Growth mindset
Also known as mindset theory, implicit theories of intelligence
Believing ability can grow does help — much less than the popular version claims, for some students, in some schools.
The claim That believing intelligence is malleable rather than fixed causes students to persist through difficulty and achieve more.
Contested Two large meta-analyses found effects near zero on average; the largest single randomised trial found a small, real effect concentrated in lower-achieving students at supportive schools.
Carol Dweck's distinction is genuinely useful as a description: some people treat ability as fixed and read difficulty as evidence of their limit, others treat it as growable and read the same difficulty as the work. The second stance is obviously better to hold. The contested question is whether teaching it to people changes what they achieve, and by how much.
For about a decade the popular answer was 'enormously', and schools bought it at scale. Then came the meta-analyses. Sisk and colleagues in 2018 pooled 273 studies on the correlation between mindset and achievement and 43 on mindset interventions. Mindset accounted for roughly 1% of the variance in academic achievement, and intervention effects were weak overall — with the interesting exception that they were less weak among students who were academically at risk or economically disadvantaged.
The other half of the story arrived in 2019 and is the reason 'contested' is the honest label rather than 'debunked'. Yeager and colleagues ran the National Study of Learning Mindsets: a nationally representative sample of about 12,500 American ninth-graders, preregistered, independently analysed. Two twenty-five-minute online sessions produced a small but real improvement in grades among lower-achieving students — and only where the school's peer norms supported it. Where the surrounding culture treated struggle as evidence of stupidity, the intervention did nothing.
Put together, the picture is coherent and much less exciting than the books. A cheap intervention produces a small effect on a subgroup under conditions. That is a normal result in education research. What was oversold was not the existence of the effect but its size and its universality, and the overselling was done largely outside the research literature.
For a self-directed adult learner — the reader of this site — the honest position is that almost none of this evidence is about you. It is about school students, mostly in one country, mostly measured by grades. What survives the translation is not a technique but a piece of framing: interpreting difficulty as information rather than as a verdict, which the desirable-difficulties literature independently says is the correct interpretation.
What to do differently
- Read a plateau as data about your current method, not as a reading of your ceiling.
- Praise and criticise the process rather than the person, including in your own head — 'that approach did not work' rather than 'I am bad at this'.
- Expect the environment to matter as much as the belief. The largest trial found the intervention worked only where the surrounding culture supported it, and that finding applies to the company you keep too.
- Do not expect a belief change to substitute for a method change. Mindset is a small effect; retrieval and spacing are large ones.
- Be wary of anyone selling mindset as a programme. The measured effects are far too small to carry that weight.
Limits and trade-offs
- The headline claim was oversold for a decade, and the meta-analytic average is close to zero. Any page presenting this as established is out of date.
- The positive trial found effects for lower-achieving students in supportive schools. Generalising it to everyone is exactly the error that caused the original problem.
- It shades easily into blaming the learner: if your beliefs are what is holding you back, then failing is your attitude's fault rather than a matter of teaching, time or circumstances.
- Almost all of the evidence concerns school-age students and grades. Its relevance to self-directed adult learning is an assumption, not a finding.
What this means for a language
The fixed-mindset belief that dominates language learning has a specific name — 'I don't have a gift for languages' — and it is unusually consequential, because it licenses quitting at exactly the point where everyone finds it hard. It is worth knowing that aptitude differences in language learning are real and measurable, so the belief is not simply false; what is false is the inference from 'this is harder for me' to 'this is not available to me'. The useful reframe is small and concrete: the intermediate plateau where nothing seems to improve is a universal feature of the process rather than a personal verdict, and it is what a change of method — more output, harder input, a specific weakness targeted — exists to break.
The research
- Sisk, Burgoyne, Sun, Butler & Macnamara (2018), Psychological Science
Two meta-analyses (273 correlational studies, 43 interventions): effects weak overall, with mindset explaining about 1% of achievement variance.
- Yeager et al. (2019), Nature
Nationally representative trial with ~12,500 ninth-graders: two 25-minute sessions improved grades for lower-achieving students, but only in schools whose norms supported it.
Related
Read next
-
273 studies on the mindset–achievement correlation and 43 on interventions. Effects were weak overall; mindset accounted for about 1% of the variance in achievement.
-
The other half of the mindset argument: a nationally representative trial with about 12,500 ninth-graders found a small but real effect on grades for lower-achieving students, and only where the school's norms supported it.
-
Popular-press, and it argues a thesis rather than surveying a field — but its chapters on interleaving, desirable difficulties and the limits of early specialisation are the most readable version of that literature anywhere.