Walk into almost any school in the United Kingdom and you will find growth mindset posters on the walls. "Mistakes help us grow." "The power of yet." "You can train your brain." These phrases have become so embedded in school culture that questioning them feels like questioning the value of effort itself. But what does the research actually show?
Carol Dweck's growth mindset framework, first popularised in her 2006 book Mindset: The New Psychology of Success, proposed a compelling idea: learners who believe their intelligence can be developed through effort and good strategies (growth mindset) outperform those who believe intelligence is fixed and unchangeable (fixed mindset). The implication was straightforward and appealing. Teach learners that their brains can grow, and they will work harder, persist longer, and achieve more.
Schools across the UK adopted this idea enthusiastically. Growth mindset became a cornerstone of personal development curricula, assembly themes, and behaviour policies. Teachers were trained to praise effort over ability. Entire school improvement plans were built around shifting learner mindsets. But between the original research and the classroom application, something important got lost.
Dweck's original studies in the 1990s and early 2000s demonstrated correlations between mindset beliefs and academic outcomes. Learners who endorsed growth mindset statements tended to show greater resilience after setbacks and, in some studies, achieved better grades. These were genuinely important findings that opened a new line of enquiry in educational psychology.
However, correlation is not causation. Showing that learners with growth mindsets tend to perform better is not the same as proving that teaching growth mindset causes better performance. The critical question for schools was always: if we intervene to change mindsets, will achievement follow? This is where the evidence becomes considerably more complicated.
Dweck's early intervention studies were conducted with relatively small samples under controlled conditions. When independent researchers attempted to replicate these findings at scale, the results were far less convincing. The replication crisis that has affected psychology more broadly has been particularly significant for growth mindset research.
Li and Bates (2019) conducted a rigorous replication of Dweck's core claims, published in the Journal of Experimental Psychology: General. Their findings were stark: growth mindset interventions changed learners' beliefs about intelligence (they came to endorse the idea that intelligence is malleable) but these new beliefs had zero significant effect on resilience to failure, cognitive ability, or educational attainment. Learners said the right things about growth mindset but did not perform any differently.
The most comprehensive challenge came from Sisk, Burgoyne, Sun, Taylor, and Macnamara (2018), who published two meta-analyses in Psychological Science. Across dozens of studies involving thousands of participants, they found that the overall effect size of growth mindset interventions on academic achievement was just 0.08. To put this in context, an effect size of 0.08 is considered negligible in educational research. For comparison, the effect of getting adequate sleep on academic performance is approximately 0.3, nearly four times larger.
For typical learners in typical school settings, growth mindset interventions frequently failed to reach statistical significance. The meta-analysis did find slightly larger effects for lower-achieving learners, but even these were modest and inconsistent across studies.
Burgoyne, Hambrick, and Macnamara (2020) went further in Psychological Science, re-analysing Dweck's own published studies. They found that some core claims were unsupported by the data, effect sizes were weaker than originally presented, and conclusions had been overstated. The foundations of mindset theory, they concluded, were shakier than the public had been led to believe.
Not all the evidence is negative. Yeager, Hanselman, Walton, and colleagues published a large national experiment in Nature in 2019 that identified specific conditions under which growth mindset interventions can produce real, if small, effects.
The study found that growth mindset interventions worked for lower-achieving learners who were enrolled in challenging, advanced courses such as honours mathematics. Crucially, the interventions only worked in schools where the peer culture already valued effort and treated mistakes as learning opportunities. In schools without this supportive culture, the interventions had no effect.
The effect sizes in Yeager's study ranged from approximately 0.1 to 0.2, which are small but meaningful. The key insight is that context matters enormously. A growth mindset intervention delivered as a one-off assembly or a poster campaign in a school that does not structurally support persistence and challenge is unlikely to change anything.
What made the difference in successful implementations was the combination of mindset messaging with specific academic challenge, peer models of persistence, and reattribution training that taught learners to interpret setbacks as signals to try a different strategy rather than as evidence of inability.
To her credit, Dweck herself has publicly acknowledged the oversimplification of her work in schools. In a 2015 piece for Education Week, she introduced the concept of "false growth mindset" to describe the gap between what she intended and what schools were actually doing.
Dweck identified several problems with how schools had implemented her ideas:
Dweck's self-critique is important because it confirms what the replication studies found independently: the simplistic version of growth mindset that most schools adopted does not work. The theory is more nuanced than the posters suggest.
For UK teachers, the most relevant evidence comes from the Education Endowment Foundation's (2019) Changing Mindsets evaluation, one of the largest randomised controlled trials of growth mindset interventions conducted in English schools.
