On this page
"Growth mindset" is everywhere in schools: posters, assemblies, "not yet" instead of "fail". The idea is that believing your abilities can grow helps you learn. But what happens when researchers test it at scale? The answer is more modest than the posters, and more useful once you see where the effect shows up.
What did the 2019 national experiment find?
David Yeager and a large team published a national experiment in Nature in 2019 with a nationally representative sample of US students starting ninth grade (age 14 to 15). The full paper describes the intervention as two self-administered online sessions of about 25 minutes each, roughly 20 days apart. It targeted specific beliefs, including "the belief that having to try hard or ask for help means you lack ability".
The results, in the paper's own numbers:
Lower-achieving students earned core-subject grade point averages 0.10 grade points higher than similar students in the control group, a standardised effect size of 0.11.
Higher-achieving students showed no significant change in grades.
Enrolment in advanced mathematics the next year rose from 33% to 36%.
The effect on grades appeared where peer norms in the school supported the intervention's message.
That is a real, carefully tested effect from less than an hour of material. It is also small, and it depended on who the student was and which school they were in.
What do the meta-analyses say?
Several reviews have pooled many studies.
Sisk and colleagues, 2018. Two meta-analyses in Psychological Science looked at the link between mindset and achievement (273 effects, 365,915 students) and at mindset interventions (43 effects, 57,155 students). Overall effects were weak for both, with some support for benefits among students from low socioeconomic backgrounds or at academic risk.
Macnamara and Burgoyne, 2023. A Psychological Bulletin review of 63 studies with 97,672 students found a small overall effect (d = 0.05) that was not significant after correcting for publication bias. Among the six highest-quality studies, the effect was also not significant.
Burnette and colleagues, 2023. A second Psychological Bulletin review of 53 samples found a positive effect on academic achievement (d = 0.14) when interventions targeted the students expected to benefit and were delivered faithfully, with wide variation between studies.
The review authors draw different conclusions, but their numbers point the same way on size: effects on grades are small on average and vary a lot.
What does this mean for a student preparing for JEE, NEET or boards?
The most useful part of the Yeager intervention is the belief it went after: that needing effort or help is a sign you are not good enough. In tough preparation, where almost everyone struggles with something, that belief can stop you asking questions or retrying hard problems. Here is a practical way to work on it without posters.
A three-question reset after a bad result
What strategy did I use? For example: "I reread the chapter twice and did no timed practice."
What will I change? For example: "Three timed sets this week, then review every mistake." See how to bounce back after a bad test.
Who or what will I ask for help? A teacher, a friend who understands it, or a worked solution. Asking is part of the work, not proof of weakness.
Phrases to swap
"I'm bad at organic" → "I haven't learned the mechanisms yet. Which one do I start with?"
"Toppers just get it" → "What do they practise that I don't?"
"I need help, so I'm behind" → "Getting help now is faster than staying stuck."
Mistakes are useful information, as our post on why making mistakes helps you learn explains. Parents can support the same ideas through praise and conversation; see how parents help without pressure.
When you do decide to get help on a stuck point, Fermi's AI tutor notices where a student is stuck and decides whether to explain, ask or show, so asking for help becomes a normal part of a study session.
After your next test, answer the three reset questions before you look at anyone else's marks.
