Grit Is Mostly Conscientiousness
Two questionnaires, one labelled grit and one labelled conscientiousness, agree more closely than two different conscientiousness questionnaires do. The half of the scale that asks about keeping at things still earns its keep, the half that asks about keeping at the same thing does not, and the most deflationary sentence in the record belongs to the person who named the trait.
Answer to the question
Is grit just another word for conscientiousness?
Largely, when the question is performance. Pooling 73 studies and 66,807 people, Credé, Tynan and Harms found that total grit scores add nothing to school grades once conscientiousness is accounted for, while conscientiousness still adds in the reverse order. The perseverance half is the exception, and grit still predicts who quits.
Key takeaways
- Credé, Tynan and Harms pooled 584 effect sizes from 88 samples in 73 studies covering 66,807 people.
- Grit and conscientiousness questionnaires agree more closely than two different conscientiousness questionnaires agree with each other.
- Account for conscientiousness first and overall grit adds nothing to school grades; reverse the order and conscientiousness still adds.
- In 11,258 West Point cadets, grit was the only reliable predictor of completing initiation and measured ability was not.
Five hundred and eighty-four separate measurements, taken from seventy-three studies, eighty-eight groups of people and sixty-six thousand eight hundred and seven lives. That is what Marcus Credé, Michael Tynan and Peter Harms pooled in 2017 to ask what grit actually predicts, and the answer came back in two pieces that are still routinely run together. Against school performance the relationship is real and weak: rank a class by score and you would guess the order of their grades a little better than chance, though not much better. The authors note that other well-known predictors of academic achievement report correlations more than twice as large. The second piece supplies this article's title, and it is stranger than a null result.
The Sentence Everybody Quotes
The line that carried this idea out of psychology and into school assemblies and hiring decks is about soldiers. In 2009 Angela Duckworth and Patrick Quinn reported that cadets who scored a standard deviation above the average on their eight-item scale were ninety-nine per cent more likely to complete West Point's summer initiation.
Ninety-nine per cent more likely sounds enormous and describes something small. About ninety-five cadets in every hundred complete that training anyway. A doubling of the odds moves you from roughly ninety-five in a hundred to roughly ninety-seven and a half, and that is the whole of it. Credé and his colleagues made exactly that point in their review. Speaking to NPR Ed in May 2016, Duckworth accepted that the sentence should have been phrased differently. She called the trait's independent contribution small to medium, and said she would prefer to think of it as a member of the conscientiousness family.
The Overlap With Conscientiousness
The 2017 review put twenty-two samples and more than eighteen thousand people behind the comparison that matters most here, and produced the oddest result in this literature.
> Two questionnaires, one labelled grit and one labelled conscientiousness, agree with each other more closely than two different conscientiousness questionnaires agree with each other.
Two questionnaires, one labelled grit and one labelled conscientiousness, agree with each other more closely than two different conscientiousness questionnaires agree with each other.
That is not a figure of speech, and the comparison figure is worth attributing properly. The typical agreement between two global conscientiousness inventories is not the review's own estimate: Credé and his colleagues take it from Valerie Pace and Michael Brannick's 2010 meta-analysis of personality scales sharing a label. Their own pooled figure for grit against conscientiousness sits above it. They write that the correlation is so strong as to limit the incremental value of the scores for predicting performance over conscientiousness, and to suggest that the construct may be redundant with it.
A twin study reads the same way from a different angle. Kathryn Rimfeld and colleagues reported in 2016, across more than two thousand British twin pairs, a genetic overlap between the two of much the same size, a figure that reaches us through our own earlier verification record rather than our own read of the paper. Say that plainly and the overlap is still hard to dismiss as an artefact of how people fill in forms.
The originator has since written the same thing in a peer-reviewed journal. In a 2020 paper first-authored by Daniel Southwick, with Duckworth as its third author, the sentence reads: "We conceive grit as a facet of conscientiousness."
The Two Halves of the Scale
The scale has always had two parts, and they were never equal partners. One half asks about keeping at things: finishing what you begin, working through setbacks, staying diligent when a project stops being interesting. The other half asks about keeping at the same thing: whether your interests shift from year to year, whether a new enthusiasm pulls you off an old one. The first is called perseverance of effort. The second is called consistency of interest.
Credé's review pulled them apart and they behaved like different variables. Against overall academic performance, the keeping-at-it half is roughly two and a half times as strongly related as the keeping-at-the-same-thing half, and against secondary school grades the gap is wider still. The authors compared each pair formally and report that all four differed reliably.
Their own summary is blunt about what that overturns. Contrary to early assertions that both facets predicted success outcomes equally well, the perseverance facet showed much stronger relations with every academic criterion than the consistency facet did.
