Nearly 50 years ago, psychologist Walter Mischel sat five-year-old children down at a table and gave them a simple choice: they could eat one marshmallow now, or two marshmallows later.
Mischel and his colleagues, conducting research at a nursery school on Stanford's campus, wanted to explore the nature of self-control. One at a time, each child was seated in front of a plate with one marshmallow (or another treat of their choice). If they could refrain from eating any until an adult returned, they were told, they could have two. Alternately, if they couldn't resist, they could ring a bell and an adult would return, allowing them to have just one (some, of course, just ate the marshmallow instead of ringing the bell).
These experiments generated all sorts of interesting findings, but they became famous because of something remarkable that happened years later: on average, the kids who could wait longer for another marshmallow were found to have higher SAT scores and got along better with others, used drugs at lower rates as young adults, and even had lower body mass indexes 30 years later at the outset of middle age. Waiting for two marshmallows turned out to be really, really important.
Many people have interpreted this to mean that our fate in life is predetermined. But Mischel, out with a new book — The Marshmallow Test: Mastering Self-Control — says the real lesson of these experiments is the exact opposite.
"The most important thing we learned is that self-control — and the ability to regulate one's own emotions — involves a set of skills that can be taught, and learned," Mischel says. "They're acquirable. Nothing is predetermined."
What's more, he says, these experiments provide concrete lessons about self-control we can use as adults. He argues that the same strategies a five-year-old uses to avoid eating the marshmallow can help you quit smoking, or stick to a diet, or save for retirement. Mischel recently spoke with me about these strategies.
Avoid the temptation
Watch a video of children taking the marshmallow test today and one thing becomes obvious: the more exposure to the marshmallow they let themselves have, the more likely they are to eat it. In other words, picking up and sniffing the marshmallow, or pretending to eat it, is not a strategy for staving off the temptation — it's a recipe for giving in imminently.
Kids who turned away from the marshmallow or pushed it farther away, on the other hand, were more successful. And other experiments found that if the marshmallow was covered by by a screen and the kids couldn't see it, they waited ten times longer, on average, before ringing the bell.
The reason, Mischel and other psychologists argue, is that the battle between instant gratification (one marshmallow now) and long-term prudence (two marshmallows later) is really a battle between two different systems of the brain. "There's the limbic system — the lower, more primitive brain, which responds immediately and emotionally, and allowed us as a species to survive a predator-filled environment in ancient times," Mischel says. "Then other parts of the brain, concentrated in the prefrontal cortex, allow us to do things like control our attention, and think about the future, and delay gratification." Neuroscientists often call this the brain's executive function.
In the book, Mischel likens the impulse-driven limbic system to "hot" thinking, and the slower, more rationally-driven executive function to "cool" thinking. These categories may oversimplify — in terms of the underlying brain structures — but they're still a helpful way of understanding our response to different sorts of stimuli. And various experiments have shown that exposure to the sight, smell, or taste of a temptation activate the sort of short-term "hot" thinking that make us most likely to give in to it.
There are all sorts of lessons for adults here, Mischel writes. If you're trying to quit smoking, surrounding yourself with smokers will inevitably drain your willpower and make you more likely to have a cigarette. "This is the reason why, even if you have resolutions about how many potato chips you're going to eat, once you open the bag, it's so hard to not empty it. It's because they taste good," he says. "Once you're focusing on that — rather than the long-term issues of cholesterol and obesity — it's incredibly hard to stop."
One of the most amusing aspects of watching marshmallow test videos is the things kids do to occupy themselves instead of eating. "They take off their shoes and play with their toes as if they were piano keys. They're inventing little songs, or exploring their nasal cavities," Mischel says. Sometimes, "they engage in Charlie Chaplin-like monologues, reminding themselves, 'If I ring the bell, then I don't get the two marshmallows.'"
These sorts of activities are adorable. But they're also an interesting look at a crucial strategy. In a sense, it's the flip side of avoiding temptation: distracting oneself by doing or thinking about something else entirely.
In experiments, when the researchers suggested that the children conjure a fun memory before leaving them alone, the kids waited about ten times longer before ringing the bell than otherwise. On the other hand, suggesting that the kids think about the delicious taste of the marshmallow pretty much guaranteed they'd ring the bell or eat it immediately.
