Tell a child they’re smart and you’d expect confidence to follow. A 1998 study found almost the opposite. Praise a child’s intelligence and they become more fragile, not more secure. Praise their effort and something sturdier takes hold.
The study, conducted by Claudia Mueller and Carol Dweck with fifth graders, is one of those findings that’s been repeated so often it’s turned into a parenting cliché. But the actual mechanism underneath it is worth slowing down on, because it explains something bigger than how to compliment a ten-year-old.
What the study actually did
Researchers gave children a set of problems to solve, then offered one of three responses after their first success: a comment on their intelligence (“you must be smart at these”), a comment on their effort (“you must have worked hard”), or neutral feedback with no attribution at all.
What followed showed just how much a single sentence of praise can redirect a child’s motivation. Given a choice between an easier task they were told they’d do well on, or a harder one they might struggle with but learn from, most children praised for intelligence picked the easy option. In one part of the study, 67 percent of intelligence-praised children chose the safer, performance-protecting task, compared to just 8 percent of the effort-praised group.
Then it got more revealing. When children later reported their scores to an anonymous peer, with no evaluator able to check, 38 percent of those praised for intelligence inflated their results. Among those praised for effort, only 13 percent did. Every single misrepresentation went in the same direction: better than the truth, never worse.
Why being called smart backfires
The intuitive read is that praising intelligence should build confidence. The study suggests it builds something narrower: a self-image tied to a fixed trait that has to be protected rather than developed.
Once a child hears “you’re smart,” a quiet equation forms. Being smart is who I am, and getting something wrong would mean I’m not. A harder problem stops being an interesting challenge and starts being a threat to that identity. Choosing the easy task isn’t laziness — it’s a rational, self-protective response to a trap the praise itself set up.
Effort praise doesn’t create that trap, because it locates success in something changeable. Working hard is a behavior, not an identity. Failing at a harder problem after trying doesn’t threaten “I worked hard.” It’s just information about where the work is currently landing.
Where the common parenting advice falls short
A lot of popular advice reduces this study to a simple swap: stop saying “you’re smart,” start saying “you worked hard.” That’s a reasonable starting point, but the deeper point isn’t really about vocabulary. It’s about what kind of self a compliment implicitly builds.
You can praise effort in a hollow, formulaic way and still miss the underlying shift, which is toward praising a process the child can control and away from an identity they have to defend. Conversely, specific attention to what a child actually did (the strategy they tried, the mistake they caught and fixed) tends to land better than either scripted phrase on its own.
The lying finding is the part that gets underplayed. It wasn’t just that intelligence-praised children performed more cautiously. They were willing to distort the truth to protect a self-image that had nothing to do with actual performance and everything to do with how performance was being interpreted by others, even hypothetically.
Why this doesn’t stop at childhood
The same mechanism plausibly shows up well past fifth grade. Adults praised primarily for being talented, brilliant, or naturally gifted often seem to develop a similar aversion to visible struggle. Taking on something you might fail publicly at feels less like growth and more like risking the label that’s become part of your identity.
This is one reason capable adults sometimes avoid the exact challenges that would develop them further. Not from lack of ability, but because the identity built around being good at things doesn’t leave much room for looking like you’re still learning.
The role of environment and attention
Environments that reward the appearance of competence over the process of developing it will reinforce this pattern regardless of what any individual believes about themselves. A classroom, or a workplace, that treats mistakes as evidence of insufficiency rather than information will train people to hide struggle instead of working through it in view of others.
I’ve noticed a version of this teaching university seminars. Students who’ve spent years being told they’re the smart ones in the room are often the most reluctant to ask a basic question out loud, not because they don’t have it, but because asking would puncture something they’ve been protecting for a long time. The students who ask anyway, plainly and without apology, tend to be the ones operating from something closer to “I can work this out” than “I’m supposed to already know.”
Attention plays a role too. Where you direct a child’s or a colleague’s attention after success (toward the outcome and what it says about them, or toward the process and what it says about the task) shapes what they optimize for next time. Outcome-focused attention narrows toward safety. Process-focused attention tends to widen toward challenge.
Unlearning the smart label
The Sovereign Mind framework is a useful way to hold this study, because it’s fundamentally about which parts of self-worth are inherited from other people’s language and which parts you can actually rebuild.
Unlearning starts with the inherited belief worth questioning: that being praised for a fixed trait like intelligence is unambiguously good, full stop.The Mueller and Dweck data says otherwise: the 38 percent of children who lied about their scores weren’t cheaters by nature, they were protecting a label someone else had handed them.
Restoration is the redirect that follows, toward process and effort instead of outcome. In practice that can look as simple as a student describing which strategy they tried on a hard problem, rather than what grade they think it proves about them, and noticing that the first conversation feels less loaded than the second.
Defense, finally, is the boundary against environments (and against your own inner monologue) that treat visible struggle as proof you don’t belong somewhere. A colleague who asks a basic question in a meeting full of “smart” people isn’t exposing a weakness. They’re doing the thing the effort-praised children in the study did without much difficulty at all.
What a more accurate self-image looks like
None of this means praise is dangerous or that intelligence shouldn’t be acknowledged at all. The issue isn’t the compliment itself, it’s what the compliment quietly teaches someone to protect.
A self-image built around “I am capable of figuring things out” tends to hold up under difficulty better than one built around “I am smart,” because the first is a process you can keep proving to yourself through effort, and the second is a label that a single hard problem can appear to threaten.
This is worth noticing in yourself as much as in how you talk to children. If a particular kind of feedback (about talent, about being naturally good at something) makes you quietly avoid the situations that would test it, that’s the same mechanism the 1998 study documented, just running on an adult timeline.
A useful way to test this in your own life is to notice your first instinct when offered something you might not be good at right away. If the instinct is to decline, or to quietly lower the stakes so a poor result wouldn’t mean much, that’s worth sitting with rather than immediately overriding. It doesn’t mean you’re fragile. It usually means some version of this loop, praise tied to identity, identity in need of defending, formed a long time ago and has simply never been named.
Where the nuance sits
It would be too simple to say effort praise is always right and intelligence praise is always wrong. Context matters. A child who is chronically anxious about ability might need a different kind of reassurance than a child who is coasting on natural talent without much effort at all.
The finding isn’t a rule to apply mechanically. It’s a lens for noticing what a particular comment is actually reinforcing, in a particular person, at a particular moment.