The findings were clear:
The disconnect is striking. Learners learned the concept of growth mindset and endorsed it when surveyed, but this new belief did not translate into different behaviour or better academic outcomes. The EEF concluded that "schools should be cautious about using growth mindset interventions alone as a mechanism for boosting academic attainment." See also our guide on visual thinking routines.
This does not mean mindset is irrelevant. It means that changing beliefs without changing the structures, strategies, and teaching practices that support learning is insufficient. A learner who believes they can improve but does not know how to improve, and is not given appropriately challenging work with adequate scaffolding, will not achieve more simply because of their beliefs.
The evidence does not suggest that mindset is meaningless. It suggests that mindset alone is not enough. The research points toward a more integrated approach that combines mindset messaging with concrete support for learning.
Mindset plus strategy. Rather than simply telling learners they can improve, teach them how to improve. Retrieval practice, metacognitive planning, and deliberate practice are evidence-based techniques that give learners the tools to translate effort into progress. When learners experience genuine improvement through specific strategies, growth mindset becomes self-reinforcing.
Mindset plus challenge. Growth mindset is most meaningful when learners face genuine difficulty. If the work is easy, effort is irrelevant. Schools that pair mindset messaging with a culture of high expectations and appropriately challenging curricula create conditions where persistence actually matters and pays off. For related guidance, see our article on translating IB for parents.
Mindset plus peer models. Yeager's research shows that peer culture is a powerful moderator. When learners see their classmates persist through difficulty, struggle openly, and improve through effort, growth mindset becomes a lived experience rather than an abstract concept. Teachers can facilitate this through collaborative learning, visible thinking routines, and normalising productive struggle.
Mindset plus feedback. The quality of feedback matters enormously. Feedback that is specific, actionable, and focused on the process of learning (not just the outcome) supports both growth mindset and actual improvement. Telling a learner "try harder" is not feedback. Telling them "your argument would be stronger if you added a counter-example in paragraph two" gives them something to act on.
Mindset plus safe culture. Learners will not take risks, make mistakes, or persist through difficulty if the classroom culture punishes failure. Schools that genuinely support growth mindset create environments where errors are treated as data, not as deficiencies. This requires changes to marking practices, classroom talk, and the way adults respond to wrong answers.
If your school has invested in growth mindset programmes, you have not wasted your time entirely. The underlying principle that learners' beliefs about their own capacity matter is well-established. What the evidence challenges is the idea that changing those beliefs through posters, assemblies, and standalone interventions will improve academic outcomes.
Consider these practical steps:
A related challenge for schools is the overjustification effect, first described by Lepper, Greene, and Nisbett (1973). When learners receive external rewards for activities they already find intrinsically motivating, their internal motivation can decrease. This is directly relevant to growth mindset implementation because many schools pair mindset messaging with reward systems such as stickers, certificates, and house points for demonstrating effort.
The research suggests this combination can be counterproductive. Learners may begin to associate effort with the reward rather than with the intrinsic satisfaction of learning and improving. When the rewards are removed, effort decreases. Deci and Ryan's (1985) self-determination theory provides a more robust framework, suggesting that learners are most motivated when they experience autonomy (choice in their learning), competence (evidence of genuine improvement), and relatedness (connection to teachers and peers who value learning).
Schools would benefit from examining whether their growth mindset programmes inadvertently rely on extrinsic motivation systems that contradict the underlying message. If the message is that effort leads to growth, but the mechanism for encouraging effort is a sticker chart, the structural incentive undermines the psychological principle.
One area where growth mindset messaging requires particular care is in relation to learners with special educational needs and disabilities (SEND). The blanket assertion that anyone can achieve anything through effort can be experienced as dismissive by learners whose difficulties are neurological, developmental, or cognitive in origin. A learner with dyslexia, for example, may interpret growth mindset messaging as an implication that their reading difficulties are caused by insufficient effort rather than a genuine processing difference.
Effective practice for SEND learners involves acknowledging both the reality of their specific challenges and the genuine capacity for growth within those parameters. This requires teachers to be precise about what growth looks like for individual learners rather than applying universal growth mindset language that may inadvertently blame the learner for difficulties beyond their control. The combination of growth mindset with realistic, differentiated expectations and targeted support is more constructive than mindset messaging alone.
One of the most common misunderstandings of Dweck's work is the treatment of mindset as a binary: you either have a growth mindset or a fixed mindset. Dweck has clarified that mindset exists on a continuum and varies across domains. A learner might hold a growth mindset about their ability in art but a fixed mindset about mathematics. A teacher might believe intelligence is malleable in general but hold fixed beliefs about certain learners' capacity for improvement.
