• Visit our partners: Our Partners:
  • Visit our partners: Our Partners:

The Problem with the Marshmallow Test: Why Self-Control is Overrated

According to Christian tradition, the original sin stemmed from a lack of willpower. In the Garden of Eden, Eve saw the delicious-looking fruit of the knowledge of good and evil, and she simply couldn’t resist.

Whether or not you believe in the old Bible story, there’s no doubt that at least one of its lessons has pervaded western culture: willpower is the key to a good life. Everyone has an angel on one shoulder and a demon on the other. Those who listen to the angel more often will do better than those who listen to the demon.

Walter mischel
Walter mischel

In the 1970s and 80s, a psychologist named Walter Mischel added fuel to this fire. He ran an experiment known as the Marshmallow Test, which challenged children to delay eating a marshmallow for just 15 minutes. Follow-up research showed that the children who passed this test lived successful, happy lives. For the children who failed it, depression and addiction were the more likely outcomes. 

This theory influenced public policy, business decisions, and parenting, among many other fields. Yet, today, it’s become clear that this dramatic, all-encompassing conclusion is based on a gross misinterpretation. Replication studies have clarified the facts—the marshmallow test is not what most people thought it was. More importantly, humankind’s understanding of self-control and willpower is entirely flawed.

Let’s explore.

The psychologist Walter Mischel was a professor and researcher at Stanford University from 1962 to 1983. In the 1960s, Mischel ascended to prominence in his field with his research on personality theory. He published a book detailing his beliefs on how personalities varied in different situations. But, as he continued his research, Mischel became absorbed with studying one particular character trait: self-control. He questioned the physical mechanisms that allowed some people to delay gratification in the face of strong temptations. 

In 1970 and ’72, Mischel ran two experiments that shed new light on this field and eventually changed how people understood willpower. Despite some variances, both tests were based on a similar procedure. All of the participants were children between three to six years old. The experimenter would lead the child into a room, show them a tasty treat like a marshmallow or an Oreo, and inform them of the rules. The child could eat the treat whenever they wanted. But, if they abstained for just 15 minutes, then they would receive a second treat. 

Marshmellow test
Marshmallow Experiment. Source: tmcblog.com

To ensure accuracy, researchers eliminated any children who couldn’t correctly repeat the directions. Throughout the series of experiments, Mischel and his team altered a handful of variables to understand which situations best allowed the children to delay gratification. For instance, the experimenters varied the number of distractions to help each child pass the waiting period. Some were given toys, others were encouraged to imagine stories or solve simple riddles. Still, others were left in empty rooms with nothing to distract themselves with. Another variable was the visibility of the reward. Sometimes the treats were placed in an open container on the table. At other times, they were hidden beneath an opaque box. In one iteration, the treats weren’t in the room, and the child needed to press a button for the experimenter to bring them in. 

In Mischel’s initial hypothesis, focusing on the reward would motivate the children to delay gratification. By the time the experiments were complete, Mischel’s assertion was proven entirely wrong. Instead of inspiring self-control, the children who focused on the treats had the hardest time waiting the full fifteen minutes. This was true whether or not the marshmallows were in the room. Children who were encouraged to picture the absent reward performed just as poorly as those who had to stare at a single marshmallow for the trial’s duration. Conversely, the kids who best delayed gratification were those who ignored the rewards, whether by playing with a toy, closing their eyes, or singing made-up songs. 

The result was clear. The best way to delay gratification was to actively avoid acknowledging the potential rewards. Mischel and his team published a series of papers detailing their experiment’s outcome. Behavioral psychologists took note, but it was by no means a groundbreaking study. For the next fifteen years, Mischel and his colleagues moved on to other things. 

Then, in 1988, Mischel returned to his earlier study. He was curious about how his test subjects, now in their late teens or early twenties, had panned out. Mischel contacted the families of the children involved in the experiment, asking them questions about their achievements in school and work, then comparing those outcomes to the results of his earlier test. It was at this time that the psychologist noticed something truly remarkable. Mischel discovered a correlation between delaying gratification on the marshmallow test and achieving positive outcomes later in life. According to the data, the kids who waited the full fifteen minutes had better SAT scores, attended better colleges, received better grades, and had better jobs. On the contrary, the children who performed the worst on the test were more likely to have lower test scores or suffer from addiction and joblessness.

Mischel and his colleagues now knew that they had something big. They published a paper that detailed the strong correlation between delayed gratification and success. Still, the team was careful to discourage drawing drastic conclusions. Mischel noted that his sample size, about 130 children, likely exaggerated the connection. With a larger pool of subjects, the correlation would be less significant. By all accounts, the team shared their results responsibly, refraining from suggesting that schools or businesses base new policies on their findings.

But that didn’t matter. Within months of publishing, the results of Mischel and Co’s paper spread across the country from psychology journals to daily newspapers to self-help books. They all seemed to focus on one portion of the conclusion: delaying gratification was a key predictor of success. As is so often the case in popular psychology, Mischel’s findings were seen as a panacea, a one-way ticket to the high life. 

Americans looked for ways to apply the lesson to their own lives and those of their children. Parents taught delayed gratification at home. Educators integrated cognitive-behavioral studies in preschool classrooms. Policy experts insisted that the government create laws that reflected this new paradigm. Even successful businesspeople incorporated tests of willpower and delayed gratification into their hiring and training processes. 

As the ideology permeated across culture, the reverse of Mischel’s findings were also treated as fact. Children who did poorly in school were told they needed more self-control. Addiction was blamed on a lack of willpower. Even those suffering from mental health ailments were encouraged to simply snap out of it or straighten up. In many ways, these beliefs still have a strong hold on many cultures, even if their veracity has been thoroughly disproven.

