Praising a child's intelligence can quietly undermine it
Carol Dweck and Claudia Mueller ran a series of experiments in the late 1990s. The finding was so counterintuitive that it still surprises educators who encounter it.
In 1998, psychologist Claudia Mueller and Carol Dweck at Columbia University gave fifth graders a moderately difficult set of problems and then praised them — but in two different ways. Half heard "You must be really smart at this." The other half heard "You must have worked really hard."
Both groups then faced a harder set of problems where most would struggle. Afterward, when given a choice of what kind of problems to try next, most children who had been praised for intelligence chose easier problems — the ones they were more likely to succeed at. Most children praised for effort chose harder ones.
The intelligence-praised group also reported enjoying the work less after the failure experience, performed worse on a final test, and were significantly more likely to misreport their scores to peers — inflating their results. The effort-praised group showed none of these patterns.
The mechanism, Dweck argued, is this: telling a child she is smart frames intelligence as a fixed trait she either has or doesn't. A setback then becomes evidence that the trait might not be there. Praising effort frames success as something the child produced through action — something repeatable and controllable.
The two praise statements are almost interchangeable in casual conversation. The effects on a child's subsequent behavior, across six separate studies, were not.