What 42,000 Participants Tell Us About How Growth Mindset Predicts Grit

A Number You Can Anchor On

Growth-mindset writing online tends to oscillate between two unhelpful poles. Either the mindset is treated as a near-magical lever that quietly determines whether you succeed at anything difficult, or it is dismissed as a self-help trope that the replication crisis has put to bed. Neither framing tells you what to do on a Tuesday morning when you're trying to stay with a hard problem.

The 2025 meta-analysis by K. K. L. Lam and M. Zhou, published in Acta Psychologica (Volume 255, Article 104872), is one of the cleanest places to anchor between those two poles. The authors aggregated 66 primary studies covering 42,112 participants and asked a narrow question: how strongly does a growth mindset correlate with grit, the construct Angela Duckworth defines as sustained passion and perseverance for long-term goals?

The headline numbers are these. The overall correlation between growth mindset and grit was ρ = 0.19. Broken out by Duckworth's two grit subcomponents, the correlation with consistency of interest was ρ = 0.20, and the correlation with perseverance of effort was ρ = 0.24.

Those are real correlations. They are also small-to-moderate, not large. The honest interpretation lives in the middle: a growth mindset is one of several things that predicts whether someone keeps going when the work gets hard, but it is not the dominant factor. This article is about what that distinction means in practice.

Reading the Effect Sizes Without Inflating or Dismissing Them

A correlation of 0.24 between mindset and effort sounds modest because, in absolute terms, it is. Squared, it accounts for roughly 5.8% of the variance in perseverance of effort. The remaining 94% lives in other factors: the person's existing skills, their environment, their physical state, the structure of the task, the people around them, and a long list of other variables that mindset does not directly address.

This is where most popular writing on mindset goes wrong in one direction or the other. The inflated reading is that mindset "drives" perseverance. The numbers do not support that. The dismissive reading is that mindset is "irrelevant" or "debunked." The numbers do not support that either. A reliable correlation of around 0.2 across more than 40,000 participants is a real signal, not noise. It is just smaller than the marketing language around growth mindset has historically suggested.

The most useful framing comes from looking at the comparison. Many factors that researchers and practitioners take seriously have correlations in this range with downstream outcomes. The correlation between conscientiousness and academic performance, for instance, sits in roughly the same band. We do not say "conscientiousness is debunked" because the correlation is 0.2 to 0.3. We say "conscientiousness is one meaningful predictor among several." The same framing applies here.

Why the Correlation Is Stronger With Effort Than With Interest

The two subcomponents of grit behave differently in the meta-analysis. Growth mindset correlates ρ = 0.24 with perseverance of effort and ρ = 0.20 with consistency of interest. The gap is small but conceptually telling.

Perseverance of effort, in Duckworth's framework, is the willingness to keep working on something hard. Consistency of interest is the willingness to stay with the same goal across long stretches of time, even when newer goals are pulling for attention.

A growth mindset speaks more directly to the first than the second. The belief that ability can grow with effort is a belief about the value of effort. It does not tell you which goal to point that effort at, or whether to switch. It tells you that the work is not wasted while you are doing it.

This is consistent with what longitudinal research on grit has found. Building Grit, a 2019 study in the Journal of Youth and Adolescence, traced the longitudinal pathways from mindset through commitment to grit and academic outcomes. Mindset predicted commitment, commitment predicted grit, grit predicted outcomes. The chain is real. But each link in it is moderate, not deterministic, and consistency-of-interest sits further down the chain than perseverance-of-effort does.

What the Mindset Intervention Literature Says, Honestly

Two important meta-analyses in 2022 reached different conclusions about whether short growth-mindset interventions actually improve student academic outcomes. One, by Macnamara and colleagues, concluded that the apparent effects were largely attributable to study-design weaknesses and publication bias, and that the highest-quality studies showed essentially no effect (d = 0.02, not significantly different from zero). Another, by Burnette and colleagues, found small but meaningful positive effects, particularly for specific subgroups.

