Praising effort or intelligence, an experiment and its aftermath
A series of experiments from 1998 made the form of a compliment a research topic in its own right. Later work confirmed a mechanism and markedly reduced its size.
The original experiment
In 1998 Claudia Mueller and Carol Dweck published six studies conducted with pupils aged around ten. After a first set of problems, the children received praise directed either at their intelligence or at their effort, then faced considerably harder problems.
After failing, the children praised for their intelligence persisted less, reported less enjoyment and performed less well on a final set. Those praised for their effort showed the opposite profile on the same measures.
One of the studies added an unexpected measure: when invited to report their score to pupils at another school, the children praised for their intelligence were more likely to exaggerate it. This detail did much to spread the result beyond academic circles.
The proposed mechanism
The authors' interpretation draws on the distinction, developed by Carol Dweck, between a view of intelligence as a fixed ability and a view of it as an ability that develops. Praise for intelligence would steer the child towards the former.
Within this framework, a later failure becomes information about an ability judged to be stable, which would make giving up more likely. Praise for effort, by contrast, would tie success and failure to a variable the child controls.
Observations in the family
A longitudinal study conducted by Elizabeth Gunderson, Susan Levine and their colleagues recorded parents with their children between the ages of one and three. The share of praise directed at the process was associated, several years later, with a more malleable view of abilities in the child.
The result is correlational. Parents who use this type of praise more probably differ on other dimensions, which the recordings do not make it possible to separate out.
What the syntheses found
Two meta-analyses published in 2018 by Victoria Sisk and her colleagues found a weak association between views of ability and academic achievement, and very small average effects of interventions aimed at changing those views.
Not all these syntheses concern praise itself: they mainly assess views of ability, which praise is supposed to influence. The link between the form of praise and those views rests mainly on laboratory experiments and correlational observations.
A national American study published in 2019, based on a brief online intervention, found a small effect on grades, concentrated among the lowest-achieving pupils. In addition, at least one attempt at a direct replication of the 1998 experiments failed to find their effects.
What remains established
The mechanism described has not disappeared, but its size outside the laboratory is well below what the spread of the result implied. It is a modest effect that varies across populations, not a general lever.
The case illustrates a common trajectory: a clear-cut experiment, wide dissemination, then larger studies that confirm the direction of the effect while sharply reducing its importance.
Key points
The 1998 experiments contrast praise for intelligence with praise for effort, with clear effects in the laboratory.
The 2018 meta-analyses find associations and intervention effects that are very small on average.
The effect persists in some populations, but on a scale well below its reputation.
This article describes documented mechanisms and disagreements. It recommends no position, no party and no vote, and commits to no reading other than that of the work it cites.
And you, where do you stand?
The Civix test situates your convictions on six dimensions and matches you with one of the twelve profiles, without ever telling you what to think.
Start my test