Growth Mindset: A National Experiment and Two Meta-Analyses
Key Points
- A 2019 pre-registered national experiment in 65 US public schools found that a short online growth mindset program raised ninth-grade grades by 0.10 grade points among lower-achieving students.
- The authors say the advanced mathematics result was exploratory and that the school-norms result cannot support causal conclusions.
- A 2018 pair of meta-analyses found weak overall effects, and a 2023 meta-analysis found a small effect that became nonsignificant after correction for publication bias.
- A 2023 commentary, whose authors include the lead author of the national experiment, argues that the averages hide effects in at-risk groups.

A growth mindset is the belief that intelligence and other personal attributes can be developed with effort. The evidence for school programs built on the idea has been argued over in print for years. This article reads four papers: a 2019 national experiment in Nature, a 2018 pair of meta-analyses, a 2023 meta-analysis and a 2023 reply to it.
What the theory claims
Sisk and colleagues, in a 2018 paper in Psychological Science, define mind-sets (also called implicit theories) as beliefs about the nature of human attributes such as intelligence2 . They describe the theory as holding that people with growth mind-sets, meaning beliefs that attributes can change with effort, enjoy many positive outcomes, including higher academic achievement, while people with fixed mind-sets experience negative ones2 . They add that, given this relationship, interventions meant to raise students’ growth mind-sets, and through that their achievement, have been put in place in schools around the world2 .
The 2019 national experiment
Yeager and colleagues, including Carol Dweck, reported the National Study of Learning Mindsets in Nature in 20191 . It was a stratified random sample of 65 regular public schools in the United States, with 12,490 ninth-grade students who were individually randomized to condition1 . The intervention was a short online program, under an hour in total, that teaches that intellectual abilities can be developed, and it uses the metaphor that the brain is like a muscle that grows stronger with rigorous learning1 . The paper says the analyses were pre-registered, that data were collected and processed by independent companies, and that a blinded Bayesian analysis was used as a robustness check1 .
The pre-registered main result concerned the 6,320 students who were lower-achieving relative to peers in the same school, defined relative to the school median grade point average in the term before assignment1 . Among them, those assigned to the intervention earned higher core-course grades at the end of ninth grade, by 0.10 grade points (95 percent confidence interval 0.04 to 0.16), a standardized effect size of 0.111 . Higher-achieving students showed no significant effect on core-course grades, 0.01 grade points (confidence interval -0.03 to 0.06)1 . The authors also report that the intervention reduced fixed-mindset beliefs, which they measured at the end of the second session1 .
The paper separates its pre-registered findings from an exploratory one1 . In 41 schools that supplied course data, the intervention raised the likelihood of taking advanced mathematics (algebra II or higher) in tenth grade by 3 percentage points, an analysis the paper says was not pre-registered and is therefore exploratory1 . The authors also estimate with a model that the intervention would cut the share of lower-achieving students ending ninth grade with a grade point average below 2.0 from 46 percent to 41 percent if applied across the United States, which is a model estimate and not an observed national result1 .
Where the effect was larger
The effect on grades varied from school to school, and the paper tested two pre-registered explanations1 . Effects among lower-achieving students were smaller in higher-achieving schools, and they were larger in schools where peer norms supported seeking out challenges, a norm the authors measured with a behavioral task given to the control group1 . The authors state a limitation of that second result: because the norms were measured and not manipulated, causal conclusions about the school norm cannot be drawn1 . They also write that not all forms of growth mindset intervention can be expected to raise grades or advanced course-taking, even in the targeted groups1 .
The 2018 meta-analyses
Sisk and colleagues ran two meta-analyses, the first on the link between mind-set and academic achievement (273 studies, 365,915 participants) and the second on mind-set interventions (43 studies, 57,155 participants)2 . Their abstract says the overall effects were weak in both2 . It adds that some results supported specific tenets of the theory, namely that students with low socioeconomic status or who are academically at risk might benefit from mind-set interventions2 .
