Angela Duckworth’s grit research became one of the most successful behavioral-science exports of the 2010s — TED talk, MacArthur grant, bestselling book, school curricula, military screening. The 2017 Credé meta-analysis of 88 samples (n > 66,000) found grit correlates r ≈ .84 with conscientiousness — essentially the same construct, repackaged. Here is the honest empirical story and what it means for leaders evaluating personality-based frameworks.

If you have sat through a corporate leadership training program in the last decade, attended a teacher in-service day at almost any American public school, watched any of the most-viewed TED talks of all time, or read any of the bestselling business books of the 2010s, you have probably encountered grit. The concept, as popularized by University of Pennsylvania psychologist Angela Duckworth, is that “grit” — defined as passion and perseverance for long-term goals — is a personality trait that predicts achievement better than talent, IQ, or socioeconomic status, and that this trait is at least partially malleable through deliberate development.

The cultural footprint is enormous. Duckworth’s 2013 TED talk has been viewed more than thirty million times. Her 2016 book Grit: The Power of Passion and Perseverance was a New York Times bestseller and has been translated into more than thirty languages. She received a MacArthur Foundation “genius” grant in 2013, partly on the strength of the grit research program. Schools across the United States — and increasingly internationally — adopted grit assessments and grit interventions. The U.S. Department of Education, under the Obama administration, considered making grit-like “non-cognitive” outcomes a measured part of school accountability. The U.S. military experimented with grit screening for selection into elite training programs. Corporate L&D vendors built entire training programs around developing employee grit. By the late 2010s, “grit” had become one of the most successful psychological constructs ever exported into popular culture, public policy, and corporate practice.

The empirical picture, however, looks substantially thinner than the popular framing suggests. The 2017 meta-analysis by Marcus Credé, Michael Tynan, and Peter Harms — synthesizing 88 independent samples covering more than 66,000 participants — found that grit correlates so strongly with the well-established personality trait of conscientiousness (estimated true-score correlation around r ≈ .84) that for most practical purposes the two constructs are indistinguishable. Grit’s correlation with performance outcomes (academic, occupational) is modest — roughly r ≈ .15 to .20 — and most of the predictive work is done by the “perseverance” facet, not the “passion for long-term goals” facet that supposedly distinguishes grit from existing personality measures. A 2018 follow-up review by Credé concluded that the case for grit interventions in schools is weak, that the higher-order grit construct lacks empirical support, and that the broader conscientiousness literature — which has been studied for fifty years — has stronger foundations for the same insights.

This article is not a takedown. Duckworth’s research is not fraudulent. The construct is not “fake.” There is a real, statistically reliable trait being measured, and that trait does correlate with achievement outcomes. The honest empirical story is that the trait Duckworth named “grit” is almost entirely conscientiousness, that its predictive power is modest, and that the policy and corporate-training enthusiasm around grit substantially outran what the meta-analytic evidence supports. For strategists, executives, and leaders evaluating L&D programs, hiring assessments, and school interventions that invoke “grit,” the practical implication is that you are paying for repackaged conscientiousness — and that the bigger story about the gap between popular framing and meta-analytic evidence is worth understanding in its own right.

What Duckworth’s 2007 Paper Actually Found

The grit research program began with Duckworth, Peterson, Matthews & Kelly (2007), “Grit: Perseverance and passion for long-term goals,” published in the Journal of Personality and Social Psychology, volume 92, pages 1087–1101. This is the foundational empirical paper. It introduced the Grit Scale (the 12-item version, later called Grit-O, and subsequently reduced to the 8-item Grit-S in Duckworth & Quinn 2009), and reported six studies establishing the construct’s predictive validity.

The headline studies are well-known to anyone who has heard a grit talk. Study 3 examined 1,218 cadets entering the U.S. Military Academy at West Point and showed that grit scores predicted completion of “Beast Barracks” — the brutal seven-week summer training program — better than the academy’s composite admissions index that combined SAT scores, high school class rank, leadership ratings, and physical aptitude scores. Study 4 examined 175 finalists in the 2005 Scripps National Spelling Bee and found that grit predicted advancement in the competition.

Across the six studies, grit accounted for an average of about 4% of the variance in success outcomes (educational attainment, GPA, West Point retention, Spelling Bee performance). The Duckworth team interpreted this as evidence that grit is a meaningful predictor of long-term achievement, distinct from existing measures of intelligence, talent, and conscientiousness.

The 2007 paper is methodologically reasonable for its era. It is not a fraud case, not a single-study finding, and not a piece of obvious researcher-degrees-of-freedom abuse. The trait Duckworth and her co-authors measured is a real trait, the predictive validity they reported is real, and the West Point and Spelling Bee samples are genuine. What the paper does not do — and what subsequent meta-analytic work would interrogate — is establish that “grit” is a meaningfully distinct construct from conscientiousness, or that the magnitudes of effect justify the policy and corporate enthusiasm that would follow.

What “Grit” Was Theoretically Supposed To Add

The reason grit became a phenomenon is not that it was the first personality trait shown to predict achievement. Conscientiousness — one of the “Big Five” personality dimensions, established in personality psychology since at least the 1980s — had already been shown to predict academic and occupational achievement across thousands of studies. By the early 2000s, the conscientiousness-predicts-achievement finding was as well-established as anything in personality psychology.

What grit was supposed to add — the theoretical claim that distinguished it from existing constructs — was twofold. First, grit was supposed to capture the long-term pursuit of a singular passion. Conscientiousness is about being organized, responsible, and dependable in the short and medium term. Grit was supposed to be about sticking with one overarching goal for years — what Duckworth often called the “consistency of interest” facet — even when easier or more interesting alternatives presented themselves. The popular framing was the violinist who practices for ten years to reach mastery, the entrepreneur who works through repeated failures toward a single vision, the student who pursues a graduate degree across years of unglamorous work.

Second, grit was supposed to be at least partially developable. The popular framing — particularly in Duckworth’s TED talk and book — emphasized that grit is not a fixed trait you either have or don’t have; it can be cultivated through specific practices, mindsets, and environments. This was the implicit policy hook: if grit predicts achievement and grit can be developed, then schools and workplaces should systematically develop grit, particularly in disadvantaged populations who might otherwise lack the gritty role models or environments that build the trait.

These two claims — grit is distinct from conscientiousness, and grit is developable — are what justified treating grit as a meaningfully new finding rather than a rebranding of an existing personality dimension. They are also the two claims that have held up least well under meta-analytic scrutiny.

The Credé, Tynan & Harms 2017 Meta-Analysis

In 2017, Marcus Credé (Iowa State University), Michael Tynan (Iowa State University), and Peter Harms (University of Alabama) published “Much ado about grit: A meta-analytic synthesis of the grit literature,” in the Journal of Personality and Social Psychology, volume 113, pages 492–511. DOI: 10.1037/pspp0000102.

This was the first comprehensive quantitative synthesis of the grit literature. The authors aggregated effect sizes across 88 independent samples covering more than 66,000 participants. They examined three questions: how strongly does grit correlate with conscientiousness; how strongly does grit predict academic and occupational performance; and does the higher-order grit construct (combining perseverance and consistency-of-interest into a single index) have better predictive validity than its component facets considered separately.

The headline finding — the one that has shaped subsequent academic debate about grit — was the correlation between grit and conscientiousness. Pooling across studies that measured both constructs, Credé and colleagues reported an estimated true-score correlation of approximately r ≈ .84. In personality-psychology terms, this is enormous. Two constructs correlated at r ≈ .84 are not meaningfully distinct — they are essentially the same construct measured with slightly different items. By comparison, two well-established but related personality traits like conscientiousness and self-control correlate around r ≈ .50 to .60; two ostensibly different traits like extraversion and conscientiousness correlate near zero.

Credé and colleagues framed this as a textbook case of the “jangle fallacy” — the error of treating two labels for the same underlying construct as if they referred to different things. (The complementary “jingle fallacy” is using the same label for two different constructs.) On this reading, “grit” is not a new personality dimension; it is a repackaging of conscientiousness — a well-established trait that has been studied for decades — with a more marketable name and a more compelling narrative.

The Credé conclusion was reinforced by a 2020 paper from a different research group — Ponnock, Muenks, Morell, Yang, Gladstone & Wigfield (2020), “Grit and conscientiousness: Another jangle fallacy,” published in the Journal of Research in Personality, volume 89, article 104021, DOI 10.1016/j.jrp.2020.104021 — that explicitly tested whether the grit-conscientiousness overlap reflects shared variance with facets of conscientiousness like industriousness and self-discipline. The Ponnock study reached essentially the same conclusion: grit, particularly its perseverance facet, is largely indistinguishable from the industriousness facet of conscientiousness as already measured in the broader personality literature.

What Credé Found On Effect Size

The second piece of the Credé 2017 meta-analysis concerned predictive validity. How strongly does grit predict the outcomes it is supposed to predict — academic achievement, occupational performance, persistence?

The aggregated results showed correlations with performance outcomes in the range of approximately r ≈ .15 to .20 for academic outcomes (GPA, academic persistence) and somewhat smaller for occupational outcomes. These are not zero — grit does predict performance — but they are modest. A correlation of r ≈ .15 to .20 accounts for between roughly 2% and 4% of variance in outcomes. By comparison, cognitive ability (general intelligence) typically correlates with academic and occupational performance at r ≈ .40 to .50, accounting for 16% to 25% of variance.

The third major finding concerned the structure of grit itself. Duckworth’s framework treats grit as a higher-order trait composed of two facets: “perseverance of effort” (how hard you work, how persistent you are) and “consistency of interest” (how stable your goals are over time, how much you stick with one passion). The theoretical case for treating grit as a unified higher-order construct depended on these two facets contributing roughly equal predictive power and showing meaningful covariance with each other.

The Credé meta-analysis found that the two facets are only modestly correlated with each other (around r ≈ .50, not strong enough to justify combining them into a single trait), and — more damaging — that the “perseverance of effort” facet does virtually all the predictive work. The “consistency of interest” facet, which is supposedly the distinctive contribution of grit (the part about long-term passion that supposedly differentiates grit from short-term conscientious behavior), adds little incremental validity beyond perseverance.

The practical interpretation: when grit predicts achievement, it is essentially because the perseverance items in the scale are measuring conscientious effort — the same thing the industriousness facet of conscientiousness measures. The “passion for long-term goals” content that supposedly makes grit special is not contributing much to the prediction.

What’s Honest To Say About Grit Now

After Credé 2017 and the subsequent literature, here is what a fair-minded reading of the evidence supports.

Grit is a real, reliably measured trait. The Grit Scale produces consistent scores across administrations, the scale has reasonable internal-consistency reliability, and the scores do correlate with achievement outcomes at small-to-modest magnitudes. The trait is not fake, the measurement is not pseudo-science, and Duckworth’s research program is not fraudulent.

Grit is essentially conscientiousness, particularly its industriousness facet. The r ≈ .84 correlation between grit and conscientiousness in the Credé meta-analysis, replicated in subsequent work by Ponnock and others, supports treating grit as a relabeling of an existing personality dimension rather than a new construct. This is not a moral failing on Duckworth’s part — the construct was developed in good faith and with reasonable theoretical motivation — but the meta-analytic evidence does not support treating grit as something separate from conscientiousness.

Grit predicts achievement at modest magnitudes, smaller than the popular framing suggests. The correlations of approximately r ≈ .15 to .20 with academic outcomes are real but not large, and they are substantially smaller than the correlations between cognitive ability and the same outcomes. The popular framing — that grit predicts achievement better than talent or IQ — is not supported by the meta-analytic evidence.

The case for grit interventions is weak. The 2018 follow-up review by Credé — Credé, M. (2018), “What shall we do about grit? A critical review of what we know and what we don’t know,” in Educational Researcher, volume 47, number 9, pages 606–611, DOI 10.3102/0013189X18801322 — concluded that the evidence base for grit interventions in schools is sparse and that what evidence exists does not support strong claims about the trainability of grit beyond what would be achieved by general motivational or study-skills interventions. The popular claim that “you can teach grit” outruns the empirical evidence.

These four points, considered together, do not justify abandoning the construct entirely. They do justify substantial deflation of the policy, corporate-training, and educational-intervention enthusiasm that grew up around grit between 2007 and the late 2010s.

What This Means For School And Workplace Interventions

The most consequential implication of the Credé meta-analysis is not academic — it is practical. If grit is essentially conscientiousness, then the case for designing and deploying “grit interventions” — as distinct from interventions that target conscientiousness, self-regulation, study skills, or motivation — is weak.

This matters because the personality psychology literature on conscientiousness has been developing for fifty years. There is a substantial evidence base on how conscientiousness relates to academic and occupational outcomes (well-established), on the heritability of conscientiousness (substantial — roughly 40% to 50% across twin studies), on the stability of conscientiousness across the lifespan (high in adulthood, with modest mean-level increases from late adolescence through middle age), and on the difficulty of changing trait-level conscientiousness through brief interventions (substantial — most interventions produce small short-term changes that do not persist).

If you are an educator or L&D professional evaluating a “grit training” program, the honest question is: what does this program offer beyond what the conscientiousness literature already tells us? In most cases, the answer will be: not much, except a more marketable name. The same is true of “grit assessments” used in hiring or admissions: a well-validated conscientiousness measure will do roughly the same predictive work as a grit measure, often with better psychometric properties from a longer research base.

The honest practical recommendation is not “stop caring about persistence.” Persistence matters. The recommendation is to ground your interventions and assessments in the broader, better-established conscientiousness literature, rather than treating grit as a separate trait that justifies separate programs. The corporate training market for “grit development” is in significant part a market for repackaged conscientiousness, and the underlying evidence base for changing conscientiousness through training is modest at best.

Duckworth’s Response And The Honest Debate

To her substantial credit, Duckworth has engaged with the meta-analytic critique rather than dismissing it. In a 2016 NPR Ed interview responding to Credé’s then-forthcoming meta-analysis, she acknowledged that grit is most accurately understood as a member of the conscientiousness family, while maintaining that the construct has independent predictive value — particularly in the long-term goal-pursuit framing that conscientiousness measures do not directly capture.

Subsequent papers from Duckworth and collaborators have shifted somewhat in framing. The strongest claims of the 2007 paper and the popular book — that grit predicts achievement better than talent or IQ, that consistency of interest is a distinct and important facet — have been moderated in more recent academic work. Duckworth has also acknowledged that the perseverance facet carries most of the predictive weight, while maintaining that the passion component is conceptually important even if its empirical contribution is smaller.

The honest scholarly debate, as of the mid-2020s, is roughly this. Credé and his collaborators argue that grit is essentially conscientiousness, that the higher-order grit construct lacks empirical support, and that the popular framing of grit substantially overstates the empirical case. Duckworth and her collaborators acknowledge the close relationship to conscientiousness but argue that the long-term goal-pursuit framing captures something practically important even if it is not statistically distinct from conscientiousness, and that the grit framework has value as a communicable, motivating construct for educational and personal-development contexts even if it is not a new personality dimension.

Both positions are defensible. The popular framing of grit — the version in TED talks, corporate training programs, and breathless media coverage — is the version that does not survive the meta-analytic evidence. The more cautious academic version, which acknowledges the conscientiousness overlap and the modest effect sizes, is consistent with the data.

What This Means For Strategists And Leaders Evaluating Personality-Based Frameworks

The pattern that grit exemplifies — a popular behavioral-science framework with a charismatic public-facing researcher, a compelling narrative, a brief and intuitive measurement instrument, broad adoption in schools and corporations, and an underlying empirical foundation that is more modest and more contested than the public framing suggests — is not unique to grit. It is one of the most common patterns in the replication crisis.

Several features characterize the pattern, and pattern-recognition matters for leaders who are repeatedly asked to evaluate, fund, and operationalize behavioral-science claims.

The first feature is the existence of a well-established prior construct that the new construct heavily overlaps with. In grit’s case, that prior construct is conscientiousness. In growth mindset’s case, it is general motivational and study-skills research. In emotional intelligence’s case, it is the broader personality-traits literature (with substantial overlap with conscientiousness, neuroticism, and agreeableness). When a new construct is being championed as transformative, ask: what is the well-established trait or finding that this new construct largely overlaps with? If the overlap is large, the new construct may be a relabeling rather than a discovery.

The second feature is the gap between popular framing and academic framing. Look at what the researcher says in TED talks and trade-press interviews versus what the same researcher writes in peer-reviewed journal articles. The popular framing is almost always more confident, more sweeping, and more strategically actionable than the academic framing. Both can be true at once — the academic version is more careful — but the difference matters when you are operationalizing the finding into policy or corporate practice. The popular framing of grit (it predicts achievement better than IQ; it can be developed; it is the secret to success) was substantially more confident than the academic framing of grit (it predicts achievement modestly; it correlates highly with conscientiousness; the case for development is weak).

The third feature is the brief, attractively-named measurement instrument that becomes the entry point into the framework. The Grit Scale is twelve items (or eight, in the short form). It takes about two minutes to complete. It produces a single score that feels meaningful. This kind of instrument is enormously valuable for academic research and for popular adoption, but the brevity is also a warning sign: a complex psychological construct that can be reliably measured in eight items is either remarkable or — more often — a slight rebranding of a construct that has been measured many ways before.

The fourth feature is rapid adoption by institutions that have decision-pressure but lack the technical capacity to evaluate the underlying evidence. Schools, government agencies, and corporate L&D departments are all under continuous pressure to identify and deploy evidence-based interventions. They are not, in general, set up to read meta-analyses and personality-psychology methodological papers. When a charismatic researcher publishes a popular book and gives a viral talk, the gap between the popular framing and the academic framing gets exploited at scale: institutions adopt the framework, the institutional adoption becomes social proof of the framework’s validity, and the resulting feedback loop is difficult to break with later corrective evidence.

The fifth feature is the difficulty of unwinding institutional adoption after the corrective evidence arrives. The Credé meta-analysis was published in 2017. By that point, grit had been incorporated into curricula at hundreds of school districts, into corporate L&D programs at major employers, and into selection processes at military and educational institutions. Removing grit from these systems is harder than installing it was, because the people who installed it had reputational and budgetary stakes in its validity. The result is that the popular framing of grit continues to circulate widely in educational and corporate settings well after the meta-analytic literature has substantially deflated the empirical case.

For leaders evaluating behavioral-science frameworks: when you encounter a new framework being championed for institutional adoption, ask about the meta-analytic evidence rather than the original study. Ask about the relationship to better-established prior constructs. Ask what the framework predicts, and at what magnitude, and how that magnitude compares to other interventions you could deploy with the same resources. Be especially skeptical when the popular framing is more confident than the academic framing, when the measurement instrument is short and intuitive, and when adoption by other institutions is being offered as evidence of validity.

Grit is the cleanest recent example of this pattern. The construct is real. The research is honest. The popular framing — and the institutional adoption that followed — outran the empirical evidence by a substantial margin. The corrective meta-analytic literature is now well-established but has not fully propagated into the institutions that adopted the framework during its peak popularity.

