with "you found a really good way to solve that." Name the strategy, not the trait. Researchers call it process praise. In the room, it's just specific noticing.
Two. Replace "good job, that was easy for you" with "I noticed you stuck with that even when it got tricky." Name the persistence. Dweck's own line from Mindset is even better: "I liked the effort you put in. Let's work together some more and figure out what you don't understand."
Three. Replace "wow, you're a natural" with "what was the bit you found hardest?" Make hard-bits visible and discussable. Boaler's point in Mathematical Mindsets is that mistakes literally grow new neural pathways. The kid who learns this isn't afraid of getting things wrong. That kid keeps going when other kids stop.
Does this mean I can't compliment my kid?
No. Compliment them constantly. Just compliment what they did, not what they are.
"You're so smart" becomes "you tried three different ways to solve that, that was clever."
"You're a great artist" becomes "the way you mixed those colours is interesting, what made you choose that?"
"You're so kind" becomes "I saw you give your sister the bigger half. That was a kind choice."
Notice the pattern. The kid is still a kid worth complimenting. The compliment just lives one layer down - in the thing they did, not the type of person they are. Same warmth. Different message.
But what about kids who are genuinely advanced?
This is the bit where most posts on this topic lose half their readers, so let me be direct.
Yes - some kids are working ahead of their year level. Yes - it's fine to acknowledge that. The point isn't to pretend ability differences don't exist. The point is to make sure the kid understands their advanced-ness came from something they did, not from a magic trait they were born with.
"You're so good at reading" locks the trait in.
"You read a lot, and you can tell - you're noticing things in this book most kids your age miss" names the cause.
There's a research angle here worth knowing about. Boaler's work on ability grouping found that kids streamed into "top" maths classes in Year 7 perform worse over five years than kids in mixed classes. The reason isn't the maths. It's that the streaming itself locks in fixed-mindset thinking. The same dynamic plays out in living rooms when "you're the smart one" gets repeated for ten years. The label is the thing that costs them later, not the early ability.