A later and much larger meta-analysis repeats the pattern rather than magnifying it. Lam and Zhou in 2022 pooled one hundred and thirty-seven studies against Credé's seventy-three, and more than two hundred and eighty-five thousand people against not quite sixty-seven thousand, so the pooled sample is better than four times the size and the study count a little under twice.
The gap between the halves came back the same shape and the same width: the keeping-at-it half related to achievement about two and a half times as strongly as the keeping-at-the-same-thing half, which is roughly where Credé's own pair of figures sat. Those figures reach us through our own earlier verification record and not from our read of the paper. Whatever else is unsettled here, the split between the two halves is not a one-off.
The Order of the Questions
Two questions, asked of the same person in either order, settle the matter the title rests on.
| Accounted for first | Added second | What the second one then adds |
|---|---|---|
| Conscientiousness | The overall score | Nothing at all to secondary school grades, and nothing to academic performance overall |
| Conscientiousness | The overall score | A trace to college grades, about four parts in a thousand of the differences between students |
| The overall score | Conscientiousness | Something, on every academic outcome |
| Conscientiousness | The keeping-at-it half on its own | Something substantial, on every academic outcome |
Ask somebody how conscientious they are and then ask how gritty they are, and the second answer tells you almost nothing more about their school grades than the first already did. Ask the two questions the other way round and the conscientiousness answer still adds. That asymmetry, from the review's own hierarchical regressions, is what licenses this article's title, and it is a claim about total scores.
Here is the precision the case turns on. Enter the keeping-at-it half on its own, after conscientiousness, and it does add, substantially, on every academic outcome the review tested. It survives controls for conscientiousness and the other half together. Credé and his co-authors call that result encouraging in their own text. The keeping-at-the-same-thing half adds essentially nothing, and its coefficients turn negative in a way that suggests it may be suppressing the half that works.
So the famous arithmetic adds a useful number to a useless one and reports the average.
So the famous arithmetic adds a useful number to a useless one and reports the average.
Those same authors then attach a caveat to their own encouraging finding, and it has to be dealt with. Such incremental validity, they write, would be observed even if the trait were simply a different manifestation of conscientiousness, as their other results suggest. The reason is the same Pace and Brannick figure: two inventories aimed at one trait agree only moderately, so a second questionnaire can always pick up a little that the first one missed without measuring anything new.
Take that seriously and the half-that-works finding stops being a discovery about a second trait. What survives is narrower and still worth having. If you must score people on something, the perseverance items beat the total score and beat a nine-item conscientiousness scale, and Credé's own text argues that small gains matter in high-stakes settings, where lifting a college retention rate by a single percentage point would help thousands of students. That is a measurement result about one questionnaire, and it is the whole of what survives.
And the same review says where the real gains are. Study skills and habits relate to college grades more than twice as strongly as the total score does, adjustment to college almost as strongly, institutional attachment is among the better predictors of staying enrolled, and class attendance is strongly related to grades, with compulsory attendance cutting failure rates sharply. Credé and his colleagues put all four forward as better targets on the ground that each one responds to intervention, which is their recommendation and not a trial result: what their own tables hold is correlations. That is the practical finding in the paper everybody cites for the demolition.
The Items and Their Wording
There is a deeper problem with treating the two halves as halves of a single thing, and it arrives from people defending the construct.
Every item on the perseverance subscale is worded positively and every item on the consistency subscale is worded negatively. Jon Jachimowicz and colleagues argued in 2019 that this design alone can manufacture the appearance of two factors, and then ran the test. They wrote an all-positive version of the eight-item scale and an all-negative version, gave each to its own sample of several hundred people, and watched the two supposed halves collapse into something very close to one measure in both. Both versions failed the standard test for telling two factors apart. Their conclusion is that the short scale measures perseverance and that none of its items capture passion.
A separate team reached a related conclusion from another direction. Annette Ponnock and five co-authors studied 1,246 ninth and tenth graders across seven north-eastern American schools, with maths and science grades taken from school records, and modelled the two sets of items together at the level of the individual item.
The model in which the two halves combine into one overall factor would not converge. The models that fitted best were those fusing conscientiousness with one half. Their recommendation is the sharpest sentence in the critical literature: that researchers stop using the short scale, and either validate a measure that justifies treating this as a distinct construct or go back to studying conscientiousness. They name their own limitation too, that the reverse wording could be inflating what they found.
Three positions are now on the table and they cannot all be right. What they agree on is that nobody defends the version in the book, in which passion and perseverance are equal partners in one trait. The originators conceded the higher-order structure themselves in a 2021 commentary, which also reaches us through that earlier record.
Eleven Thousand Cadets at West Point
Anyone who dismisses grit outright has not read what follows, and the strongest evidence for it is the author's own.
