Carol Dweck’s pioneering research on growth mindset demonstrated that individuals’ beliefs about the malleability of intelligence (whether or not academic ability can be changed) have important implications for development (Dweck, 2006[4]). For example, a growth mindset has predicted achievement (Blackwell et al., 2007) and long-term educational attainment such as postsecondary enrolment (Hecht et al., 2024[34]). Intervention programs that foster a growth mindset have influenced student motivation, challenge-seeking, and achievement (Blackwell, Trzesniewski and Dweck, 2007[35]; Yeager et al., 2019[14]). This foundational work inspired extensive educational experimentation and led to large-scale efforts to foster adaptive learning beliefs (Dweck and Yeager, 2019[8]; Hecht, 2021[36]). Yet, as the evidence base has expanded, research has sought to put research on growth mindset into context, by showing that mindset effects on achievement vary depending on learner profiles such as students’ prior achievement levels (Paunesku et al., 2015[37]), school contexts (Yeager et al., 2019[14]; Lou and Li, 2022[38]), the teacher’s classroom culture (Hecht et al., 2022[39]), and program implementation quality (Dweck and Yeager, 2019[8]). Meta-analyses have documented substantial heterogeneity in findings (Sisk et al., 2018[40]; Burnette et al., 2023[41]), with treatment effects varying across diverse populations (Burnette et al., 2013[6]) and program implementation characteristics (Sisk et al., 2018[40]). This shift from universal optimism to contextual nuance underscores the continuing policy relevance of the concept: understanding when and where such beliefs support learning can inform more equitable (Claro, Paunesku and Dweck, 2016[15]; OECD, 2021[7]) and cost-effective interventions (Paunesku et al., 2015[37]).
Mindset theory stems from the study of implicit theories – the lay beliefs individuals hold about whether abilities are fixed or malleable (Burnette et al., 2013[6]; Dweck and Yeager, 2019[8]). Two core orientations are typically distinguished: entity theories (fixed mindsets), which view intelligence as innate and unchangeable, and incremental theories (growth mindsets), which regard intelligence as improvable through effort, strategy and support (Dweck, 2006[4]; Lou and Li, 2022[38]; Yeager et al., 2019[14]). These mindsets act as “meaning systems” that integrate people’s goals, beliefs and behaviours (Dweck and Yeager, 2019[8]). Large, representative-sample survey studies (see (Yeager and Dweck, 2020[42])) have found that individuals endorsing more of a fixed mindset tend to prioritise tasks with a low likelihood of failure and to avoid challenges that might expose limitations (because of their goal is to maintain the appearance or feeling of high ability). These are called performance-avoidance goals). In contrast, those holding more of a growth mindset tend to focus on developing competence and interpret setbacks as information for improvement (because of their goal of developing their abilities, called a learning goal) (Dweck, 2006[4]; Dweck and Yeager, 2019[8]). Those with more of a fixed mindset also tend to view effort as a bad sign that one lacks potential, because they attribute struggle or difficulty to a lack of ability. Moving beyond the core predictions from mindset theory, other studies have found that a growth mindset is linked to fewer negative emotions when learning and stronger expectations of success, although the magnitudes of those correlations are variable and depend on the specific instruments or populations in a study (Burnette et al., 2013[6]). Neuroscientific evidence further shows that growth mindsets are linked to greater attentional engagement with feedback and error correction (Lou and Li, 2022[38]; Moser et al., 2011[43]), and PISA data suggest a positive association between growth mindset and a stronger sense of belonging at school and positive feelings more generally (OECD, 2021[7]).
These mechanisms have inspired numerous interventions translating laboratory findings into more real-world programs and techniques (Yeager and Walton, 2011[44]; Paunesku et al., 2015[37]; Mueller and Dweck, 1998[45]; Aronson, Fried and Good, 2002[46]; Blackwell, Trzesniewski and Dweck, 2007[35]). Early studies demonstrated that teaching students about the malleability of intelligence (i.e., the idea that the “brain is like a muscle”) and that effort could improve one’s abilities, could increase persistence and academic outcomes (Dweck, 2006[4]; Dweck and Yeager, 2019[8]; Yeager and Walton, 2011[44]; Burnette et al., 2023[41]). Later, researchers designed brief, scalable online modules requiring minimal teacher input to maximise reach (Dweck and Yeager, 2019[8]; Paunesku et al., 2015[37]; Yeager et al., 2016[47]). Randomised field trials, such as the National Study of Learning Mindsets (Yeager et al., 2019[14]), showed that such interventions can improve grades and resilience, particularly among lower-achieving or disadvantaged students (Dweck and Yeager, 2019[8]; Paunesku et al., 2015[37]). Growth mindset has also been associated with higher enrolment in advanced mathematics courses (Yeager et al., 2019[14]) and stronger learning motivation among economically disadvantaged learners (Claro, Paunesku and Dweck, 2016[15]). International replications have confirmed these effects in diverse contexts such as large samples in Norway (Bettinger et al., 2018[48]; Rege et al., 2021[49]), illustrating that, when aligned with local conditions, mindset initiatives can contribute to reducing inequality (Dweck and Yeager, 2019[8]; Lou and Li, 2022[38]).