The more recent 2025 review by Gazmuri and colleagues in the Review of Education, looking systematically at the existing evidence, leans toward the cautious interpretation. Short, scalable mindset interventions appear to produce small-to-zero effects on average academic achievement. The 2019 Yeager and colleagues national experiment in Nature, which is one of the better-designed studies in the field, found measurable improvements concentrated in lower-achieving students in supportive school cultures, with smaller or null effects elsewhere.

The 2025 Lam and Zhou meta-analysis fits cleanly into this picture. A growth mindset is a real correlate of grit. Brief interventions to change someone's mindset, however, produce small effects on the outcomes that follow grit, like grades. The correlations exist. The cheap, fast routes to changing them often do not.

This is the honest version of where the literature sits. For deeper context, see our companion piece on why growth mindset alone is not enough, which walks through the contested intervention findings in more detail.

What This Means For You On A Tuesday Morning

The practical implication of these effect sizes is not "growth mindset doesn't matter." It is "growth mindset matters, modestly, and the place it matters most is in how you stay with hard work."

Three concrete uses of this finding stand up to the data.

Use Growth Mindset Where It Has the Strongest Pull

The 0.24 correlation with perseverance of effort is the strongest signal in the meta-analysis. The most leveraged application of growth mindset is therefore in the moment when you are deciding whether to keep going on a difficult task. Not in motivation generally, not in goal-setting generally, but in the specific micro-decision of whether to continue with a hard piece of work for another twenty minutes.

The internal move is small. When you notice the impulse to stop because the work is uncomfortable, label the discomfort as the condition under which the mindset can do its job. The discomfort is the wedge through which growth happens. This framing is consistent with what the synergistic mindsets research (see growth mindset paired with stress reappraisal) has shown about the value of pairing the mindset with a useful interpretation of physical activation.

Don't Expect Mindset to Solve Goal-Selection Problems

The weaker correlation with consistency of interest, ρ = 0.20, is the part of grit that mindset does not address well. If you keep starting goals and abandoning them, that is not a mindset problem. It is a problem with how you are choosing goals, what your environment rewards, or what you actually want.

This matters because a lot of growth-mindset advice tries to apply the framework to motivation in general, including the question of which projects to pursue. The correlations suggest mindset is not the right tool for this. For the question "should I keep at this particular goal?" you are better served by tools like values clarification, environmental design, and explicit cost-benefit analysis than by mindset reframing.

Use the Mindset Frame, But Don't Mistake It For The Habit

Mindset is a belief. Grit is a pattern of behaviour over years. A correlation of ρ = 0.19 between the two means that having the belief makes the pattern modestly more likely, but the pattern still has to be built behaviour by behaviour. The 2009 University College London research by Phillippa Lally and colleagues on habit formation found that habits take a median of 66 days to form, with significant individual variation. Mindset is the orientation. Habits are the substrate. Both are needed, but they are not the same.

For the underlying mechanics of the second piece, see our piece on the psychology of habit formation, which goes into the specific design of small reliable behaviours rather than the orientation toward them.

Where Growth Mindset Has the Strongest Practical Case

Stepping back, the cleanest version of what mindset is good for emerges from looking at where the correlations are strongest and where the intervention literature is most positive.

The strongest practical case is for moments of acute difficulty in pursuit of a clear goal. Not for goal-setting. Not for vague life direction. Not as a substitute for skill, environment, or rest. For the specific psychological micro-moment when a piece of hard work is uncomfortable and you are deciding whether to push through. In that moment, the belief that effort produces growth makes the decision to push through measurably more likely. That is what the data supports, and it is enough to be useful.

For leaders who care about how this translates into team performance and how to build the surrounding conditions that let mindset do its job, The Leader's Table covers the management side of this question, including how to give feedback in a way that reinforces the mindset rather than undermining it.

A One-Sentence Takeaway

A growth mindset is one of several modest predictors of whether you stay with hard work. The honest expectation is that, in the right moment and with the right surrounding habits, it tilts the decision in your favour by something like a 5 to 7 percent margin. Not transformational. Not nothing. Useful.

Use it where it has the strongest pull. Build the habits and environment around it that the correlations alone cannot replace.