In 2019 Duckworth, as first author, published an analysis of a full decade of West Point data: eleven thousand two hundred and fifty-eight cadets across nine entering cohorts, preregistered, with the data posted openly. The authors note that the sample is fifty times larger than the average study in a high-impact psychology journal, and they tightened their own significance threshold to account for five outcomes.
On completing the summer initiation, cadets a standard deviation higher in grit had fifty-four per cent greater odds of finishing. Neither cognitive ability nor physical ability cleared the tightened bar in that model, and the paper's own section heading says so: only grit predicts completing Beast Barracks. Its discussion repeats the wording, calling it the only reliable predictor of completing the training during which attrition peaks.
That is a strong claim and it needs its fine print, so we went back to the paper for the model behind it. The figures come from the first model of its supplementary table seven, a single logistic regression with all three attributes entered together and gender, race, age and cohort year controlled, as the figure caption confirms. The table itself sits in a separate appendix we could not reach, so the caption is as close as we got.
The abstract and significance statement read differently, and the difference is worth knowing. There the authors group physical ability with grit as the noncognitive attributes more prognostic of other achievement outcomes, naming completion of initiation training and four-year graduation in the same breath. Follow that to the results and physical ability does its work on graduation, where its odds match grit's exactly, and not on the summer.
On grades the picture inverts completely, and the same paper prints it. Cognitive ability predicts academic marks several times more strongly than the trait does, and physical ability dominates physical marks the same way. Past the ninety-ninth percentile, more of the trait predicted slightly lower odds of graduating, though the curve is not symmetric and the worst odds belonged to cadets entering with the least of everything.
Then the deflation, in the authors' own words: their whole battery of cognitive and non-cognitive attributes, plus a rich set of demographic covariates, collectively explained less than four per cent of the variance in graduating four years later. Twenty-four parts in twenty-five were something else. They close by writing that there may be severe limits to how well any set of personal attributes can forecast an individual's destiny.
One thing that study cannot do, and its authors say so: the Big Five were never measured, so it runs no test of the trait against conscientiousness.
The Evidence for Keeping It
Three further results sit on the same side of the ledger and none of them is soft.
Lauren Eskreis-Winkler and colleagues in 2014 followed 677 candidates through a twenty-four-day Army Special Operations selection course that fewer than six in ten completed. The trait did add to the prediction of who finished, over the army's own measures of intelligence and fitness. It also explained roughly a quarter as much as fitness did.
In the same paper, 442 sales employees of a vacation ownership company, fewer than half of them still in post at the end. Here grit was the only personality variable to reach significance, and conscientiousness fell away to nothing beside it. The book's claim on this study is accurate and in fact conservative. The caveat travels with it: the single largest predictor of staying was how many weeks the person had already been employed.
And the thing moves. Sule Alan, Teodora Boneva and Seda Ertac ran a randomised educational intervention in two independent elementary school samples and reported, in a follow-up two and a half years on, an effect of about a fifth of a standard deviation on a standardised maths test, with outcomes measured partly through an incentivised real-effort task. Credé's review had written in 2017 that whether this could be enhanced by intervention was not yet clear. Two years later it was answered, in the construct's favour.
[Nobody Is Born Consistent](https://thewhyweproject.com/articles/nobody-is-born-consistent) argued that consistency is something a routine produces in you. This piece argues something adjacent: that the trait the culture named to explain who keeps going already had a perfectly good name.
The Author's Own Narrowing
Grit is a better book than its reputation in either direction suggests. Its author concedes limits constantly, often unprompted:
- that her theory is incomplete because it ignores coaches, teachers, circumstances and luck;
- that her age data is a single snapshot, with two rival explanations she declines to choose between;
- that reading for fun had no relationship at all to spelling ability, a null of her own that she prints;
- that her sample of exemplars excludes the people who would falsify her claim about purpose;
- that the scale is easily faked.
On two occasions she goes further than any author writing for a general readership needs to go, and the first of them is a line about the evidential standing of an entire chapter:
> "All the research I talked about in this chapter is nonexperimental."
> "There's a lot of research on parenting, and some research on grit, but no research yet on parenting and grit."
The conclusion states that this trait is far from the most important aspect of character, that goodness and greatness are different things, and that forced to choose she would put goodness first. Her published position outside the book is stronger still: her own laboratory's distribution terms discourage using her scales where faking is a concern, naming admissions and hiring, and a peer-reviewed paper she co-authored with David Yeager concluded that no available measure of personal qualities suits between-school accountability judgements. That caution belongs beside the scale every time the scale comes up.
A note on what the book is and is not being asked to carry here. It is a trade account written for a general reader, it is cited in this piece for framing, and every magnitude in the dispute here comes from the journal literature.