The lesson is that daydreaming or distracting yourself — instead of directly thinking about the temptation — is a way of quelling the sort of "hot" thinking that leads to giving in. This, too, can be useful to adolescents and adults as well as five-year-olds, regardless of the temptation at hand.
Mentally transform your temptation
One of factors that helped kids most effectively delay gratification was remarkably simple. "If the child looked at the cookie but pretended it weren't real — they were told beforehand to put an imaginary picture frame around it and pretend it was just a picture — the child who would previously ring the bell within 30 seconds could now wait 10 or 15 minutes, on average," Mischel says.
Simply transforming the desired object inside your brain affects your response to it — and how much willpower it will require to abstain. An adult might not be as easily persuaded that a tempting item is simply a picture, but it can be transformed in plenty of other ways. "For instance, instead of a drug being a fantastic high, pretend it's a poison," Mischel says. "How you think about it makes a huge difference in cooling your desire for it."
This sort of mental transformation is often more effective than thinking about long-term consequences. "To cool your desire for a piece of chocolate fudge, instead of thinking about your next blood test, just imagine that a cockroach nibbled on it," Mischel suggests. This is why, in other countries, cigarette packs often come with graphic photos of cancer-scarred lungs, rather than mere text.
Many psychologists argue that our sense of disgust evolved to protect us from unsanitary foods. If that's the case, triggering a feeling of disgust might be effective because it taps into the power of the impulse-driven "hot" thinking, but instinctively pushes us in the opposite direction.
This sort of thinking can even be leveraged to make long-term priorities seem more like urgent temptations. One recent study, for instance, found that people committed 30 percent more money towards their 401(k) savings accounts if they saw a photo of themselves that had been digitally altered to look about 68 years old. Just being able to see themselves in old age made the need to save for retirement seem much more urgent.
Create an "if-then" plan
"If-then" planning is now a tenet of many productivity coaches, but Mischel observed it way back in the 1970s, as part of an experiment that came shortly after the marshmallow tests.
In the experiment, four- and five-year-olds were put in a room with something called Mr. Clown Box — essentially, a talking, light-up box with toys that invited kids to come play with it. Like the marshmallow test, they were told that if they could ignore the short-term temptation (and complete a more boring task), they'd be rewarded later (by getting to play with both the clown and other toys).
These young children understandably had a tough time completing their boring task. But they were dramatically more successful when the researchers suggested that if the clown talked to them, they should do something specific, like turn away from it, or tell it to shut up. In essence, this is rudimentary "if-then" planning — explicitly creating a plan to follow in the event of a specific stimulus.
"The lesson that comes out of that, and has now been studied by other people in great detail, is that very simple "if-then" plans work," Mischel says. "For instance, 'If the alarm rings at 7:00, then I will get out of bed — I won't hit snooze. If I am working on a homework assignment, then I will turn off my phone.'"
This seems simple, but research has shown it's dramatically more effective than vaguer sorts of resolutions on a consistent basis. One study found that 91 percent of participants who created an if-then exercise plan ('If it's a Monday, Wednesday, or Friday, then I will go to the gym') stuck with it, compared with just 39 who created a vague plan ('I will go to the gym more often'). "Articulating targeted plans beforehand, based on the areas of our lives in which we want to exert self-control, makes a huge difference," Mischel says.
Many different experiments have shown that, both in kids and adults, high levels of stress increase the chance we'll give in to short-term "hot" impulses instead of prioritizing long-term "cool" thinking. fMRI studies have also shown that stress affects the prefrontal cortex — the region most heavily involved in long-term thinking — more detrimentally than any other area of the brain.
Over time, chronic levels of high stress appear to alter the structure of this area, disrupting someone's baseline capacity to engage in long-term thinking even on a non-stressful day. "Kids who are living under chronic stress have to get help in having that reduced to do well in school," Mischel says. In other words, children dealing with a difficult home life can't just suddenly turn on the ability to control themselves and concentrate on a lesson when they enter the schoolhouse door. (Somewhat related is the fact that, as Mischel found and other scientists have since confirmed, kids growing up in unstable home settings are less likely to delay gratification for a good reason: there's no guarantee that an adult is telling the truth about the two marshmallows later.)
This goes for adults as well, and explains why many of us find ourselves giving in to temptation during difficult times — whether it's something as mundane as stress-eating or more destructive long-term decisions.
Self-control is more important than you realize
The marshmallow test is famous not because of the self-control strategies it's taught us, but because of the striking way it has served as an indicator of its subjects long-term success.
