Psychologist Walter Mischel developed the “marshmallow test” in the 1960s as a way of studying preschoolers’ ability to delay gratification. Children were left alone in a room, sitting in front of a single marshmallow, and provided with a bell that they could ring at any time to call back the researcher and then eat the marshmallow. But if they could refrain until the researcher came back of her own accord, they would be given two marshmallows. You have doubtless heard of the test: as a seminal experiment that spawned countless research programmes on self-control, it is a staple of works on the topic. Now, with The Marshmallow Test, which combines scholarly argument with self-help tips, policy proposals and anecdotes from his career, Mischel himself joins the fray.
As a postgraduate student in clinical psychology keen to make a difference to people’s lives, Mischel recounts, he thought he could apply some of the concepts he was learning to work he was doing with a group of disadvantaged teenagers. One evening, the teenagers seemed particularly attentive, and before long Mischel spotted that one of them had set the back of his jacket on fire. The concepts Mischel was being taught were irrelevant to these young people, he realised, and this spurred him to pursue a research career that would allow him to help people to improve their lives.
If self-control is a skill that some of us have more of, then how can successful people such as Bill Clinton or Tiger Woods have had such public lapses?
The starting point for his self-control research was his close observation of the children in the marshmallow test. Tactics used by those who succeeded in resisting temptation included closing their eyes, avoiding looking at the treat, repeating aloud that they could have two treats if they waited, pushing the bell away, singing to themselves and even falling asleep. One particularly successful tactic involved distraction. Mischel investigated further and discovered that when the marshmallows were covered, children found it much easier to wait. He wondered if they could use abstraction to similar effect, drawing on Freud’s idea that creating a “hallucinatory image” of objects of desire (for Freud, the mother’s breasts) enabled infants to inhibit the impulse for immediate gratification. Indeed, Mischel found that giving the children a life-sized, realistic photo instead of an actual marshmallow doubled the average time that children waited. It was possible to get the same increase in waiting time when an actual marshmallow was in front of them if they were instructed to use mental abstraction: “imagine it’s a picture; put a frame around it in your head”. What is more, the effect of substituting a picture was reversed if the children were told to make believe that the reward in the picture was real.
He explains these results by invoking the sort of “two systems” hypothesis popularised by Daniel Kahneman in Thinking, Fast and Slow: the idea that the brain has one fast automatic system for decision-making and another slower and effortful one. Mischel connects these to levels of emotional arousal, calling them the “hot” and “cool” systems. Marshmallows have arousing and motivating “hot” qualities, such as being tasty, which make you want to eat them, and other non-emotional cognitive “cool” features, such as being white and soft. Arousing the hot representations automatically triggers the action of eating the marshmallow, whereas cool representations support the system involving cognitive processing, which enables delay. Changing how you mentally represent a stimulus can change its impact on what you feel and what you do. When Mischel gave up smoking, he observes, he enlisted a memory of a cancer victim being taken for radiation, which he conjured up when he felt a craving.
Nevertheless, it is not always a good thing to be shielded from emotional cues. Self-control can have paradoxical effects. People with high self-control tend to stay calm when making risky decisions, so they are shielded from the stress that people with low self-control feel on making losses. When making a series of risky decisions that are turning out badly, stress and anxiety leads people with low self-control to quit before they go broke. In contrast, under experimental conditions, people who were high in self-control failed to react to danger signals and lost more money than people with low self-control – which Mischel offers as an intriguing explanation for behaviour in the financial sector leading up to the 2008 financial crisis.
Twelve years after the original marshmallow test, follow-up studies revealed that, as adolescents, the children who had waited longer displayed more self-control, had better social and cognitive functioning, and scored higher on intelligence tests. As young adults, they had reached higher educational levels, were less likely to use drugs and had a lower body mass index. Brain scans showed that they had more activity in parts of the brain associated with effective problem solving than those who had not been able to wait.
However, Mischel is at pains to stress that self-control is not genetically predetermined, but can be taught. He is particularly clear on the importance of mindset, the idea that a person’s theory about the possibility of mastery influences how much self-control he or she can actually exert. He cites research by Carol Dweck that shows that if one is to successfully develop self-control, then it is important to think of it as a skill that can be honed and not as an ability that is fixed at birth.
If self-control is a skill that some of us have more of, then how can successful people such as Bill Clinton or Tiger Woods, who must have plenty of self-control in order to achieve their professional success, have had such public lapses? Mischel’s answer builds on his own devastating critique of traditional notions of personality: he proposes that personality consists of “if-then” signatures. These work as follows: someone who is aggressive will not be aggressive in all situations; rather, he or she will be aggressive in response to particular cues. For instance, an aggressive child may be aggressive when other children tease her or when adults punish her. She need not be aggressive in response to all possible cues for aggression, but she will consistently be aggressive in response to particular cues. If those cues occur, then she will exhibit aggressive behaviour. Extrapolating to self-control, it is possible to exercise self-control in many areas of one’s life but to lapse when presented with a specific type of cue, such as sexual temptation. It also follows that it may be valuable to keep a diary in order to learn the cues that lead to destructive behaviours.