Nevertheless, growth mindset is not universally associated with outcomes in every context or for every population (Burnette et al., 2013[6]; Lou and Li, 2022[38]). Meta-analyses averaging across very different studies that reported a correlation between self-reported mindset and measured achievement find modest average associations (Sisk et al., 2018[40]; Lou and Li, 2022[38]; Costa and Faria, 2018[50]) with high heterogeneity (Sisk et al., 2018[40]; Burnette et al., 2013[6]; Lou and Li, 2022[38]). This mixed empirical landscape has driven closer scrutiny of the foundational claims and proposed mechanisms (Burnette et al., 2013[6]; Li and Bates, 2019[51]; Burnette et al., 2023[41]). Specifically, some small-sample longitudinal analyses conducted in China, where the PISA has found weak links between mindset and achievement, did not find an association between self-reported mindset and grade trajectories during the challenging transition from school to university, nor evidence that benefits are strongest for initially low achievers (Li and Bates, 2020[52]; Li and Bates, 2019[51]). These findings, which were also observed in other studies conducted in contexts like East Asia, align with the notion that growth mindset effects can be heterogeneous and context-dependent (Lou and Li, 2022[38]; OECD, 2021[7]). Indeed, at the same time, other longitudinal studies using large, representative samples in the U.S. have found meaningful links between self-reported mindset and educational trajectories, such as eventual postsecondary enrolment years later (Hecht et al., 2024[34]). In practice, wide dissemination of growth mindset has also led to misapplication –so-called “false growth mindset” – where teachers inaccurately equate the concept with praising effort rather than guiding students towards effective strategies (Dweck, 2006[4]; Dweck and Yeager, 2019[8]). Such findings underline the necessity of addressing heterogeneity and ensuring rigorous implementation and theoretical precision in policy design (Lou and Li, 2022[38]; Bryan, Tipton and Yeager, 2021[53]).
Recent research therefore reframes the discussion through the Mindset × Context paradigm (Dweck and Yeager, 2019[8]; Walton and Yeager, 2020[54]), which emphasises that the success of mindset interventions depends on the environmental conditions that allow beliefs to translate into adaptive action (Walton and Yeager, 2020[54]). Evidence shows that interventions are more consistently effective when peer and teacher norms reinforce effort and persistence (Yeager et al., 2019[14]; Walton and Yeager, 2020[54]). Teachers’ own mindsets shape classroom climates (Murphy and Dweck, 2009[55]): in certain contexts, fixed-mindset professors have been associated with achievement gaps among minority or vulnerable groups (Dweck and Yeager, 2019[8]; Canning et al., 2019[56]) but also with less supportive pedagogical practices, such as diminished process-focused teaching (Muenks et al., 2020[57]; Kroeper, Fried and Murphy, 2022[58]). At the societal level, mindset benefits have been found to be stronger in cultures where growth beliefs are the norm; conversely, in societies with fixed-mindset norms, growth-oriented students may experience lower well-being (Lou and Li, 2022[38]). These findings converge with calls for a “heterogeneity revolution” in education research, which views variability in outcomes not as error but as evidence of meaningful contextual moderation (Bryan, Tipton and Yeager, 2021[53]).
In summary, growth mindset is best understood not as a universal lever (Sisk et al., 2018[40]; Lou and Li, 2022[38]) but as part of a broader motivational ecosystem that includes learning strategies, self-beliefs, socio-emotional skills and supportive school culture (Burnette et al., 2013[6]; Dweck, 2006[4]; Dweck and Yeager, 2019[8]; Yeager et al., 2016[47]; Hecht, 2021[36]; Burnette et al., 2023[41]). The literature collectively suggests that fostering a growth mindset (the "seed") can complement structural reforms (Yeager and Walton, 2011[44]; Dweck and Yeager, 2019[8]) when embedded within supportive environments (the "soil") (Walton and Yeager, 2020[54]; Lou and Li, 2022[38]) that provide both the belief in improvement (Dweck, 2006[4]) and the necessary means to achieve it (Yeager and Walton, 2011[44]; Paunesku et al., 2015[37]). For policy, the implication is clear: cultivating adaptive beliefs among students, teachers and institutions requires parallel investment in the conditions that make those beliefs actionable – including feedback-rich classrooms (Dweck, 2006[4]; Kroeper, Fried and Murphy, 2022[58]), equitable opportunities (Claro, Paunesku and Dweck, 2016[15]; Dweck and Yeager, 2019[8]), and systems that normalise effort and growth (Dweck, 2006[4]; Yeager et al., 2019[14]; Walton and Yeager, 2020[54]; Lou and Li, 2022[38]).