The famous Stanford 'marshmallow test' suggested that kids with better self-control were more successful. But it's being challenged because of a major flaw.
- The Stanford marshmallow tests have long been considered compelling evidence for the need to teach kids how to delay gratification and exercise restraint.
- Some studies have shown that a child's ability to resist temptation (by not immediately eating a marshmallow, for example) can predict their future success in school and life.
- But new research is beginning to question the original Stanford findings, and suggests instead that a kid's success may have more to do with their early environment.
For decades, psychologists have suggested that if a kid can't resist waiting a few minutes to eat a marshmallow, they might be doomed in some serious, long-term ways.
In the early 1970s the soft, sticky treat was the basis for a groundbreaking series of psychology experiments on more than 600 kids, which is now known as the marshmallow study.
It worked like this: Stanford researchers presented preschoolers with a sugary or salty snack . Sometimes the kids were placed in front of a marshmallow; other times it was a different food, like a pretzel or cookie. Then, the children were told they'd get an additional reward if they could wait 15 or 20 minutes before eating their snack. If they held off, they would get two yummy treats instead of one.
Researchers then traced some of the young study participants through high school and into adulthood. They discovered that a kid's ability to resist the immediate gratification of a marshmallow tended to correlate with beneficial outcomes later, including higher SAT scores, better emotional coping skills, less cocaine use, and healthier weights.
Ever since those results were published, many social scientists have trumpeted the marshmallow-test findings as evidence that developing a child's self-control skills can help them achieve future success.
But the science of good child rearing may not be so simple.
A new troupe of researchers is beginning to raise doubts about the marshmallow test. They've designed a set of more diverse and complex experiments that show that a kid's ability to resist temptation may have little impact on their future as a healthy, well-adapted adult.
The original marshmallow experiment had one fatal flaw
For a new study published last week in the journal Psychological Science, researchers assembled data on a racially and economically diverse group of more than 900 four-year-olds from across the US. The data came from a nationwide survey that gave kindergartners a seven-minute long version of the marshmallow test in 1998 and 1999. For the updated test, kids got to choose their preferred treat: M&Ms, marshmallows, or animal crackers.
The original studies at Stanford only included kids who went to preschool on the university campus, which limited the pool of participants to the offspring of professors and graduate students. So for this new study, the researchers included data on preschoolers whose parents did not have college degrees, along with those whose parents had more higher education.
The statisticians found that generally speaking, kids who showed greater self-control when presented with a treat like a marshmallow or candy seemed to be marginally better at math and reading by age 15. But it wasn't predictive of better overall behavior as a teen.
Then the number scientists crunched their data again, this time making only side-by-side comparisons of kids with nearly identical cognitive abilities and home environments. They took into account socio-economic variables like whether a child's mother graduated from college, and also looked at how well the kids' memory, problem solving, and verbal communication skills were developing at age two.
They found that when all of those early childhood measures were equal, a young kid's ability to wait to eat a marshmallow had almost no effect on their future success in school or life.
"Take two kids who have the same ethnicity, the same gender, the same type of home environment, the same type of parents, the same sort of general cognitive ability, measured very early on," lead study author Tyler Watts told Business Insider as he explained his new study. "One of them is able to wait longer on the marshmallow test. Our results suggest that it doesn't matter very much, once you adjust for those background characteristics."
That's an important finding because it suggests that the original marshmallow test may only have measured how stable a child's home environment was, or how well their cognitive abilities were developing.
If a marshmallow test is only a "symptom of all this other stuff going on," as Watts put it, then improving a kid's ability to resist a marshmallow is no silver bullet for success.
"Just narrowly focusing on this one skill, without taking into consideration the broader elements of a child's life, probably isn't likely to make a big difference down the road, based on our results," Watts said.
Other new studies call the marshmallow test into question, too
Other new research also suggests that kids often change how much self-control they exert, depending on which adults are around. A 2012 study from the University of Rochester found that if kids develop trust with an adult, they're willing to wait up to four times longer to eat their treat.
"It occurred to me that the marshmallow task might be correlated with something else that the child already knows - like having a stable environment," one of the researchers behind that study, Celeste Kidd, said in 2012. "If you are used to getting things taken away from you, not waiting is the rational choice."
Researchers have recently pointed out additional culturally significant quirks in the marshmallow test.
A group of German researchers compared the marshmallow-saving abilities of German kids to children of Nso farmers in Cameroon in 2017. They found that the Cameroonian children were much better at restraining themselves from eating treats than German kids. The researchers behind that study think the hierarchical, top-down structure of the Nso society, which is geared towards building respect and obedience, leads kids to develop skills to delay gratification at an earlier age than German tots. Kids in Germany, on the other hand, are encouraged to develop their own interests and preferences early on.
Some new data also suggests that curiosity may be just as important as self-control when it comes to doing well in school. Scientists who've studied curious kids from all walks of life have discovered that inquisitive question-askers performed better on math and reading assessments at school regardless of their socioeconomic background or how persistent or attentive they were in class.
These findings all add to a fresh and compelling pile of scientific evidence that suggests raising high-performing kids can't be boiled down to a simple formula. Nor can a kid's chances of success be accurately assessed by how well they resist a sweet treat.