The Marshmallow Test is a tour de force. Despite its serious academic content, it wears its learning lightly. Mischel has never been one to shy away from conflict, taking on not only the personality psychologists but also the clinicians to whom he dared to point out that their professional diagnosis was not predictive of which psychiatric patients would be readmitted – instead, the most reliable indicator was the weight of their files. And here, he takes issue with the popular theory of ego depletion, the idea that self-control draws on a limited pool of mental energy. It is valuable for readers to be reminded that science is contested, as popular books frequently give a false impression of consensus.
Although Mischel’s experiments have become justly famous, it has been largely through the popularising work of other people. It is to be hoped that this book will make him as much of a household name as his marshmallows are.
In 1938, the eight-year-old Walter Mischel – today the Robert Johnston Niven professor of humane letters in the department of psychology at Columbia University – fled Nazi-occupied Vienna with his family for the US. His parents opened a shop in Brooklyn, and Mischel studied psychology at New York University and then Ohio State University.
His first paper detailing the “marshmallow test”, conducted at Stanford University’s Bing Nursery School, was published in the Journal of Personality and Social Psychology in 1972. Mischel then moved on to other subjects, thinking, as he recalled in a recent interview, that “there are only so many things you can do with kids trying not to eat marshmallows”. But serendipitous follow-up research led to papers in Science (1989) and Developmental Psychology (1990), and informed countless studies by others.
Indeed, as Michael Bourne noted in The New York Times, the marshmallow test is not only a tale that “captured the public imagination because it is a funny story, easily told”, but is also an object of fascination academy-wide. “If you doubt the ubiquity of the Mischel study,” said Bourne, “try this simple experiment: put a few social-policy geeks in a room and ask them about willpower, then see how long it takes before somebody brings up the four-year-olds and then the marshmallows. My bet is you wouldn’t have to wait more than a minute.”
Mischel sees the studies not as predictors of later success or failure, but as clues to self-improvement strategies. He told the New Yorker, “What we’re really measuring with the marshmallows isn’t willpower or self-control. It’s much more important than that. This task forces kids to find a way to make the situation work for them. They want the second marshmallow, but how can they get it? We can’t control the world, but we can control how we think about it.”
The Marshmallow Test: Understanding Self-control and How to Master It
By Walter Mischel
Bantam, 336pp, £20.00 and £11.98
ISBN 9780593071311 and 97814481542 (e-book)
Published 25 September 2014
The Marshmallow Study Revisited
October 11, 2012
Delaying Gratification Depends as Much on Nurture as on Nature
For the past four decades, the "marshmallow test" has served as a classic experimental measure of children's self-control: will a preschooler eat one of the fluffy white confections now or hold out for two later?
Now a new study demonstrates that being able to delay gratification is influenced as much by the environment as by innate ability. Children who experienced reliable interactions immediately before the marshmallow task waited on average four times longer12 versus three minutesthan youngsters in similar but unreliable situations.
Study reenactment: Evelyn Rose, 4, of Brighton, N.Y. participates in a reenactment of the marshmallow experiment. The study found that children's decisions to delay gratification is influenced as much by the environment as by their innate capacity for self-control. The study was conducted at the University of Rochester Baby Lab. Photo by J. Adam Fenster / University of Rochester
"Our results definitely temper the popular perception that marshmallow-like tasks are very powerful diagnostics for self-control capacity," says Celeste Kidd, a doctoral candidate in brain and cognitive sciences at the University of Rochester and lead author on the study to be published online October 11 in the journal Cognition.
"Being able to delay gratificationin this case to wait 15 difficult minutes to earn a second marshmallownot only reflects a child's capacity for self-control, it also reflects their belief about the practicality of waiting," says Kidd. "Delaying gratification is only the rational choice if the child believes a second marshmallow is likely to be delivered after a reasonably short delay."
The findings provide an important reminder about the complexity of human behavior, adds coauthor Richard Aslin, the William R. Kenan Professor of brain and cognitive sciences at the University. "This study is an example of both nature and nurture playing a role," he says. "We know that to some extent, temperament is clearly inherited, because infants differ in their behaviors from birth. But this experiment provides robust evidence that young children's action are also based on rational decisions about their environment."
The research builds on a long series of marshmallow-related studies that began at Stanford University in the late 1960s. Walter Mischel and other researchers famously showed that individual differences in the ability to delay gratification on this simple task correlated strongly with success in later life. Longer wait times as a child were linked years later to higher SAT scores, less substance abuse, and parental reports of better social skills.