The 2023 meta-analysis
Macnamara and Burgoyne published a systematic review and meta-analysis in Psychological Bulletin in 2023 that covered 63 studies with 97,672 participants3 . They report major shortcomings in study design, analysis and reporting, and suggestions of researcher and publication bias, noting that authors with a financial incentive to report positive findings published significantly larger effects than authors without one3 . Across all studies the overall effect was small (d of 0.05, 95 percent confidence interval 0.02 to 0.09) and became nonsignificant after correction for potential publication bias3 .
Restricting the analysis to 13 studies (18,355 participants) that showed the intervention changed students’ mindsets as intended gave a nonsignificant d of 0.043 . Restricting it to the 6 highest-quality studies (13,571 participants) gave a nonsignificant d of 0.023 . The authors concluded that the apparent effects of growth mindset interventions on academic achievement are likely attributable to inadequate study design, reporting flaws and bias3 .
The reply
A commentary in the same journal, whose authors include Elizabeth Tipton and David Yeager, compared the 2023 meta-analysis with a second one by Burnette and colleagues that examined the same literature and reached different conclusions4 . It says the first used a traditional approach, averaging effect sizes within each study and testing moderators one at a time, while the second modeled variation in effects within studies and across subgroups and outcomes4 . The commentary argues that this heterogeneity-attuned approach is the better practice, and reports an exploratory re-analysis of the Macnamara and Burgoyne data that, in its authors’ words, confirmed a meaningful, significant effect in focal (at-risk) groups4 . Yeager was the first author of the 2019 national experiment, so the commentary comes from authors with a stake in the question. Only the abstracts of the 2018 meta-analyses, the 2023 meta-analysis and this commentary were read for this article, and the second meta-analysis it mentions was not opened.
Where the evidence stands
The papers agree on some points and split on others. A pre-registered national experiment found a modest grade gain for lower-achieving students, concentrated in some kinds of schools. Two meta-analyses found weak or near-zero average effects across the wider literature, and the 2023 one attributes the positive findings to design flaws and bias. The 2023 commentary answers that averages can hide real differences between groups. So the claim that a growth mindset lifts achievement is neither settled nor refuted. What the papers dispute is whether the average effect is close to zero or whether it is real for particular students and schools.
This article draws no advice for any student, parent or teacher from these papers. The photo at the top shows a classroom at Oregon State University, chosen as a general picture of a teaching room, and has no connection to the schools in the papers above.
Trivia
- In which journal did the 2019 National Study of Learning Mindsets appear? Nature1.
- How many public schools were in that national experiment? 651.
- By how many grade points did lower-achieving students in the intervention group gain in core courses, per the 2019 paper? 0.10 grade points1.
- Which journal published the 2018 pair of growth mindset meta-analyses by Sisk and colleagues? Psychological Science2.
- What overall effect size (d) did the 2023 Macnamara and Burgoyne meta-analysis report across all studies? 0.053.
Sources
- Yeager DS, Hanselman P, Walton GM, et al. A national experiment reveals where a growth mindset improves achievement. Nature, 573(7774), 364-369, 2019. Accessed November 6, 2026.
- Sisk VF, Burgoyne AP, Sun J, Butler JL, Macnamara BN. To what extent and under which circumstances are growth mind-sets important to academic achievement? Two meta-analyses. Psychological Science, 29(4), 549-571, 2018. Accessed November 6, 2026.
- Macnamara BN, Burgoyne AP. Do growth mindset interventions impact students' academic achievement? A systematic review and meta-analysis with recommendations for best practices. Psychological Bulletin, 149(3-4), 133-173, 2023. Accessed November 6, 2026.
- Tipton E, Bryan C, Murray J, McDaniel M, Schneider B, Yeager DS. Why meta-analyses of growth mindset and other interventions should follow best practices for examining heterogeneity: commentary on Macnamara and Burgoyne (2023) and Burnette et al. (2023). Psychological Bulletin, 149(3-4), 229-241, 2023. Accessed November 6, 2026.