This domain specificity has important implications for classroom practice. Rather than delivering generic growth mindset messaging, teachers are better served by identifying the specific domains and contexts where fixed mindset beliefs are most limiting and addressing those directly. A Year 9 learner who says "I am not a maths person" needs domain-specific reattribution (showing them that mathematical thinking develops with practice and specific strategies) rather than a poster about the power of yet.
The evidence from cognitive science supports this targeted approach. Cognitive load theory demonstrates that effective learning requires appropriate challenge at the boundary of current understanding. When tasks are properly calibrated and learners experience genuine progress through effort, growth mindset develops as a natural consequence rather than requiring explicit instruction.
Growth mindset is not irrelevant, but it is not the transformative intervention many schools believed it to be. The evidence shows that mindset beliefs alone do not improve achievement. What matters is combining growth-oriented beliefs with specific learning strategies, genuine challenge, and supportive school culture. Mindset is one ingredient, not the whole recipe.
Focus on evidence-based strategies with stronger effect sizes. The EEF's Teaching and Learning Toolkit identifies metacognition and self-regulation (+7 months), feedback (+6 months), and collaborative learning (+5 months) as high-impact approaches. These give learners the tools to improve, rather than just the belief that improvement is possible.
Dweck's original findings about the correlation between mindset and achievement have been broadly supported. What has not been supported is the assumption that teaching mindset through interventions causes improved achievement. Li and Bates (2019), Sisk et al. (2018), and the EEF trial all found that changing beliefs did not change outcomes. Dweck herself acknowledges that schools have oversimplified her work.
Growth mindset is intuitively appealing, easy to explain to parents, and aligns with teachers' genuine desire to help every learner succeed. Commercial programmes and CPD providers have marketed mindset interventions aggressively. The theory also arrived during a period when schools were seeking non-academic approaches to improvement. These factors created a perfect storm of adoption that outpaced the evidence.
Growth mindset that works is embedded in school culture, paired with specific learning strategies, delivered alongside genuinely challenging content, and reinforced through quality feedback. Growth mindset that does not work is delivered as a standalone intervention, focuses on posters and language alone, praises effort without strategy, and operates in schools where the structural conditions for persistence are absent (Yeager et al., 2019).
Walk into almost any school in the United Kingdom and you will find growth mindset posters on the walls. "Mistakes help us grow." "The power of yet." "You can train your brain." These phrases have become so embedded in school culture that questioning them feels like questioning the value of effort itself. But what does the research actually show?
Carol Dweck's growth mindset framework, first popularised in her 2006 book Mindset: The New Psychology of Success, proposed a compelling idea: learners who believe their intelligence can be developed through effort and good strategies (growth mindset) outperform those who believe intelligence is fixed and unchangeable (fixed mindset). The implication was straightforward and appealing. Teach learners that their brains can grow, and they will work harder, persist longer, and achieve more.
Schools across the UK adopted this idea enthusiastically. Growth mindset became a cornerstone of personal development curricula, assembly themes, and behaviour policies. Teachers were trained to praise effort over ability. Entire school improvement plans were built around shifting learner mindsets. But between the original research and the classroom application, something important got lost.
Dweck's original studies in the 1990s and early 2000s demonstrated correlations between mindset beliefs and academic outcomes. Learners who endorsed growth mindset statements tended to show greater resilience after setbacks and, in some studies, achieved better grades. These were genuinely important findings that opened a new line of enquiry in educational psychology.
However, correlation is not causation. Showing that learners with growth mindsets tend to perform better is not the same as proving that teaching growth mindset causes better performance. The critical question for schools was always: if we intervene to change mindsets, will achievement follow? This is where the evidence becomes considerably more complicated.
Dweck's early intervention studies were conducted with relatively small samples under controlled conditions. When independent researchers attempted to replicate these findings at scale, the results were far less convincing. The replication crisis that has affected psychology more broadly has been particularly significant for growth mindset research.
Li and Bates (2019) conducted a rigorous replication of Dweck's core claims, published in the Journal of Experimental Psychology: General. Their findings were stark: growth mindset interventions changed learners' beliefs about intelligence (they came to endorse the idea that intelligence is malleable) but these new beliefs had zero significant effect on resilience to failure, cognitive ability, or educational attainment. Learners said the right things about growth mindset but did not perform any differently.
