The Marshmallow Test (2014) is a psychology book detailing the story of the famous, same-name experiment on self-control, delayed gratification, and their link to personal success.
TPM Takeaway: The time horizon impact of life strategies
Human decision-making is a conflict and tradeoff between short-term gratification impulses and long time-horizon. However, it’s not true that waiting is superior and it depends on context.
Waiting within a long-term horizon works when the odds of larger future rewards are high, but immediate consumption works when rewards may disappear. In general, for most men in civil societies, extending time horizons works, but not always.
Overview
- Children who were able to resist the urge of eating the treat showed higher concentration and scored higher on SATs
- Self-control is genes, nurture, and also the context
- We can increase our self-control with techniques and mind re-framing
Marshmallow Theory
The Marshmallow Theory refers to the idea that delayed gratification—the ability to wait for a larger future reward instead of taking an immediate smaller one—is a key predictor of long-term success.
It originates from Walter Mischel’s “marshmallow test”, where children were given a choice:
One marshmallow now, or two marshmallows if they waited 15 minutes.
The follow-up studies revealed a surprising pattern:
Children who were able to wait tended to do better later in life—academically, socially, and even in terms of health and career success.
What the Theory Suggests
Self-control is foundational: The ability to delay gratification is linked to executive function and emotional regulation.
Environment matters: Later studies showed that children’s trust in the reliability of adults (and their environment) also plays a big role in their decision to wait or not.
It’s not just about willpower: Cognitive strategies (like distraction or reframing) can improve impulse control—skills that can be taught and trained.
Related Concepts:
Impulse control
Executive function
Future orientation
Trust in authority/environment
Is It True?
The marshmallow theory is valid, and the ability to delay gratification is an important trait for long-term success in many environments, but not all. And it may not be as important as it was originally thought.
Read our criticism section for more.
Notes
The Marshmallow Test details the famous experiment involving children’s capacity to resist temptation. It then expands on the importance of delaying gratification and how we can improve our emotional intelligence to delay gratification.
How the Test Worked
This is how the marshmallow test worked:
The children would first pick their favorite treat. The researcher then would tell them they would leave the room and they could either have the treat now or enjoy it later once they’d come back.
The researcher also told the children they could call them back if they couldn’t wait any longer.
And then, they would leave them alone and observe.
Later in Life
Children who resisted for longer were better at concentrating as adults, scored higher on their SATs, and were better at maintaining personal relationships.
The children who were able to resist showed more activation in their prefrontal cortex, while those who ate the marshmallow right away showed more activation in their ventral striatum, which is linked with pleasure and addiction.
Self-Control vs. Urges
We have two systems in our body and mind: one that instantly reacts to the environment -and wants to eat-. And one that controls and, if necessary, restrains our behavior.
The author calls the former system “hot system” and the delaying strategies the “cold system”.
When we engage in activities that distract us from the stimuli, we are engaging in the cooling system.
The hot system is ready to grow from birth, but the cooling system is only developed during childhood. And this is one of the reasons why very small children find it harder to resist: they haven’t developed cooling and coping mechanisms yet.
Children under four indeed usually don’t have any cooling systems. But the cooling system can keep developing until we’re adults.
Self-Control is Genes, Nurture, & Context
As for many other personal traits, self-control is partly genetic and partly environmental.
For example, the authors say that a parent can teach children distracting techniques by handing them a toy when he’s upset.
But self-control is also environmental. For example, sexual restraint doesn’t necessarily mean that the same lack of control will be exhibited in different environments and situations.
However, here is the most interesting thing for me: if you think self-control is limitless you’ll be able to control yourself much better.
This is another great example of how our own beliefs shape our reality.
TPM Note: Unstable environments prime for immediate gratification
In some contexts immediate gratification works better. This is also related to fast life theory, also linked to psychopathy. Children in unstable and risky environments are primed for faster life and immediate gratification to secure early reproduction.
Teaching Children Self-Control
The authors have a few tips for teaching children self-control:
- Reward effort (not outcome, as also recommended for a growth mindset)
- Be consistent
- Teach that actions have consequences (good actions lead to good results and vice versa)
- Be a good role model
- Teach them to always give their best
- Teach “what if”: “if my hand moves towards the sugary food, then do X to distract yourself”
Personally, you can also:
- Focus on the long-term meaning and consequences
- Distance yourself from the source of temptation
Does the Theory Hold?
As it’s been the case for many classic psychology experiments, the marshmallow test has received criticism and faced some replication issues.
The researchers were clear that a bigger sample size could reduce the link between the delay of gratification and overall success in life.
And they were also clearly not advocating any policy changes because being able to delay gratification in children wouldn’t necessarily mean they would be more successful in life.
Yet the experiment became so popular that, as is sometimes the case, it has taken on a whole different significance. Namely, that children who could control themselves were going to do great things in life and that children who could not control themselves were doomed.
Of course, the reality is often slightly more complex than that.
Indeed, the original marshmallow experiment has several drawbacks, including:
- Some children might obviously be less interested in sugary treats than others
- Children in the tests were few
- Children in the test were all from Stanford University: either children of professors or students
- Poorer children might have a predisposition and drive to consume faster
Indeed, when the study was replicated, it was found that:
- Success of the marshmallow test at age 4 predicted success at age 14 with a correlation that was 50% smaller than the original test
- Correlation almost vanished when taking into account intelligence and family background
Addendum: Replication with broader samples shows weaker effects
In short:
Watts, Duncan & Quan (2018) conducted a high-powered replication using a much larger and more diverse sample (from the NICHD study).T
They did find a link, but the effect became much smaller after controlling for:
Socioeconomic status (SES)
Parental education
Home environment quality
Tyler Watts, who replicated the test, sums it up as follows:
Take two kids with same background, parenting, ethnicity, gender and cognitive ability and their ability to delay the marshmallow gratification matters little.
That means that the ability to delay gratification is not a unique lever but a consequence of a much bigger picture. And the second study shows that teaching children how to delay gratification might not have such a major impact on their lives.
Conclusion: The ability to delay gratification is not uniquely predictive of success—it’s one factor among many, and heavily influenced by environment.
Trust and context matter more than we thought
Kids from less stable environments may eat the marshmallow early not due to poor self-control, but because:
- They’ve learned that delayed rewards often never arrive.
- They’re rational actors in an untrustworthy environment.
This makes sense from an evolutionary psychology perspective as recent theories have also suggested.
Conclusion: The marshmallow test is as much about environmental trust as about individual willpower.
Cultural variation
In some cultures, group conformity or respect for adults may alter how children respond to authority and delayed rewards.
Delayed gratification may be less predictive in non-Western societies.
Bottom Line: What the Science Now Suggests
The marshmallow theory is partly true: The ability to delay gratification can be helpful when rewards are certain, and it reflects executive functioning that are often important for long-term success.
But:
- It’s not destiny
- Outcomes are mediated by socioeconomic context, trust in adults, and learned behavior
- Training, context, and support matter more than just willpower
You can read more here:
Review
The Marshmallow Test is a good book on a famous experiment and later theory that showed that impulse control supports delayed gratification and is an important element of long-term thinking and long-term success.
Like many other classical experiments, the marshmallow test has been called into question by the replication crisis.
While some of the criticism is valid, we must be cautious not to throw away the baby with the bathwater. The ability to resist urges and temptations is a crucial factor for one’s level of long-term success and fulfillment.
The bottom line is this: long-term success often requires some short-term pain. And that’s the message we agree with: for a better tomorrow, learn to bite the bullet today.
However, for long-term success, you also need more. See, for example:
