Delayed Gratification: What the Marshmallow Story Got Wrong, and What Still Holds

The story usually goes like this. A four-year-old sits alone at a table with a marshmallow. If she can wait fifteen minutes, she gets two. Some children wait, some don't, and the ones who wait grow up to do better in school, cope better with stress, and generally win at life. The moral writes itself: the capacity to delay gratification is the secret of success, and you either have it at four or you don't. Most adults hear the story and quietly sort themselves. I'd have eaten it. That explains a lot.

It is a good story, which is the problem. The marshmallow test became a parable about character, but the best replication found most of the famous link to later success fades once you account for family background, and an earlier study showed children wait far longer when they trust the adult; what still holds is practical, not moral: the strategies that helped four-year-olds work for adults too, and the modern instant-gratification problem is mostly engineered. Here is what the original studies found, what the replication did to the parable, and how to delay gratification when the thing in front of you is a phone.

What is delayed gratification?

Delayed gratification is choosing a larger or better reward later over a smaller one now. In the lab it is measured in seconds: how long a child will wait for the preferred treat before settling for the lesser one available now.

The original research was never about character. Walter Mischel's 1972 study with Ebbe Ebbesen and Antonette Zeiss was titled "Cognitive and attentional mechanisms in delay of gratification," and that is what it tested. Preschoolers could take a less-preferred reward immediately or wait for a better one, and the experimenters varied what the children were thinking about and looking at. Children distracted from the rewards waited far longer than children who attended to them. Thinking about something fun worked as a distraction; thinking sad thoughts did not, and thinking about the treats themselves made waiting harder, even when the treats were not in the room.

In other words, the original finding was that attention and framing decide how long a child can wait. Not grit. That is the part of the research that survived everything that came after, and the part almost nobody talks about.

What did the marshmallow test find about later life?

A follow-up found that longer waits in preschool went with better outcomes years later, under specific conditions. In 1989, Mischel, Yuichi Shoda and Monica Rodriguez reported in Science that four-year-olds who waited longer in certain lab situations became adolescents who were more cognitively and socially competent, did better in school, and coped better with frustration and stress. A 1990 paper by Shoda, Mischel and Philip Peake in Developmental Psychology refined this: the correlations showed up only in particular versions of the task, the ones the authors argued would reflect a child's own cognitive and attentional skills rather than the setup.

This is where the parable began. The findings were real, but they were correlations from children in a single university community, and the authors framed them in terms of cognitive and attentional competence. The culture took them as a verdict on willpower, which, as our post on whether willpower is a myth explains, has had a rough decade of its own.

What did the 2018 replication find?

That the famous link mostly disappears once you account for who the child's parents are. In 2018, Tyler Watts, Greg Duncan and Haonan Quan published a conceptual replication in Psychological Science using 918 children from a national study. Their focal analysis looked at 552 children whose mothers had not finished college.

Without controls, delay time at age four predicted achievement at age fifteen, but the association was about half the size of the original. Once the researchers added family background, early cognitive ability and home environment, it fell by about two thirds, to a coefficient of 0.08. Adding measures of the child's cognition and behavior at the same age as the test brought it to 0.05, not distinguishable from zero. Most of the remaining association came from whether a child could wait at least twenty seconds; beyond that, longer waits made no measurable difference.

The honest reading is not that the marshmallow test was wrong: delay time does track something. But much of what it tracks is the child's environment: a stable home, educated parents, an early sense that promised things arrive. The marshmallow was a thermometer, and the parable mistook it for the weather.

Does trust change whether children wait?

Yes, and by a lot. In 2013, Celeste Kidd, Holly Palmeri and Richard Aslin published a study in Cognition with the pointed title "Rational snacking." Before the marshmallow task, each child did an art project with an adult. For half the children, the adult promised better supplies, left, and came back with them. For the other half, the adult came back empty-handed and apologized.

Children who had met the reliable adult waited an average of twelve minutes, and nine of the fourteen held out the full fifteen. Children who had met the unreliable adult waited an average of three minutes, and one of fourteen made it. The only difference was whether the second marshmallow seemed likely to exist. The sample was small, 28 children in all, but the gap was large, and the point stands: eating the marshmallow in an unreliable world is not a failure of self-control but the correct decision.

The adult translation is uncomfortable. If your history says the second marshmallow doesn't come (the promotion that went elsewhere, the diet that cost a month and gave back nothing, the promise you made yourself and broke), then grabbing what is in front of you is rational, and shame will not change it. Evidence will: small waits that pay off, reliably, on a timescale you can see. It is slower than a speech, and the only fix that works on what is actually driving the behavior. When the payoff you are waiting for is years away, our post on linking short-term and long-term goals covers how nearer goals can carry it.

