A growth mindset can help, but teaching it does much less than its popularity suggests. The best studies of Carol Dweck’s idea find small average effects on grades. They’re a little larger for struggling students, and they mostly show up where the adults around them back the message.
Say you’re a deputy head planning the next INSET day. Someone’s pencilled in a growth mindset session, and there’s a box of ‘Mistakes help me grow’ posters in the staff room. Have a look at the research before you give it the morning. It’ll change where you put the effort.
What a growth mindset is, and what Dweck didn’t say
A growth mindset is the belief that your abilities can be developed, and a fixed mindset is the belief that they’re set. Carol Dweck, the Stanford psychologist, came up with the idea. Her theory says the belief shapes how people handle challenge, effort and setbacks.
The popular version gets one part wrong. In 2015, Dweck wrote about an outbreak of what her Australian colleague Susan Mackie called “false growth mindset” (Carol Dweck in Education Week, 2015). Part of it was students being praised for effort while they weren’t learning. Her reminder was that “effort is a means to an end to the goal of learning and improving”. So if someone tells you growth mindset means praising effort, they’ve got half of it. Dweck’s own advice is that students should try new strategies and ask others for input when they’re stuck.
What the research says about growth mindset interventions
The average effect on grades is small, and researchers still argue about whether it’s anything at all.
A first big check came in 2018, from Victoria Sisk, Brooke Macnamara and colleagues. They pooled 273 studies, covering 365,915 people, on whether mindset tracks achievement. A second analysis pooled 43 intervention studies, with 57,155 people, on whether changing it helps. Both links came out weak. The hopeful sign was that pupils from low-income families, or already at academic risk, might gain something (Sisk and colleagues, Psychological Science, 2018).
In 2023, Psychological Bulletin ran two big reviews of the same research. They didn’t agree. Macnamara and Burgoyne graded 63 studies, with 97,672 students, against best-practice standards for drawing causal conclusions. Pooled, the lessons shifted achievement by 0.05 of a standard deviation. That’s very small, and it wasn’t even significant once they’d adjusted for publication bias, the tendency for disappointing studies to stay in the drawer. In the six best-designed trials, the effect was 0.02, too small to tell apart from zero. Researchers with a financial stake in positive results also reported bigger effects (Macnamara and Burgoyne, Psychological Bulletin, 2023).
Burnette and colleagues asked who it works for, and how. Across 53 randomised samples, lessons aimed at the students expected to benefit most, and delivered as designed, had a small positive effect (0.14). Results varied a lot (Burnette and colleagues, Psychological Bulletin, 2023). In a commentary, Elizabeth Tipton, David Yeager and colleagues reran Macnamara and Burgoyne’s data using Burnette’s methods. They found a meaningful effect for at-risk students (Tipton and colleagues, Psychological Bulletin, 2023).
It’s been tested in England too. An Education Endowment Foundation trial with 5,018 Year 6 pupils in 101 schools tested a programme that taught pupils their brain potential isn’t fixed and can grow with effort. The evaluation, published in 2019, found no evidence that they made extra progress in Key Stage 2 literacy or numeracy (Education Endowment Foundation, 2019).
And a 2025 review by a Durham University researcher looked at 24 trials with school-age pupils. The strongest showed null or very small effects. She advises schools and governments not to put significant time or resources into these programmes, unless they cost next to nothing (Gazmuri, Review of Education, 2025).
On this evidence, an assembly and a set of posters won’t move your results.
Where a growth mindset does seem to help
It helps most for students who are struggling, in places where the people around them support the idea.
The strongest positive result is the US National Study of Learning Mindsets. It covered 12,490 ninth-graders, roughly Year 10 in England, in 65 US public schools. Yeager, Dweck and colleagues randomly assigned them to a growth mindset programme or a control. The programme was two online sessions of about 25 minutes. Lower-achieving students’ grades in core subjects rose by about a tenth of a grade point. That’s small. The effect was bigger in schools where peer norms supported taking on challenges, but the authors are clear they can’t prove the norms caused it (Yeager and colleagues, Nature, 2019).
Later, the same team asked why it worked in some classrooms and not others. They matched 9,167 student records to 223 maths teachers. With teachers who held more of a growth mindset, the lesson lifted maths grades by 0.11 of a grade point. Where the teacher leaned towards a fixed mindset, it did close to nothing (Yeager and colleagues, Psychological Science, 2022). So the same lesson worked in one classroom and not the next, and the thing that tracked it was the teacher’s own mindset. That’s the finding I’d underline for any school leader.
Why the adults matter more than the message
People take their cue from how adults react to mistakes, more than from what adults say about ability.
