Sisk et al. (2018) synthesized 273 studies of the mind-set–achievement association (N = 365,915) and 43 intervention effect sizes (N = 57,155) and found weak overall effects (pooled intervention effect d=0.08d=0.08), tentative signs that low-socioeconomic-status and academically at-risk students benefit, and an anomaly: pooled intervention effects were significant only when manipulation checks were absent or had failed, with 46% of checked studies failing to shift students' mind-sets. Whether these patterns reflect real boundary conditions and a genuine mechanism, or artifacts of weak designs, contestable measurement, and publication bias, remains unresolved. Later work sharpened but did not settle the question: a preregistered U.S. national experiment found benefits concentrated among lower-achieving students where peer norms supported the message, while two 2023 meta-analyses reached opposite conclusions — d=0.14d=0.14 for at-risk subgroups under high implementation fidelity versus d=0.05d=0.05 overall (nonsignificant after publication-bias correction, with no theoretically meaningful moderators) — and the accompanying exchange exposed unresolved disagreement over meta-analytic method and the validity of mind-set measurement itself. The open problem is to establish, with evidence meeting current causal best practices, whether growth mind-set interventions improve academic achievement, for whom and under what implementation and contextual circumstances, through what mechanism, and at what bias-corrected magnitude.

Background

Mind-sets (also called implicit theories) are lay beliefs about the malleability of personal attributes: a growth mind-set is the belief that intellectual abilities can be developed through effort and strategy, whereas a fixed mind-set is the belief that they are essentially static. Mind-set theory, developed largely by Carol Dweck and colleagues, predicts that growth mind-sets promote adaptive academic behavior and achievement, and short psychological interventions — typically brief readings and writing exercises teaching the malleability of intelligence — have been deployed in schools worldwide on that premise. The standard self-report instrument is the Implicit Theories of Intelligence Scale (ITIS), which conventionally treats fixed and growth mind-sets as opposite ends of one dimension by reverse-scoring one subscale. Throughout, effect sizes are reported as standardized mean differences (dd), where values around 0.10–0.20 are conventionally considered small. A manipulation check is a pre/post measure verifying that an intervention actually changed the psychological construct it targets; without one, observed outcome gains cannot be confidently attributed to mind-set change. Publication bias denotes the distortion of a literature's pooled effect by selective publication of positive results, typically assessed with funnel-asymmetry tests (which check whether small studies report systematically larger effects than large ones) and selection models (which estimate how much the pooled effect would change if unpublished results were included). Preregistration — committing to an analysis plan before data collection — is a key safeguard against researcher degrees of freedom. Two lines of evidence set up the problem. First, large observational datasets show that growth mind-set predicts achievement and tempers the achievement gap associated with poverty: Claro, Paunesku, and Dweck (2016) found in Chile's national assessment data that students from lower-income families held more fixed mind-sets and that a growth mind-set buffered the poverty–achievement relationship, and Claro and Loeb (2024), following more than 200,000 California students in grades 4–7, estimated that students with a growth mind-set learn roughly 0.066 standard deviations more per year in English language arts (about 18% of average annual growth) than observationally similar peers — but such designs cannot establish causality. Second, Sisk, Burgoyne, Sun, Butler, and Macnamara (2018) conducted two meta-analyses: one on the correlation between measured mind-set and achievement (273 effect sizes, N = 365,915) and one on interventions (43 effect sizes, N = 57,155). Both overall effects were weak (interventions: d=0.08d=0.08, 95% CI [0.02, 0.14]); the only statistically supported hints of moderation (i.e., of the effect differing across student subgroups) were that low-socioeconomic-status (SES) and academically at-risk students might benefit more, based on few effect sizes; publication-bias analyses were mixed and cautioned by heterogeneity. Most strikingly, the 28 of 43 intervention effect sizes that included manipulation checks yielded a nonsignificant pooled effect (d=0.04d=0.04, 95% CI [−0.03, 0.10]), the 15 without checks a significant one (d=0.18d=0.18, 95% CI [0.05, 0.31]), and 46% of checked studies failed to produce a significant pre-to-post increase in growth mind-set — raising the possibility that apparent intervention benefits arise where nobody verified the presumed mediator actually moved. Subsequent work deepened rather than resolved these worries. Yeager et al. (2019), in a preregistered, nationally representative U.S. randomized experiment, found a short online growth-mindset intervention improved grades among lower-achieving students and increased advanced-mathematics course-taking, with effects sustained in schools whose peer norms aligned with the intervention message — evidence for boundary conditions at scale, though not a direct test of low-SES moderation or of mind-set change as mediator. The two 2023 Psychological Bulletin syntheses then split the field. Burnette, Billingsley, Banks, Knouse, Hoyt, Pollack, and Simon (2023), synthesizing 53 randomized samples, reported that targeted (expected-to-benefit) subsamples under high implementation fidelity showed an academic-achievement effect of d=0.14d=0.14 (95% CI [0.06, 0.22]), with very wide 95% prediction intervals (−0.08 to 0.35; a prediction interval is the range in which a future study's effect is expected to fall), concluding that heterogeneity — who benefits, how, and why — is the field's central open question. Macnamara and Burgoyne (2023), synthesizing 63 studies (N = 97,672) with explicit best-practices criteria, found an overall effect of d=0.05d=0.05 (95% CI [0.02, 0.09]) that was nonsignificant after publication-bias correction; restricting to the 13 studies that demonstrably changed students' mind-sets gave d=0.04d=0.04 (95% CI [−0.01, 0.10], nonsignificant), and the 6 highest-quality studies gave d=0.02d=0.02 (nonsignificant), with no theoretically meaningful moderators and larger effects published by authors with a financial stake. A commentary exchange — Tipton, Bryan, Murray, McDaniel, Schneider, and Yeager (2023) reanalyzing Macnamara and Burgoyne's data with Burnette et al.'s approach to claim positive effects; Macnamara and Burgoyne's (2023) reply documenting altered effect sizes, moderator codings, and study inclusion; plus Yan and Schuetze (2023) arguing that current theory and measurement leave "growth mind-set" itself open to interpretation and Oyserman (2023) analyzing the field's cultural dynamics — established that the disagreement is substantially about method, coding, and construct validity. Independently, Scherer and Campos (2022), in an item-level meta-analysis of the ITIS (32 primary studies, N = 27,328), showed that fixed and growth mind-set items form two only moderately correlated factors (ρ=0.63\rho=0.63–0.65) even after reverse-scoring, undermining the mutual-exclusivity assumption built into most manipulation checks and mediation analyses. No study or synthesis since 2023 has adjudicated between these positions; no at-scale replication or bias adjudication of the dispute has appeared.