Yuichi Shoda

Yet, these applications of the marshmallow test’s results were based on a thorough misunderstanding. On the simplest level, Mischel’s colleague Yuichi Shoda pointed out that the study made no claims about the possible outcomes of teaching delayed gratification. After all, imposing self-control isn’t actually self-control, is it? 

Furthermore, while succeeding in the marshmallow test indicated the desirable quality of self-control, it was only one way of showing it. In Mischel’s words, “the idea that your child is doomed if she chooses not to wait for her marshmallows is really a serious misinterpretation.” 

But, the marshmallow test’s most significant issue was one that Mischel was well aware of: the lack of diversity amongst his subjects. The researcher pointed out that his test reflected more than just self-control. It was a reflection of the child’s background. In the case of the Marshmallow Test, all of the children were enrolled at the nearby Bing Nursery School, ran by Stanford University. Most of the kids were children of Stanford faculty and thus came from similar backgrounds. Yet, when the subtle differences between their backgrounds were controlled for, the correlation between delayed gratification and success decreased.

The impact of this adjustment was shown much more clearly in a 21st-century study done by an NYU professor named Tyler Watts. Watts dug up data from a replication marshmallow test ran by the National Institue of Health in 1990. This NIH test was similar to Mischel’s but with two differences. First, it had a much more diverse sample, with more than ten times as many subjects as the original. Second, no one had followed up on the study’s participants in more than 20 years. 

Watts’s team found that, like Mischel, a child’s ability to delay gratification seemed to correlate with adolescent success— things like higher test scores. However, as Mischel predicted, the correlation was much smaller than the initial experiment found. Watts also noticed that the correlation weakened as subjects got older. Success on the marshmallow test could predict favorable outcomes until participants were about 20-years old, but after that, the connection vanished.

More important, though, was the NYU team’s other discovery. When controlling for factors like economic background, the correlation completely disappeared. In Watts’s words, “if you have two kids [with] the same background, the same kind of parenting, same ethnicity, same gender, similar home environment, similar early cognitive ability, [and] one of them is able to delay gratification, and the other one isn’t, does that matter? Our study says…probably not.”

In other words, a child’s ability to delay gratification was a decent indicator that they would find success as an adolescent. However, that success stemmed much more from their socioeconomic background than their willpower or self-control. This example is clearly illustrated in comparisons between the results of children from high-income and low-income households. 

Children of upper-class parents receive a better education. They are more likely to have two parents at home. Those parents are, by definition, not living paycheck to paycheck, allowing them to think and plan for the future and worry less about the present. The child won’t experience excessive hunger and will trust adults to consistently put plenty of food on the table. If they accidentally miss a meal, these kids know they will have plenty of other opportunities to eat. In other words, children in high-income households will likely see their parents delay gratification and learn that habit themselves. Critically, those same children are also much more likely to be enrolled in an expensive SAT prep course— paid for by the parents. 

On the other hand, children of lower-income families learn to value the present above all. Researchers from the University of Oregon showed that impoverished people put little faith in the future, as it’s much less certain than for people with higher incomes. For families who struggle to put food on the table, children learn to take what they can get when they can get it. As the saying goes, a bird (marshmallow?) in the hand is worth two in the bush.

Of course, other factors from the children’s backgrounds influenced success on the marshmallow test. The subjects’ results were highly affected by their general trust in adults or authority figures. That trust could even be impacted by the experimenter’s identity. One example involved a marshmallow experiment in Trinidad. Lower-class black residents there performed poorly when the test was administered by white people, who had a history of breaking their promises on the island.

Altogether, Watts’s research disproved the public belief in the immense power of delayed gratification. It wasn’t that Mischel’s study was incorrect, but the way that it was interpreted by the public was blatantly false. The new conclusion was simple: willpower and delayed gratification aren’t a one-way ticket to success. They are essential traits, but working to instill them in your children will have much less impact than simply raising them in a stable environment and educating them well. 

Still, follow-up research into self-control shows that our understanding of the elusive trait is similarly flawed. Most people see self-control as one’s ability to control impulses, but that isn’t the case. Rather, people who appear to have strong willpower are better at avoiding temptations, not resisting them. Like the marshmallow test showed, self-control is largely a result of one’s environment— how visible are the temptations and how intriguing are the alternatives.

For instance, imagine you determined to abstain from drinking alcohol for one month, but that all your friends love to go out to the bar for a drink on the weekends. Traditionally, we would view willpower as going to the bar but not ordering a drink. In reality, the people who best delay gratification are those who don’t go to the bar in the first place. 

More than ever, the globalized world is proving to be a place filled with temptations. Our phones are designed to entice us with the things we most want. Processed foods line the counters of the grocery store. Avoiding temptations has become nearly impossible.

But hope is not lost. Despite its flaws, Mischel’s earliest experiments reveal essential strategies for resisting temptations and delaying gratification. Distract yourself with the things you enjoy. Avoid environments filled with proverbial marshmallows. Learn to find the fun in the process, like by gamifying your studies or working out with a friend. 

Unsurprisingly, the media and society as a whole have been much slower to latch onto this more grounded view of self-control. Pop psychology has long produced the impression that there are quick, easy answers to complex problems. The marshmallow test is just one of many examples. In reality, the solutions are as complicated as the questions. 

But what do you think? What is the value of the marshmallow test? What other lessons can we draw from this experiment that has, for so long, been misinterpreted? How do you define willpower or self-control? Is delayed gratification still among the most essential traits, or is there something else that you believe is the ultimate determinant of success?

Related Articles


Please enter your comment!
Please enter your name here

Stay Connected


Random article


Who Was Really the Worst Dictator of the 20th Century?

It was perhaps the golden age of dictators.  During the 20th Century, regime after regime rose up across the planet, headed by some of the...

Latest Articles