neur.ro

Growth mindset

type: effect · origin: study · evidence: contested · themes: habits and self-control, learning and memory

Claim assessed: Teaching students that abilities can grow through effort improves their results.

Published · how we work

In short

A growth mindset is the belief that intelligence and other abilities can be developed through effort, good strategies and help. Its opposite, a fixed mindset, is the belief that abilities are given and don’t change. The theory, popularised by Carol Dweck, says that students with a growth mindset do better, and that short interventions teaching it can raise their grades. The interventions spread to schools around the world. How much they help is disputed, though.

What it says

The first meta-analysis. Sisk et al. (2018) analysed the link between mindset and school results in 273 studies with over 365,000 students, then the effect of interventions in 43 studies with over 57,000 students. Both effects were weak. Some results still suggested that students from low-income families or at academic risk might benefit.

The national experiment. Yeager et al. (2019) tested an online intervention of less than an hour on 12,490 ninth-grade students in 65 US public schools chosen to represent the country. Students were randomly assigned, and the analysis was registered in advance. The intervention explained, among other things, that the brain “grows” like a muscle when it is challenged. For students with weaker previous results, the grade average in core subjects rose by 0.10 points, a small effect (d = 0.11). For the other students, the effect was practically zero. The effect was larger in schools where classmates were willing to try hard problems. Separately, in an analysis decided after the data were collected, the share of students choosing advanced mathematics the next year rose from 33% to 36%.

The critical meta-analysis. Macnamara and Burgoyne (2023) reanalysed 63 studies with almost 98,000 students. Across all studies, the effect was very small (d = 0.05) and no longer significant after correcting for selective publication. In the 6 highest-quality studies, the effect was 0.02, not significant. They also found problems of method: in many studies, the group receiving the intervention also received other things, such as encouragement or study strategies, and teachers often knew who was in which group. Studies with authors who had a financial interest in the result reported significant effects more than two and a half times as often.

The effect size of growth mindset interventions on grades. In Yeager and colleagues' national experiment: 0.11 for lower-achieving students and 0.01 for the rest. In Macnamara and Burgoyne's meta-analysis: 0.05 across all studies and 0.02 in the highest-quality studies. A small effect, and where it shows Yeager et al., 2019; Macnamara, Burgoyne, 2023 Effect size on grades (d) 0.2 would be a small effect by the usual convention The national experiment, 12,490 students lower-achieving students 0.11 the other students 0.01 The critical meta-analysis all 63 studies 0.05 the 6 highest-quality studies 0.02 Small in students who struggle, close to zero on average.
Yeager et al. (2019), grade point average in core subjects at the end of ninth grade. Macnamara and Burgoyne (2023), the average effect before correcting for selective publication, after which even 0.05 became non-significant.

The reply. Tipton et al. (2023), Yeager among them, argue that an average across all studies hides where the intervention works. The theory doesn’t promise effects in everyone, but mainly in students who struggle. Reanalysing Macnamara and Burgoyne’s data with methods that take these differences into account, they found a significant effect in the targeted groups.

Example

A student who gets a bad maths grade can conclude “I’m not good at maths” or “I haven’t yet found how to learn this”. The second reading may make her try again, with a different strategy. The example is an illustration, not a study cited here.

How to apply it

The steps below are a practical way to apply it that we propose, based on the studies cited.

  1. Don’t expect a transformation. At best, the measured effect is small and shows mainly in those who struggle.
  2. Tie the belief to concrete things. The idea that you can improve helps only if it comes with learning strategies, time and help. See the entry on spacing and retrieval.
  3. The environment matters. In the national experiment, the effect held up better where classmates accepted challenges. A motivating message doesn’t make up for a class where effort is mocked.
  4. Be sceptical of programmes sold to schools. Some large studies have authors with financial interests, and the effects they reported were larger.

Limits and nuances

Sources

See also: Delayed gratification, Ego depletion, Stereotype threat