Problem Statement

Determine, with evidence meeting current causal best practices, to what extent and under which circumstances growth mind-sets matter for academic achievement. Concretely: (1) whether growth mind-set interventions causally improve students' academic achievement, or whether the apparent effects in the literature are attributable to weak study designs, failed or absent manipulation checks, and publication/researcher bias; (2) whether intervention effects are genuinely moderated by student socioeconomic status and academic risk — i.e., whether the at-risk subgroup benefits suggested by Sisk et al. (2018), Yeager et al. (2019), and Burnette et al. (2023), and the low-SES benefits suggested by Sisk et al. (2018) and Burnette et al. (2023), but not found by Macnamara and Burgoyne (2023), are reliable boundary conditions of mind-set theory rather than small-sample artifacts; (3) through what mechanism and under what implementation and contextual conditions (manipulation fidelity, classroom implementation, peer norms, combination with complementary supports) any real effect operates — in particular, whether verified change in mind-set, measured with instruments whose validity is itself at issue, mediates achievement gains; and (4) the bias-corrected magnitude of both the mind-set–achievement association and of intervention effects on achievement. The effects and associations at issue are to be established with quantified uncertainty. The criteria for admissible causal evidence must be stated explicitly and applied consistently. If item (1) yields no reliable causal effect, items (2) and (3) are resolved by stating that they are moot and why.