Sources

  • Duckworth, A. L., Peterson, C., Matthews, M. D., & Kelly, D. R. (2007). Grit: Perseverance and passion for long-term goals. Journal of Personality and Social Psychology, 92(6), 1087–1101. DOI: 10.1037/0022-3514.92.6.1087
  • Duckworth, A. L., & Quinn, P. D. (2009). Development and validation of the Short Grit Scale (Grit-S). Journal of Personality Assessment, 91(2), 166–174. DOI: 10.1080/00223890802634290
  • Duckworth, A. L. (2016). Grit: The Power of Passion and Perseverance. New York: Scribner.
  • Credé, M., Tynan, M. C., & Harms, P. D. (2017). Much ado about grit: A meta-analytic synthesis of the grit literature. Journal of Personality and Social Psychology, 113(3), 492–511. DOI: 10.1037/pspp0000102
  • Credé, M. (2018). What shall we do about grit? A critical review of what we know and what we don’t know. Educational Researcher, 47(9), 606–611. DOI: 10.3102/0013189X18801322
  • Ponnock, A., Muenks, K., Morell, M., Yang, J., Gladstone, J., & Wigfield, A. (2020). Grit and conscientiousness: Another jangle fallacy. Journal of Research in Personality, 89, 104021. DOI: 10.1016/j.jrp.2020.104021
  • Tynan, M. C., Credé, M., & Harms, P. D. (2020). Are individual differences in cognitive ability and personality traits incrementally predictive of standardized achievement test scores? Educational and Psychological Measurement.
  • Kamenetz, A. (2016, May 25). MacArthur ‘Genius’ Angela Duckworth responds to a new critique of grit. NPR Ed. URL: https://www.npr.org/sections/ed/2016/05/25/479172868/angela-duckworth-responds-to-a-new-critique-of-grit

FAQ

Should I stop using grit assessments in hiring?

Not necessarily, but understand what you are buying. A grit score is roughly a conscientiousness score, particularly its industriousness facet. If your hiring process already includes a well-validated conscientiousness measure (or a structured interview assessing dependability and persistence), adding a separate grit assessment is largely redundant. If your hiring process does not include any conscientiousness measure, a grit scale will provide modest incremental predictive validity for performance outcomes — at roughly the same magnitude that a longer-established conscientiousness scale would provide. The choice between “grit assessment” and “conscientiousness assessment” is mostly a branding question; the underlying predictive validity is similar.

What about conscientiousness? Is that a more reliable thing to measure?

Yes, with caveats. Conscientiousness is one of the most-studied personality dimensions, with a fifty-year evidence base, well-validated measurement instruments (NEO-PI-R, IPIP scales, Big Five Inventory), and substantial meta-analytic evidence linking it to academic and occupational performance. The typical correlation between conscientiousness and job performance is roughly r ≈ .20 to .25 across occupations — modest but real, and well-documented. The caveats: conscientiousness is heritable and relatively stable in adulthood, so the case for “developing conscientiousness through brief interventions” is also weak, just as it is for grit. The honest takeaway is that conscientiousness predicts performance modestly, that grit is essentially conscientiousness, and that both are useful for selection but limited as targets for short-term intervention.

What about the military’s grit screening programs?

The U.S. military and several other organizations experimented with grit-based screening for selection into elite training programs in the 2010s, following the West Point findings in Duckworth’s 2007 paper. The honest interpretation is that those screening programs were likely capturing the predictive validity of conscientiousness — the same predictive validity that a well-validated conscientiousness measure would have captured. If you are operating in a setting that uses grit screening, the predictive value is real but modest, and you would get similar value from any well-constructed measure of trait-level dependability and persistence. The marketing of grit as uniquely military-relevant or uniquely predictive of elite-program completion is not strongly supported by the meta-analytic literature.

Is Angela Duckworth wrong?

