You have probably heard the story: a four-year-old who can resist a marshmallow is quietly headed for a better life, and one who grabs it is not.
In 2018, a large new study took that idea apart. A team led by Tyler W. Watts at NYU Steinhardt, with Greg J. Duncan and Haonan Quan at UC Irvine, ran a fresh version of the famous “marshmallow test.”
It was a conceptual replication, meaning they tested the same idea rather than copying the exact setup. They looked at 918 children whose ability to wait for a treat had been measured at about four and a half years old.
The link between waiting and later success was there. It was just far smaller than the story we have all been told.
That matters because the marshmallow test has traveled a long way from the lab. It shows up in parenting books, TED talks, and the small pang of worry many of us feel when a child grabs the cookie the second it lands on the table. So it is worth being clear about what the original study claimed, what this larger one found, and what neither can tell you about your own child.
What the original marshmallow test actually claimed
In 1990, Yuichi Shoda, Walter Mischel, and Philip Peake reported that preschoolers who held out longer before eating a marshmallow later scored higher on the SAT and did better as teenagers. The numbers looked strong. Wait time lined up with SAT math scores at around 0.57 and verbal scores at around 0.42, big enough to make the finding feel almost like a law of nature.
But the group of children was small and unusual. They came from a nursery school at Stanford, tested between 1968 and 1974, mostly the kids of faculty and staff. Only a few dozen had follow-up SAT scores. Mischel and his colleagues knew this and urged caution. The wider culture did not. One number, waiting time at age four, turned into a tidy prophecy about willpower and success.
What the 2018 replication found
Watts and his colleagues used a much larger and more varied group of children from the NICHD Study of Early Child Care and Youth Development, measured at about four and a half. Because family circumstances matter so much here, they zeroed in on the children of mothers who had not completed college, a group of 552 kids, roughly ten times the size of the original study.
Once the researchers accounted for family background, a child’s early thinking skills, and the home environment, the link was small and disappeared, more or less ceasing to mean anything.
One detail rarely survives the retellings. This was a new take on the idea, not a re-staging of the original. Here, children were counted as delaying if they waited 7 minutes instead of 20, and the data and methods differed from Mischel’s. Do not that this is one careful study, not a final verdict.
Why family background swallows the effect
The plainer way to read it is this. A child’s ability to sit and wait at age four is not some standalone trait. It grows out of everything around them: how stable things are at home, how often the promise of “wait and you’ll get more” has actually been kept, what a parent’s own schooling and resources look like. Those same conditions also shape how a teenager does in school.
So the waiting and the later achievement both trace back to the same roots. When you account for family circumstances, most of the apparent power drains out of the marshmallow itself. The willpower was never the engine. It was more like a reading on a gauge that reflects a lot of other things at once.
What the study does not prove
This is a study about links, not causes, and it cannot tell us that self-control has no value. Some researchers think the 2018 team may have overcorrected. In a published commentary, Sabine Doebel, Laura Michaelson, and Yuko Munakata argued that the researchers may have controlled for too much.
The study’s lead author is careful here too. Watts argues that “these new findings should not be interpreted to suggest that gratification delay is completely unimportant, but rather that focusing only on teaching young children to delay gratification is unlikely to make much of a difference.” Waiting is not worthless, but treating it as the one lever to pull is what the evidence doesn’t support.
What delay of gratification still is, and is not
If you have ever watched a small child bargain with themselves over a treat, you have seen something real. Self-control is a genuine ability, and it develops. What this work pushes back on is a narrower, stranger belief: that a single waiting-room test at age four is a keyhole you can peek through to see a child’s whole adult life.
Watts sums up the practical point in terms of where effort is best spent. He suggests that “if intervention developers hope to generate the kinds of improvements associated with the original marshmallow study, it is likely to be more fruitful to target the broader cognitive and behavioral abilities related to gratification delay.” Teaching a child to wait, without building the wider thinking and behavior skills underneath it, probably won’t move much. The waiting sits downstream of things that are broader and slower to build.
For a parent, that lands as something close to relief. The child who eats the cookie right away has not failed a test of their future, and the child who waits has not passed one. What the bigger study gives back is permission to stop reading a four-year-old’s patience as a forecast, and to keep caring about the ordinary, unglamorous conditions that shape a kid over years rather than seconds.