References

  1. Victoria F. Sisk, Alexander P. Burgoyne, Jingze Sun, Jennifer L. Butler, Brooke N. Macnamara (2018). To What Extent and Under Which Circumstances Are Growth Mind-Sets Important to Academic Achievement? Two Meta-Analyses. Psychological Science 29:549–571. DOI: 10.1177/0956797617739704. https://doi.org/10.1177/0956797617739704 (source paper)
  2. David S. Yeager, Paul Hanselman, Gregory M. Walton, Jared S. Murray, Robert Crosnoe, Chandra Muller, Elizabeth Tipton, Barbara Schneider, Chris S. Hulleman, Cintia P. Hinojosa, David Paunesku, Carissa Romero, Kate Flint, Alice Roberts, Jill Trott, Ronaldo Iachan, Jenny Buontempo, Sophia Man Yang, Carlos M. Carvalho, P. Richard Hahn, Maithreyi Gopalan, Pratik Mhatre, Ronald Ferguson, Angela L. Duckworth, Carol S. Dweck (2019). A national experiment reveals where a growth mindset improves achievement. Nature 573:364–369. DOI: 10.1038/s41586-019-1466-y. https://doi.org/10.1038/s41586-019-1466-y
  3. Susana Claro, David Paunesku, Carol S. Dweck (2016). Growth mindset tempers the effects of poverty on academic achievement. Proceedings of the National Academy of Sciences 113(31):8664–8668. DOI: 10.1073/pnas.1608207113. https://doi.org/10.1073/pnas.1608207113
  4. Jeni L. Burnette, Joseph Billingsley, George C. Banks, Laura E. Knouse, Crystal L. Hoyt, Jeffrey M. Pollack, Stefanie Simon (2023). A systematic review and meta-analysis of growth mindset interventions: For whom, how, and why might such interventions work? Psychological Bulletin 149(3-4):174–205. DOI: 10.1037/bul0000368. https://doi.org/10.1037/bul0000368
  5. Brooke N. Macnamara, Alexander P. Burgoyne (2023). 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. DOI: 10.1037/bul0000352. https://doi.org/10.1037/bul0000352
  6. Elizabeth Tipton, Christopher Bryan, Jared Murray, Mark A. McDaniel, Barbara Schneider, David S. Yeager (2023). 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. DOI: 10.1037/bul0000384. https://doi.org/10.1037/bul0000384
  7. Brooke N. Macnamara, Alexander P. Burgoyne (2023). A spotlight on bias in the growth mindset intervention literature: A reply to commentaries that contextualize the discussion (Oyserman, 2023; Yan & Schuetze, 2023) and illustrate the conclusion (Tipton et al., 2023). Psychological Bulletin 149(3-4):242–258. DOI: 10.1037/bul0000394. https://doi.org/10.1037/bul0000394
  8. Veronica X. Yan, Brendan A. Schuetze (2023). What is meant by "growth mindset"? Current theory, measurement practices, and empirical results leave much open to interpretation: Commentary on Macnamara and Burgoyne (2023) and Burnette et al. (2023). Psychological Bulletin 149(3-4):206–219. DOI: 10.1037/bul0000370. https://doi.org/10.1037/bul0000370
  9. Daphna Oyserman (2023). Culturally fluent theories, metascience, and scientific progress: A case example: Commentary on Macnamara and Burgoyne (2023) and Burnette et al. (2023). Psychological Bulletin 149(3-4):220–228. DOI: 10.1037/bul0000376. https://doi.org/10.1037/bul0000376
  10. Susana Claro, Susanna Loeb (2024). Students With Growth Mindset Learn More in School: Evidence From California's CORE School Districts. Educational Researcher 53:389–402. DOI: 10.3102/0013189X241242393. https://doi.org/10.3102/0013189X241242393
  11. Ronny Scherer, Diego G. Campos (2022). Measuring those who have their minds set: An item-level meta-analysis of the implicit theories of intelligence scale in education. Educational Research Review 37:100479. DOI: 10.1016/j.edurev.2022.100479. https://doi.org/10.1016/j.edurev.2022.100479