No, not in the sense of being fraudulent or producing bad research. Duckworth’s empirical work is methodologically reasonable and the trait she measured is real. What is more accurate to say is that the popular framing of grit — the version in the TED talk, the bestselling book, and the corporate training programs — substantially outran the empirical evidence base, and that the meta-analytic literature has subsequently shown the construct to be essentially a relabeling of conscientiousness with modest predictive validity. Duckworth herself has engaged with this critique and moderated some of her claims in subsequent academic work. The honest summary is that her research is real, her popular framing was overstated, and the gap between the two is the actual story.

Can grit be taught? What about school-based interventions?

The 2018 Credé review concluded that the evidence base for grit interventions in schools is sparse and that what evidence exists does not strongly support the claim that grit interventions improve achievement beyond what general motivational, study-skills, or behavioral-engagement interventions would produce. The broader literature on changing trait-level personality through brief interventions is similarly modest — conscientiousness is relatively stable in adulthood, and most interventions produce small short-term changes that do not persist. The honest practical recommendation for schools is to focus on well-established educational interventions (high-quality teaching, more time on task, individualized tutoring, evidence-based curricula) rather than on grit-specific training, which does not appear to add measurable value beyond what those broader interventions produce.

What is the “jangle fallacy” Credé invokes?

The jangle fallacy is the error of treating two different labels as if they referred to different underlying constructs, when in fact they refer to the same construct. (The complementary jingle fallacy is using one label for two genuinely different constructs.) Credé’s argument is that “grit” and “conscientiousness” are largely the same construct (the r ≈ .84 correlation in the meta-analysis is the evidence), and that treating them as separate constructs is a jangle fallacy. The 2020 Ponnock paper explicitly framed the grit-conscientiousness overlap in jangle-fallacy terms. The methodological point is broader: any time a new personality construct is introduced that correlates very highly with an existing well-established trait, the burden of proof is on the new construct to demonstrate incremental predictive validity beyond the existing one — and the grit literature has not strongly met that burden.

How does this compare to growth mindset, which also had a similar arc?

The two stories are structurally similar but empirically distinct. Growth mindset (Carol Dweck) and grit (Angela Duckworth) both became enormously popular behavioral-science exports in the 2010s, both were championed through TED talks and bestselling books, both were adopted at scale by schools and corporations, and both have been subjected to meta-analytic critiques that substantially deflated the empirical case. The key difference is the nature of the critique. Growth mindset’s meta-analytic critique (Sisk 2018) found small effects but real ones — the construct exists, the interventions produce modest effects in specific subgroups. Grit’s meta-analytic critique (Credé 2017) found that the construct is essentially a relabeling of conscientiousness — not that grit doesn’t exist, but that it isn’t distinct from what was already being measured. Both are cases where the popular framing outran the empirical evidence, but the specific empirical story is different. The growth-mindset story is “real but a tenth the size you were told.” The grit story is “real, but barely distinguishable from a trait that was already well-studied for fifty years.”

Should I read Duckworth’s book?

It’s a well-written popular-science book with engaging narratives and useful framing for thinking about persistence and long-term goal pursuit. It is also more confident in its empirical claims than the academic literature supports. If you read it, read it knowing that the meta-analytic evidence base (Credé 2017, Credé 2018, Ponnock 2020) substantially complicates the central claims about grit being a distinct trait and a strong predictor of achievement. A more accurate practical takeaway from the book is something like: persistence and effort matter (which is true and well-supported), and conscientiousness — the trait that “grit” mostly measures — is a modest but real predictor of long-term achievement. The strongest claims in the popular framing should be discounted accordingly.

replication-crisis grit-duckworth personality-psychology education-policy evidence-evaluation
Share this article
LinkedIn (opens in new tab) X / Twitter (opens in new tab)
Atticus Li

Experimentation and growth leader. CXL-certified CRO practitioner, Mindworx-certified behavioral economist (1 of ~1,000 worldwide). 200+ A/B tests across energy, SaaS, fintech, e-commerce, and marketplace verticals.

About LinkedIn Newsletter