Haimovitz and Dweck tested this with parents in 2016. Parents who saw failure as debilitating focused on their children’s performance and ability rather than their learning. Their children tended to believe intelligence is fixed. The children could read their parents’ views of failure accurately, but not their views of intelligence (Haimovitz and Dweck, Psychological Science, 2016).
Universities show how much the adult’s belief matters. Canning and colleagues studied 150 science, technology, engineering and maths professors and more than 15,000 students at one US university. Racial achievement gaps were twice as large in courses taught by professors who saw ability as fixed. Students were less motivated in those classes, too. The professors’ mindset predicted this better than their gender, race, age, teaching experience or tenure (Canning and colleagues, Science Advances, 2019).
For teams at work, treat this part as my view rather than a finding. Tell people mistakes are fine, then go quiet when one lands on your desk, and they’ll believe the silence. It’s the same reason I’ve said organisations should stop buying AI and start building learners.
What to do in the minute after a mistake
Respond to the attempt before the result, and make the next try the thing you’re talking about.
BRAVE’s got a letter for this. R stands for Reframe Failure, next to B for Build Curiosity, and I explained the whole framework in my BRAVE talk at TEDxSwansea. It’s my method for putting this research to work. It isn’t a tested intervention like the studies in this piece, so weigh it that way.
In practice, that’s the adult meeting a mistake with curiosity rather than disappointment. It’s built on a line of mine, “When failure doesn’t feel scary, learning becomes so, so powerful.”
At the close of a lesson or a project review, I’d ask the question that goes with it. “What did we learn today because we failed that we could not have learned if we got it right the first time?”
Dweck’s 2015 piece has a script I’d pair with it. It moves the talk from the person to the strategy, and I’d use it in a one-to-one at work as readily as at a desk in Year 8. In her words, “Let’s talk about what you’ve tried, and what you can try next.”
Curiosity, the intelligence I’d build next to it
Of the five human intelligences AI can’t automate, Curiosity is the one I’d put next to any growth mindset work. It gives people a reason to try again after a mistake.
The five are Curiosity, Creativity, Critical Thinking, Adaptive Leadership and Emotional Intelligence. It’s my framework, so read it as a point of view rather than a result.
There’s lab evidence that curiosity helps memory, though. In a 2014 brain-imaging study, people remembered answers better when they’d been curious about the question, both straight away and a day later. They also remembered unrelated material they’d seen while curious (Gruber, Gelman and Ranganath, Neuron, 2014). It’s a lab study, not a classroom, so I treat it as a reason to try something, not proof it’ll work in yours. Start with one habit. Ask people what they expect before you show them the answer. It’s what I want from students using AI too, which is why I’ve written about teaching them to question what AI tells them.
Growth mindset in the classroom: what to do on Monday
Change how the adults respond to mistakes and effort, and measure it, before you spend a penny on posters. I’d run it in four steps over half a term.
- Watch the 30 seconds after a mistake. In three or four lessons, note what happens straight after a pupil gets something wrong. Does the adult ask about the thinking, or move on to someone who’s right?
- Change the script. Swap ‘good effort’ on work that isn’t improving for Dweck’s question about what they’ve tried and what they’ll try next.
- Check what you reward. Look at your merits, reports and certificates. When rewards only celebrate getting it right first time, people learn to hide their attempts.
- If you run a mindset lesson, aim it and measure it. Point it at struggling students, brief their teachers first, and decide now which grades you’ll check at the end of term.
None of this needs a school. If you lead a team, try step one in your next three meetings. Listen to what happens when someone says ‘I got that wrong’.
How Colin helps
My keynotes and advisory work for schools, universities and leadership teams can focus on how the adults in a room respond to mistakes, because that’s where this evidence keeps pointing. I founded The Human Behaviour Institute. In a growth mindset conversation, I’d start with Reframe Failure, from the BRAVE framework I presented at TEDxSwansea.
If you’d like to talk about a keynote or advisory work for your school, university or business, get in touch.
Related questions
Can you change your mindset?
Beliefs can shift, but that doesn’t automatically shift results. In Macnamara and Burgoyne’s 2023 review, 13 studies showed their lesson had changed students’ mindsets as intended. Even there, the effect on achievement wasn’t significant. So change what’s around the belief too, like how mistakes are handled and what gets praised.
Is a growth mindset just about praising effort?
No. In 2015 Dweck warned against praising students who are trying hard but not learning, as part of what her colleague Susan Mackie called “false growth mindset”. Effort matters because it leads to learning. Praise the strategy and the progress, and help people change approach when the effort isn’t working.
Does a growth mindset work for adults at work?
The studies I’ve cited here come from schools, universities and families, so I’d be wary of anyone promising big results for a workforce. But the parent and university studies carry one lesson into any team. People read your reaction to failure more than your words about it, so watch your face the next time someone brings you a mistake.