"We found that there are correlations between the number of seconds a child delayed the marshmallow for, and outcomes in early adolescence, including early SAT scores, and teacher and parent ratings of how well the children were doing socially and cognitively in their teens," Mischel says. Teens who had waited longer for the marshmallow also reported they were better at dealing with stress and frustration.
Later, the researchers found, adults who had waited longer reported they were better at pursuing long term goals and were found to have lower body mass indexes and were less likely to use drugs. Even more recently, brain scans of the adults who'd waited longest as kids and those who waited least showed differences in prefrontal cortex activity.
In short, being able to choose two marshmallows later instead of one marshmallow now turns out to be crucial in navigating many of life's biggest challenges. "The stuff needed to delay gratification on the marshmallow test — namely, executive function — is exactly what's needed for school success," Mischel says. Kids who have that are "ready to learn, to focus on the teacher, to concentrate, to not become distracted, to keep the goal in mind."
This also applies to many other areas. "As we grow up, we have to deal with not only not gobbling up every temptation that's in sight, but we have to be able to control our negative emotions," Mischel says. "We won't do well in kindergarten if, every time we become angry, we punch everyone around us. So the ability to recognize and inhibit one's own negative emotions is another part of executive function."
But self-control isn't a fixed trait
This is the thing that many people have gotten wrong about the marshmallow test. It may be true that, on the whole, children who waited longer for the marshmallow did better, by some measures, later in life.
But the test wasn't designed to "pass" or "fail" kids. It was designed to see what circumstances and strategies make delaying easier, and which make it harder. And more than anything, it shows that self-control is a mutable skill that's highly dependent on the choices we make and the conditions we find ourselves in. It's not a fixed trait like height or eye color, but something that we can all constantly work on — like, say, the ability to play the piano.
"For me, the marshmallow test is not an indicator that our futures are already determined when we're four years old, but that our potential for maximizing our lives involves a set of skills that are already visible and teachable at age four," Mischel says. "[It] involves a set of skills that can be taught, and learned. They're acquirable. Nothing is predetermined."
Joseph Stromberg: How did you come up with the original marshmallow test?
Walter Mischel: Well previously, in Trinidad, I had done work trying to understand what were the determinants of the choice to go for the little candy now, or wait for the big one later. In economics, this is now called "intertemporal discounting." It's a fancy way of saying, 'would you rather have $1 today, or $1.10 tomorrow?'
So for about five years, we were trying to understand what determines this choice. And clearly, one of the most important determinants is how strongly the choice maker believes that the delayed outcome will actually come. For instance, kids who live in environments in which promises are regularly broken aren't going to be very likely to wait for a bigger candy later. Other crucial determinants, we found, were the subjective value of the rewards, as well as the actual delay time.
We mapped all that out, and then I got to Stanford. And I myself had three closely-spaced daughters, from age zero to age five. So I wanted to ask a different question about self-control: once you've made the choice that you want two cookies, or two marshmallows, why are the cognitive and emotional skills that make it possible for you to see it through?
It's not about initially choosing. I can choose as I'm walking in to a restaurant, for instance, that I'm not going to have desert. But when the server comes with the dessert cart and starts describing to me how delicious the chocolate fudge cake is, my choice might be forgotten, and all of a sudden I have the cake in my mouth. So the difference between making the choice and actually having the ability to sustain it is what the marshmallow test measures.
Joseph Stromberg: In a nutshell, what did the test teach us?
Walter Mischel:The whole point of the experiment was to identify the cognitive and emotional mechanisms that make it possible for kids to engage in sustained self-control. And looking back on it, what makes the marshmallow experiments so interesting is that they tap what neuroscientists now call executive function.
It depends on a few things. You have to have a delayed goal in mind: I want to wait and not eat this marshmallow so I get an extra one. To not eat the marshmallow, you have to start by giving yourself as little chance to eat it as possible. You have to think, 'I can't start sniffing this marshmallow, I need to push this marshmallow as far away as possible so I won't be tempted.' By analogy, if we're talking about kids and homework assignments, they're not going to be able to concentrate if they keep their phone next to them — so to reach the goal, they have to turn off their phone.
And thirdly, you have to use your attention control and imagination to reach the goal too. When we're talking about four-year-old kids, they do things like turn around in the chair, so they can't see the marshmallow. They take off their shoes and play with their toes as if they were piano keys. They're inventing little songs, or exploring their nasal cavities.