The most comprehensive challenge came from Sisk, Burgoyne, Sun, Taylor, and Macnamara (2018), who published two meta-analyses in Psychological Science. Across dozens of studies involving thousands of participants, they found that the overall effect size of growth mindset interventions on academic achievement was just 0.08. To put this in context, an effect size of 0.08 is considered negligible in educational research. For comparison, the effect of getting adequate sleep on academic performance is approximately 0.3, nearly four times larger.
For typical learners in typical school settings, growth mindset interventions frequently failed to reach statistical significance. The meta-analysis did find slightly larger effects for lower-achieving learners, but even these were modest and inconsistent across studies.
Burgoyne, Hambrick, and Macnamara (2020) went further in Psychological Science, re-analysing Dweck's own published studies. They found that some core claims were unsupported by the data, effect sizes were weaker than originally presented, and conclusions had been overstated. The foundations of mindset theory, they concluded, were shakier than the public had been led to believe.
Not all the evidence is negative. Yeager, Hanselman, Walton, and colleagues published a large national experiment in Nature in 2019 that identified specific conditions under which growth mindset interventions can produce real, if small, effects.
The study found that growth mindset interventions worked for lower-achieving learners who were enrolled in challenging, advanced courses such as honours mathematics. Crucially, the interventions only worked in schools where the peer culture already valued effort and treated mistakes as learning opportunities. In schools without this supportive culture, the interventions had no effect.
The effect sizes in Yeager's study ranged from approximately 0.1 to 0.2, which are small but meaningful. The key insight is that context matters enormously. A growth mindset intervention delivered as a one-off assembly or a poster campaign in a school that does not structurally support persistence and challenge is unlikely to change anything.
What made the difference in successful implementations was the combination of mindset messaging with specific academic challenge, peer models of persistence, and reattribution training that taught learners to interpret setbacks as signals to try a different strategy rather than as evidence of inability.
To her credit, Dweck herself has publicly acknowledged the oversimplification of her work in schools. In a 2015 piece for Education Week, she introduced the concept of "false growth mindset" to describe the gap between what she intended and what schools were actually doing.
Dweck identified several problems with how schools had implemented her ideas:
Dweck's self-critique is important because it confirms what the replication studies found independently: the simplistic version of growth mindset that most schools adopted does not work. The theory is more nuanced than the posters suggest.
For UK teachers, the most relevant evidence comes from the Education Endowment Foundation's (2019) Changing Mindsets evaluation, one of the largest randomised controlled trials of growth mindset interventions conducted in English schools.
The findings were clear:
The disconnect is striking. Learners learned the concept of growth mindset and endorsed it when surveyed, but this new belief did not translate into different behaviour or better academic outcomes. The EEF concluded that "schools should be cautious about using growth mindset interventions alone as a mechanism for boosting academic attainment." See also our guide on visual thinking routines.
This does not mean mindset is irrelevant. It means that changing beliefs without changing the structures, strategies, and teaching practices that support learning is insufficient. A learner who believes they can improve but does not know how to improve, and is not given appropriately challenging work with adequate scaffolding, will not achieve more simply because of their beliefs.
The evidence does not suggest that mindset is meaningless. It suggests that mindset alone is not enough. The research points toward a more integrated approach that combines mindset messaging with concrete support for learning.
Mindset plus strategy. Rather than simply telling learners they can improve, teach them how to improve. Retrieval practice, metacognitive planning, and deliberate practice are evidence-based techniques that give learners the tools to translate effort into progress. When learners experience genuine improvement through specific strategies, growth mindset becomes self-reinforcing.
Mindset plus challenge. Growth mindset is most meaningful when learners face genuine difficulty. If the work is easy, effort is irrelevant. Schools that pair mindset messaging with a culture of high expectations and appropriately challenging curricula create conditions where persistence actually matters and pays off. For related guidance, see our article on translating IB for parents.
Mindset plus peer models. Yeager's research shows that peer culture is a powerful moderator. When learners see their classmates persist through difficulty, struggle openly, and improve through effort, growth mindset becomes a lived experience rather than an abstract concept. Teachers can facilitate this through collaborative learning, visible thinking routines, and normalising productive struggle.
Mindset plus feedback. The quality of feedback matters enormously. Feedback that is specific, actionable, and focused on the process of learning (not just the outcome) supports both growth mindset and actual improvement. Telling a learner "try harder" is not feedback. Telling them "your argument would be stronger if you added a counter-example in paragraph two" gives them something to act on.
Mindset plus safe culture. Learners will not take risks, make mistakes, or persist through difficulty if the classroom culture punishes failure. Schools that genuinely support growth mindset create environments where errors are treated as data, not as deficiencies. This requires changes to marking practices, classroom talk, and the way adults respond to wrong answers.
