Praise, Growth Mindset, and What the Research Actually Shows

Few ideas have travelled from psychology into parenting as fast as growth mindset. Somewhere along the way, 'praise the process, not the person' hardened into dogma, schools painted the word YET on walls, and parents began policing themselves for ever saying 'clever'. It is worth stepping back and asking what the research actually supports — because the honest answer is: some of it, solidly, and some of it much less than the posters suggest.

The core finding that has held up

The foundational studies by Carol Dweck and colleagues, notably Mueller and Dweck's 1998 experiments, compared children praised for ability ('you must be smart at this') with children praised for effort ('you must have worked hard') after a task. The ability-praised children subsequently chose easier tasks, enjoyed the work less, gave up sooner when it got hard, and were more likely to misrepresent their scores. The effort-praised children sought challenge and persisted. The mechanism is intuitive once seen: if success proved I am smart, failure threatens to prove I am not — so the safest move is never to risk failure. Ability praise quietly raises the cost of trying.

This specific finding — that the type of praise shifts how children respond to difficulty — has replicated reasonably well and is worth taking seriously. Longer-term work has even linked the kind of praise toddlers receive to their attitudes toward challenge years later.

What the research does not show

The stronger claims have fared worse. Large-scale studies and meta-analyses of mindset interventions — programmes teaching students that intelligence is malleable — find effects on actual grades that are, on average, small. Not zero: the best evidence, including a national study of thousands of US students, suggests a modest benefit concentrated among lower-achieving students, at very low cost. But small. Mindset training does not turn average students into top students, several attempts to replicate headline results have struggled, and Dweck herself has objected to the caricatured version spread through schools.

Two practical warnings fall out of this. First, a growth mindset is not a substitute for the boring machinery of achievement — good teaching, actual practice, sleep, and feedback. Telling a child to believe in growth while changing nothing about how they study is motivational wallpaper. Second, beware what researchers call false growth mindset: praising effort itself, detached from results. 'At least you tried hard' delivered after every failure becomes a consolation prize, and children decode it quickly — effort praise with no interest in what the effort produced reads as 'you failed, but you are fragile'. The point was never effort for its own sake. It was effort as the thing that drives improvement.

Praise that actually helps

Pulling the reliable threads together, good praise has three properties: it is true, it is specific, and it points at something the child can do again on purpose. 'You are a genius' fails all three. 'You redid the corrections two days in a row and it shows in Section A' passes all three. Strategy praise, in fact, may be the most useful kind — better than generic effort praise — because it names the behaviour worth repeating.

Everyday language swaps

• Instead of 'You are so smart' — try 'You found a method that worked. Walk me through it.'
• Instead of 'You are just not a maths person' (or agreeing when they say it) — try 'You are not fast at this yet. Fractions took time too, remember?'
• Instead of 'At least you tried' after a failure — try 'That approach did not work. What would you change for the next attempt?'
• Instead of 'Wow, full marks, so clever!' — try 'Full marks — those practice papers clearly paid off.'
• Instead of 'This is easy, just think!' — try 'This one is genuinely tricky. Which part is blocking you?'
• Instead of 'Why did you get this wrong?' — try 'Interesting mistake — what was the thinking here?' Mistakes examined calmly are the raw material of improvement.

The part adults forget: children copy what you model

Language swaps are easily undone by behaviour. A parent who says 'mistakes help you learn' but visibly deflates at a bad grade has taught the real lesson. Children calibrate to what adults do about difficulty, not what they say about it — so let them see you attempt things badly, say 'I got that wrong, let me try another way', and treat your own errors with curiosity rather than shame. In a results-focused system like ours, where a child can come to believe their PSLE score is a measurement of their worth, that modelling is not a soft extra. It is the most credible growth-mindset intervention a family can run — and unlike the posters, it is one the research quietly supports: not believing your way to better grades, but lowering the cost of difficulty enough that a child keeps showing up to it.