Growth mindset: what the replication crisis did to one of education's biggest ideas

Carol Dweck spent the 1980s and 1990s studying how people think about their own ability. Some treated intelligence as fixed; others treated it as malleable. The two groups responded to challenges differently. Her synthesis — what she eventually called growth mindset — became one of the most influential ideas in twentieth-century educational psychology.

Then the replication crisis arrived.

1. The original story

Dweck and colleagues' research, summarized in her 2006 book Mindset, made two claims:

1. People differ in their implicit theory of intelligence — fixed vs. growth
2. Brief interventions teaching students that intelligence is malleable can improve academic performance, particularly for students at risk

The interventions were small — 30-60 minute single sessions teaching neuroplasticity concepts — but the effects on grades were striking in the original studies, particularly for low-achieving students (Blackwell, Trzesniewski, & Dweck, 2007).

The framework became enormously influential. School districts ran growth-mindset training. Corporate trainings adopted it. Parenting books rewrote praise practices ("praise effort, not ability").

2. The replication record

Several major studies have attempted to replicate the academic-performance effects:

Sisk et al. (2018): Meta-analysis of 273 effect sizes. Found very small effects on academic outcomes overall, with substantial variation. Effects largest in high-risk samples; near-zero in general populations (Sisk et al., 2018).

Yeager et al. (2019): Large-scale RCT in the United States with 12,000+ ninth-graders. Found small but significant effects on low-achieving students' grades. Effect was concentrated in supportive school environments. Did not replicate in unsupportive ones (Yeager et al., 2019).

Macnamara & Burgoyne (2023): Another meta-analysis, more critical. Argued that publication bias and methodological inflation accounted for much of the apparent effect. Concluded effects were near-zero after correction (Macnamara & Burgoyne, 2023).

3. The careful summary

The strong original claims — that brief growth-mindset interventions can transform academic trajectories — are not supported by the replication record at full strength. The weaker claims — that for high-risk students in supportive environments, mindset training produces small reliable gains — are partially supported.

The general population effects, in well-controlled studies, are small. The interventions are cheap. Whether they're worth deploying depends on values and context, not just effect sizes.

4. The framework vs. the intervention

There's a useful distinction often muddled in this discussion. Growth-mindset training as a clinical intervention is the thing under empirical scrutiny. The broader cultural framing of effort vs. talent is a different question with mixed evidence.

The intervention literature has been disappointing. The descriptive framework — people do differ in how they think about ability, and those differences correlate with persistence under challenge — remains reasonably well-supported as a psychological phenomenon (Burgoyne, Hambrick, & Macnamara, 2020).

5. The honest takeaway

For educators considering growth-mindset interventions: they probably help a little, in some contexts, for some students. They are not the educational silver bullet the 2010s hype suggested.

For individual readers: the descriptive claim that believing ability is malleable tends to produce more persistence than believing ability is fixed is reasonable. The implementation effect of a 30-minute intervention to shift that belief is smaller and more contingent than the popular framing claims.

This is a common pattern: useful descriptive psychology converted into prescriptive intervention has often proven much harder than the original observation suggested.