If your school has invested in growth mindset programmes, you have not wasted your time entirely. The underlying principle that learners' beliefs about their own capacity matter is well-established. What the evidence challenges is the idea that changing those beliefs through posters, assemblies, and standalone interventions will improve academic outcomes.
Consider these practical steps:
A related challenge for schools is the overjustification effect, first described by Lepper, Greene, and Nisbett (1973). When learners receive external rewards for activities they already find intrinsically motivating, their internal motivation can decrease. This is directly relevant to growth mindset implementation because many schools pair mindset messaging with reward systems such as stickers, certificates, and house points for demonstrating effort.
The research suggests this combination can be counterproductive. Learners may begin to associate effort with the reward rather than with the intrinsic satisfaction of learning and improving. When the rewards are removed, effort decreases. Deci and Ryan's (1985) self-determination theory provides a more robust framework, suggesting that learners are most motivated when they experience autonomy (choice in their learning), competence (evidence of genuine improvement), and relatedness (connection to teachers and peers who value learning).
Schools would benefit from examining whether their growth mindset programmes inadvertently rely on extrinsic motivation systems that contradict the underlying message. If the message is that effort leads to growth, but the mechanism for encouraging effort is a sticker chart, the structural incentive undermines the psychological principle.
One area where growth mindset messaging requires particular care is in relation to learners with special educational needs and disabilities (SEND). The blanket assertion that anyone can achieve anything through effort can be experienced as dismissive by learners whose difficulties are neurological, developmental, or cognitive in origin. A learner with dyslexia, for example, may interpret growth mindset messaging as an implication that their reading difficulties are caused by insufficient effort rather than a genuine processing difference.
Effective practice for SEND learners involves acknowledging both the reality of their specific challenges and the genuine capacity for growth within those parameters. This requires teachers to be precise about what growth looks like for individual learners rather than applying universal growth mindset language that may inadvertently blame the learner for difficulties beyond their control. The combination of growth mindset with realistic, differentiated expectations and targeted support is more constructive than mindset messaging alone.
One of the most common misunderstandings of Dweck's work is the treatment of mindset as a binary: you either have a growth mindset or a fixed mindset. Dweck has clarified that mindset exists on a continuum and varies across domains. A learner might hold a growth mindset about their ability in art but a fixed mindset about mathematics. A teacher might believe intelligence is malleable in general but hold fixed beliefs about certain learners' capacity for improvement.
This domain specificity has important implications for classroom practice. Rather than delivering generic growth mindset messaging, teachers are better served by identifying the specific domains and contexts where fixed mindset beliefs are most limiting and addressing those directly. A Year 9 learner who says "I am not a maths person" needs domain-specific reattribution (showing them that mathematical thinking develops with practice and specific strategies) rather than a poster about the power of yet.
The evidence from cognitive science supports this targeted approach. Cognitive load theory demonstrates that effective learning requires appropriate challenge at the boundary of current understanding. When tasks are properly calibrated and learners experience genuine progress through effort, growth mindset develops as a natural consequence rather than requiring explicit instruction.
Growth mindset is not irrelevant, but it is not the transformative intervention many schools believed it to be. The evidence shows that mindset beliefs alone do not improve achievement. What matters is combining growth-oriented beliefs with specific learning strategies, genuine challenge, and supportive school culture. Mindset is one ingredient, not the whole recipe.
Focus on evidence-based strategies with stronger effect sizes. The EEF's Teaching and Learning Toolkit identifies metacognition and self-regulation (+7 months), feedback (+6 months), and collaborative learning (+5 months) as high-impact approaches. These give learners the tools to improve, rather than just the belief that improvement is possible.
Dweck's original findings about the correlation between mindset and achievement have been broadly supported. What has not been supported is the assumption that teaching mindset through interventions causes improved achievement. Li and Bates (2019), Sisk et al. (2018), and the EEF trial all found that changing beliefs did not change outcomes. Dweck herself acknowledges that schools have oversimplified her work.
Growth mindset is intuitively appealing, easy to explain to parents, and aligns with teachers' genuine desire to help every learner succeed. Commercial programmes and CPD providers have marketed mindset interventions aggressively. The theory also arrived during a period when schools were seeking non-academic approaches to improvement. These factors created a perfect storm of adoption that outpaced the evidence.
Growth mindset that works is embedded in school culture, paired with specific learning strategies, delivered alongside genuinely challenging content, and reinforced through quality feedback. Growth mindset that does not work is delivered as a standalone intervention, focuses on posters and language alone, praises effort without strategy, and operates in schools where the structural conditions for persistence are absent (Yeager et al., 2019).