Children who experienced unreliable interactions with an experimenter waited for a mean time of three minutes and two seconds on the subsequent marshmallow task, while youngsters who experienced reliable interactions held out for 12 minutes and two seconds. Error bars show 95% confidence intervals, meaning that the findings statistically are highly reliable. Credit: University of Rochester
Because of the surprising correlation, the landmark marshmallow studies have been cited as evidence that qualities like self-control or emotional intelligence in general may be more important to navigating life successfully than more traditional measures of intelligence, such as IQ.
The Rochester team wanted to explore more closely why some preschoolers are able to resist the marshmallow while others succumb to licking, nibbling, and eventually swallowing the sugary treat. The researchers assigned 28 three- to five-year-olds to two contrasting environments: unreliable and reliable. The study results were so strong that a larger sample group was not required to ensure statistical accuracy and other factors, like the influence of hunger, were accounted for by randomly assigning participants to the two groups, according to the researchers. In both groups the children were given a create-your-own-cup kit and asked to decorate the blank paper that would be inserted in the cup.
In the unreliable condition, the children were provided a container of used crayons and told that if they could wait, the researcher would return shortly with a bigger and better set of new art supplies for their project. After two and a half minutes, the research returned with this explanation: "I'm sorry, but I made a mistake. We don't have any other art supplies after all. But why don't you use these instead?" She then helped to open the crayon container.
Next a quarter-inch sticker was placed on the table and the child was told that if he or she could wait, the researcher would return with a large selection of better stickers to use. After the same wait, the researcher again returned empty handed.
The reliable group experienced the same set up, but the researcher returned with the promised materials: first with a rotating tray full of art supplies and the next time with five to seven large, die-cut stickers.
The marshmallow task followed, with the explanation that the child could have "one marshmallow right now. Or if you can wait for me to get more marshmallows from the other room you can have two marshmallows to eat instead." The researcher removed the art supplies and placed a single marshmallow in a small desert dish four inches from the table's edge directly in front of the child. From an adjoining room, the researchers and the parent observed through a computer video camera until the first taste or 15 minutes had lapsed, whichever came first. All children then received three additional marshmallows.
"Watching their strategies for waiting was quite entertaining," says Holly Palmeri, coauthor and coordinator of the Rochester Baby Lab. Kids danced in their seats, sang, and took pretend naps. Several took a bite from the bottom of the marshmallow then placed it back in the desert cup so it looked untouched. A few then nibbled off the top, forgetting they could then no longer hide the evidence since both ends were eaten, she said.
"We had one little boy who grabbed the marshmallow immediately and we thought he was going to eat it," recalled Kidd. Instead he sat on it. "Instead of covering his eyes, he covered the marshmallow."
Children who experienced unreliable interactions with an experimenter waited for a mean time of three minutes and two seconds on the subsequent marshmallow task, while youngsters who experienced reliable interactions held out for 12 minutes and two seconds. Only one of the 14 children in the unreliable group waited the full 15 minutes, compared to nine children in the reliable condition.
"I was astounded that the effect was so large," says Aslin. "I thought that we might get a difference of maybe a minute or so You don't see effects like this very often."
In prior research, children's wait time averaged between 6.08 and 5.71 minutes, the authors report. By comparison, manipulating the environment doubled wait times in the reliable condition and halved the time in the unreliable scenario. Previous studies that explored the effect of teaching children waiting strategies showed smaller effects, the authors report. Hiding the treat from view boosted wait times by 3.75 minutes, while encouraging children to think about the larger reward added 2.53 minutes.
The robust effect of manipulating the environment, conclude the authors, provides strong evidence that children's wait times reflect rational decision making about the probability of reward. The results are consistent with other research showing that children are sensitive to uncertainly in future rewards and with population studies showing children with absent fathers prefer more immediate rewards over larger but delayed ones.
The findings, says Kidd, are reassuring. She recalls reading about the predictive power of these earlier experiments years ago and finding it "depressing." At the time she was volunteering at a homeless shelter for families in Santa Ana, California. "There were lots of kids staying there with their families. Everyone shared one big area, so keeping personal possessions safe was difficult," she says. "When one child got a toy or treat, there was a real risk of a bigger, faster kid taking it away. I read about these studies and I thought, 'All of these kids would eat the marshmallow right away.' "
But as she observed the children week after week, she began to question the task as a marker of innate ability alone. "If you are used to getting things taken away from you, not waiting is the rational choice. Then it occurred to me that the marshmallow task might be correlated with something else that the child already knowslike having a stable environment."
So does that mean that if little ones gobble up desert without waiting, as is typical of preschoolers, parents should worry that they have failed to be role models of reliability every minute?
Not necessarily, say the researchers. "Children do monitor the behavior of parents and adults, but it is unlikely that they are keeping detailed records of every single action," says Aslin. "It's the overall sense of a parent's reliability or unreliability that's going to get through, not every single action."
Adds Kidd: "Don't do the marshmallow test on your kitchen table and conclude something about your child. It especially would not work with a parent, because your child has all sorts of strong expectations about what a person who loves them very much is likely to do."