So, the verdict, last, because it only means anything after the rest. This trait is mostly conscientiousness when the question is how well somebody performs, and it is still worth something when the question is whether somebody is still there. Those are different questions and the field ran them together. The problem was never a failed replication, which distinguishes this case from most of what [The Psychology You Learned Didn't Replicate](https://thewhyweproject.com/articles/the-psychology-you-learned-didnt-replicate) covers. Grit replicates. What replicates is also what conscientiousness already measured.
Popular psychology has a habit of swapping a plain process for a glamorous trait, which is the move [How Long a New Habit Really Takes](https://thewhyweproject.com/articles/how-long-a-new-habit-really-takes) tracks through timelines and [Why Announcing Your Goal Can Backfire](https://thewhyweproject.com/articles/why-announcing-your-goal-can-backfire) tracks through goals. Which half of a questionnaire somebody scores on is a measurement result rather than an instruction, and once Credé's caveat is taken seriously it may not even be a real division.
What the same review hands over instead is a list with four usable items on it. The four it recommends, study skills and habits, adjustment to the place, attachment to the institution and plain attendance, are all more strongly related to grades than the total score is, and the review's own closing argument is that each of them answers to intervention.
Common questions
Is grit just another word for conscientiousness?
Close to it, for one purpose. On school performance, total scores add almost nothing once conscientiousness is accounted for, while conscientiousness still adds something in the reverse order, and the two questionnaires agree more closely than two conscientiousness questionnaires do. The perseverance half is the exception and genuinely adds. A 2020 paper on which Duckworth is third author describes the trait as a facet of conscientiousness.
Does grit predict school grades?
Weakly, on both of the pooled estimates, the larger of which is Lam and Zhou's. Rank a class by score and you would guess the order of their grades slightly better than chance. The meta-analysts note that other well-known predictors of academic performance report correlations more than twice as large, and recommend study skills, academic adjustment and class attendance as better targets for anyone hoping to shift results.
Which half of grit actually earns its keep?
Perseverance of effort, the keeping-at-it half. It predicts academic outcomes more strongly than the total score does, and it adds over conscientiousness on every academic criterion the 2017 review tested. Consistency of interest predicts very little and may suppress the half that works. The authors caution that even this increment could reflect another manifestation of conscientiousness.
Sources
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.
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.
Duckworth, A. L., Quirk, A., Gallop, R., Hoyle, R. H., Kelly, D. R., & Matthews, M. D. (2019).
Eskreis-Winkler, L., Shulman, E. P., Beal, S. A., & Duckworth, A. L. (2014). 'The grit effect: predicting retention in the military, the workplace, school and marriage.' Frontiers in Psychology, 5, article 36.
Ponnock, A., Muenks, K., Morell, M., Yang, J. S., Gladstone, J. R., & Wigfield, A. (2020).
Jachimowicz, J. M., Wihler, A., Bailey, E. R., & Galinsky, A. D. (2019). 'Reply to Guo et al.
Alan, S., Boneva, T., & Ertac, S. (2019). 'Ever Failed, Try Again, Succeed Better: Results from a Randomized Educational Intervention on Grit.' The Quarterly Journal of Economics, 134(3), 1121-1162.
Southwick, D. A., Tsay, C.-J., & Duckworth, A. L. (2020). 'Grit at work.' Research in Organizational Behavior, 39, article 100126.
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.
Duckworth, A. L., Quinn, P. D., & Tsukayama, E. (2021). 'Revisiting the Factor Structure of Grit: A Commentary on Duckworth and Quinn (2009).' Journal of Personality Assessment, 103(5), 573-575.
Lam, K. K. L., & Zhou, M. (2022). 'Grit and academic achievement: A comparative cross-cultural meta-analysis.' Journal of Educational Psychology, 114(3), 597-621.
Rimfeld, K., and colleagues (2016), Journal of Personality and Social Psychology. 2,321 United Kingdom twin pairs; genetic correlation between grit and conscientiousness of .86; grit added little over the Big Five to academic achievement.
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.
ABSENCE CLAIMS, EACH SEARCHED BEFORE IT STANDS AND EACH PHRASED IN THE ARTICLE AS A NEGATIVE FINDING FROM SEARCH. No formal rejoinder by Duckworth appeared in the Journal of Personality and Social Psychology accompanying the 2017 meta-analysis, and none accompanied Credé's 2018 Educational Researcher essay, searched on this pass and by the vault on 2026-09-24.
PAGES THAT COULD NOT BE READ, RECORDED RATHER THAN SMOOTHED OVER. pmc.ncbi.nlm.nih.gov returned a reCAPTCHA challenge; the PNAS 2019 full text was obtained instead from the Europe PMC REST full-text XML service.