When they get a little bit older — say, five — they engage in Charlie Chaplin-like monologues, reminding themselves, "If I ring the bell, then I don't get the two marshmallows." It's adorable, but it's also a fantastic window for seeing how executive function works and develops.
Joseph Stromberg: What other strategies affected whether the kids could wait or not?
Walter Mischel: A big one was imagination. If the child looked at the cookie but pretended it weren't real — they were told beforehand to put an imaginary picture frame around it and pretend it was just a picture — the child who would previously ring the bell within 30 seconds could now wait 10 or 15 minutes, on average.
What that tells us is that what enables self-control is the ability of the individual to transform the meaning of the stimulus. That's a skill that, if you have it, it's terrific. Rather than being the victim of the chocolate fudge, or the victim of the cigarette or drug, you're the one who's in charge. You can change how you represent it. For instance, instead of a drug being a fantastic high, pretend it's a poison.
Another interesting factor came out of our later experiments with something called "Mr. Clown Box" [a talking box with toys that invited kids to come play with it]. Kids came up with very simple explicit "if-then" plans beforehand — if the clown box is tempting me, I'm not going to look at him, or I'll tell him to shut up. The lesson that comes out of that, and has now been studied by other people in great detail, is that very simple "if-then" plans work. If the alarm rings at 7:00, then I will get out of bed — I won't hit snooze. If I am working on a homework assignment, then I will turn off my phone. Articulating targeted plans beforehand, based on the areas of our lives in which we want to exert self-control, makes a huge difference.
Joseph Stromberg: What were some of the other circumstances beyond kids' control that made a difference?
Walter Mischel: One of the things that make control extremely difficult is when one is experiencing chronic stress levels. One useful way to think about the brain is to divide it into two parts. There's the limbic system — the lower, more primitive brain, which responds immediately and emotionally, and allowed us as a species to survive a predator-filled environment in ancient times. Back then, long-term self-control was less important. There was no need for retirement planning.
Then the other parts of the brain, concentrated in the prefrontal cortex, allow us to do things like control our attention, and think about the future, and delay gratification.
The problem is that, often, the fast, emotional system snaps into place too quickly. When it comes to retirement planning, for instance, I'd rather use my paycheck today and buy something fun, rather than set it aside for the future and have it when I'm old.
If we want to change that, it becomes imperative to make the delayed consequences hot. So in this case, imagine yourself in old age, without any money. That's not an easy thing to do. But the point is, if you want to exert self-control, you need to make the distal consequences hot and control the immediate temptation.
Joseph Stromberg: What are some things that did the opposite — that made it more difficult for kids to wait?
Walter Mischel: All they have to do is start thinking about how delicious it is. They just have to start looking at it, and focusing on the "hot" features — how yummy it is, how chewy it is.
This is the same reason why, even if you have resolutions about how many potato chips you eat, once you open the bag, it's so hard to not empty it. It's because they taste good. Once you're focusing on that — rather than the long-term issues of cholesterol and obesity — it's incredibly hard to stop.
So the key is that we have the freedom to change how we represent our objects of desire. We can think about the potato chip being poison, rather than delicious and crunchy. How you think about it makes a huge difference in cooling your desire for it. To cool your desire for a piece of chocolate fudge, instead of thinking about your next blood test, just imagine that a cockroach nibbled on it.
Joseph Stromberg: One of the really interesting things about this research is how you and colleagues found these long-term correlations between test results and all sorts of measures. Could you tell me a bit about that?
Walter Mischel: Well, essentially, we found that there are correlations between the number of seconds a child delayed the marshmallow for, and outcomes in early adolescence, including early SAT scores, and teacher and parent ratings of how well the children are doing socially and cognitively in their teens.
In the next follow-up, when the kids were young adults between the ages of 27 and 32, we again found correlations between how well they felt they were able to pursue their long-term goals, how well they could deal with frustration, and so on. In one study, we could show there was even a correlation between seconds of delay time when they were four years old and their body mass index at age 32.
Then, in the next wave, which we conducted when they were in their early forties, we saw differences between the consistently high delay kids and the consistently low self-control kids. When we compared those two extreme groups, we saw differences in their fMRIs, in terms of activity in areas of the brain involved in resisting temptation.
What we're doing currently is working with an economist at Harvard and following this sample as they're in their early fifties to look at their economic outcomes — their credit card default rates and so on. Those data will be published sometime early next year.
Joseph Stromberg: Why do you think our ability to exert self-control is so important?
Walter Mischel: I think the stuff needed to delay gratification on the marshmallow test — namely, executive function — is exactly what's needed for school success. So if a kid is already able to control oneself in preschool, they've already got a leg up on academic and social success. With academic, it's clear: they're ready to learn, to focus on the teacher, to concentrate, to not become distracted, to keep the goal in mind. Those fundamentals are in place.
But they also have a much better chance to regulate their negative emotions in general, which is crucial for social success. As we grow up, we have to deal with not only not gobbling up every temptation that's in sight, but we have to be able to control our negative emotions. We won't do well in kindergarten if, every time we become angry, we punch everyone around us. So the ability to recognize and inhibit one's own negative emotions is another part of executive function. These things add up over the course of development.
Joseph Stromberg: What's the relative role of genetics and environment in all this? How much of our ability to exert self-control seems to be predetermined?
Walter Mischel: It's becoming increasingly clear that the nature-nurture question really needs to be rephrased: how do nature and nurture interact to produce who we are?
We're not a genetic text that simply plays out from birth on. We are dealt a genome, but there are also switches that determine which parts of our DNA is turned on, and which parts are turned off. These switches are profoundly influenced by the environment — both the biological environment and the psychological environment, so the food we eat and toxins we inhale, but also the emotional experiences we have, the chronic stress levels we experience.
So in essence, environments are much more influential than we used to think they were, and genetics are much more malleable than we used to think we were. Both are continuously interacting, over the entire course of life. For instance, if we become depressed, our biology actually changes. Our brain structures change. The same thing happens if we have vibrant and joyous experiences.
Joseph Stromberg: What do you think are the public policy implications of this sort of research?
Walter Mischel: We've reached a point where there is a pretty clear consensus about executive function, and how it hinges on the reduction of chronic stress levels. This is essential for education. Kids who are living under chronic stress have to get help in having that reduced to do well in school. And secondly, they need to have learning experiences in which their executive skills are maximized.
This means that school curricula, beginning as early in life as possible, need to include those ingredients. There are a number of programs that have demonstrated — for example, the TK diamond tools of the mind program — that it is possible for very young children to learn strategies, as part of games in nursery school, that will have important consequences and improve their executive function skills.
These ideas have been put into place in a number of school programs. Here's a simple example: the research suggests that it's very important for us to be able to cool down our negative emotions before go violent. So some classrooms are now featuring something called "thinking chairs." If a child feels he or she is about to lose it, they can go sit there. It's not like standing in the corner, it's not a punishment, it's a chance for a kid to withdraw for a few minutes and think cool thoughts — sometimes the kids are suggested they can think about their anger floating off like a balloon, or count backwards from 100.
Joseph Stromberg: What are some of the most common criticisms of the test? How do you respond to them?
Walter Mischel: It's often said that the marshmallow test doesn't take trust into account. This is dead wrong. The work started, in Trinidad, with us showing the tremendously important role of trust in making these choices.
Another is that people think the test means you should delay gratification all the time. That's not it at all. If we did that, life wouldn't be worth living. This isn't about constant, rigid self-control — it's about making decisions on when it's worth waiting, and when it's time to ring the bell.
In terms of the long-term correlations, lots of people misinterpret this to mean it's true in every case. But there are lots and lots of exceptions. The correlations just show that, on average, someone who could delay longer during the marshmallow test had different outcomes than someone who could not. It doesn't necessarily mean anything about your child's destiny.
For me, the the most important thing we learned is that self-control — and the ability to regulate one's own emotions — involves a set of skills that can be taught, and learned. They're acquirable. Nothing is predetermined. For me, the marshmallow test is not an indicator that our futures are already determined when we're four years old, but that our potential for maximizing our lives involves a set of skills that are already visible and teachable at age four.
Joseph Stromberg: Why do you think this research has gained so much cultural currency over the last few decades?
Walter Mischel: I think there's something about "two marshmallows later or one marshmallow now" that's intuitively appealing. Everyone can imagine themselves in those situations. It may not be marshmallows, but we have all experiences those kinds of dilemmas. It feels natural, and feels right.
When we talk about executive function, or the prefrontal cortex, that's less appealing. But there's something really vivid about marshmallows.
Note: This interview has been edited for length and clarity.