Why is instant gratification so hard to resist now?

Because the marshmallow sat still and the phone doesn't. The parable blames the person. The better explanation for most adult struggles with instant gratification is design.

A marshmallow is a single, fixed reward. A feed refreshes, a notification arrives, a video is followed by a slightly better video, and the reward is not just immediate but unpredictable, which is the pattern that keeps people pulling at any lever. This is not a claim that your phone has rewired your brain; our honest review of the dopamine detox explains why that language outruns the evidence. The four-year-olds in 1972 waited longer when the treat was out of sight; most adults now carry the treat in their pocket, and it taps them on the shoulder. Our guide to quitting doomscrolling treats that as the engineering problem it is. The question is not "why am I so weak" but "why is the thing I want now so close," and distances can be changed.

How to delay gratification as an adult

Use the strategies that worked on four-year-olds, because they were never about age. Then add the two tools adults have that children don't: precommitment and a track record.

1. Put the marshmallow out of sight. Distraction from the reward was the strongest effect in the original research. The phone charges in another room. The app is deleted on weekdays, not muted. The card is not saved in the browser. You are not resisting; you are making resisting unnecessary.
2. Change what you are looking at. Mischel's children waited longer when they thought about something else. The adult version is naming the cost at the moment of wanting: not "I shouldn't" but "this is forty minutes I'll feel bad about." Then look at something else, deliberately, and make it something you enjoy; grim distraction didn't work for the children either.
3. Bundle the temptation with the task. In a 2014 field study, Katherine Milkman, Julia Minson and Kevin Volpp gave people tempting audiobooks they could only listen to at the gym. Early on, that group went to the gym 51 percent more often than controls. The effect faded after Thanksgiving, but 61 percent of participants then chose to pay for a gym-only device loaded with audiobooks. Pair the thing you want now with the thing you want later, and only allow the pairing.
4. Make the second marshmallow visible and short. If your past says waiting doesn't pay, run small experiments where it does. A week, not a year. Something you can measure. Our post on doing hard things is about how to pick the first one.
5. Precommit while you are calm. Decide the rule before the wanting starts: the thirty-minute wait before any purchase over a set amount, the automatic transfer on payday, the no-ordering-in rule on weeknights. A decision made at 2pm is easier to keep than one made at 9pm.
6. One wait at a time. The people who seem to delay gratification across their whole lives are not running twelve simultaneous acts of restraint. They have one or two that became habits and a lot of environment doing the rest. Holding high standards in one place beats enforcing medium ones everywhere.

Ordinary examples: cooking tonight instead of ordering, writing the page before opening the feed, leaving the shoes in the cart for three days, doing the dull part of the project first. None of these are heroic, which is the point. If one keeps failing in the same place, our page on quitting a bad habit walks through the situational approach.

A version of self-improvement treats every small wait as a referendum on your character. That is the parable talking, and it was wrong about the children, so it is probably wrong about you. Becoming a better person without toxic positivity means getting the distances right, keeping your promises to yourself small enough to keep, and letting evidence do what shame never could.

Frequently asked questions

What are some examples of delayed gratification? Saving part of a paycheck, cooking rather than ordering in, finishing the dull half of a task first, waiting a few days before a purchase. The useful ones are small and specific: "being more disciplined" is not an example; "the phone charges in the kitchen" is.

What are the benefits of delayed gratification? The honest answer is narrower than the parable. The trait correlates with later outcomes, but the replication suggests much of that reflects family background. The benefits of specific delays are concrete: money that is still there, work that is finished, mornings that are not wrecked.

Is instant gratification always bad? No. Taking the reward now is sometimes the right call, especially when the later reward is uncertain, as the children in the trust study understood. The problem is not wanting things now; it is an environment built to serve that want constantly, and a self-story that reads every lapse as proof of weakness.

Why can't I delay gratification no matter how hard I try? Usually because the fix you are trying is willpower in the moment, the weakest tool available. Move the temptation, decide in advance, pair the want with the task, and build a short track record of waits that paid. If the pattern runs across your whole life with restlessness, low mood or trouble finishing anything, talk to a professional: ADHD and depression both affect this and neither is a character flaw. If you are in crisis, call or text 988 in the US, 9-8-8 in Canada, or find your country's line at findahelpline.com.

Sources and further reading

• Mischel W, Ebbesen EB, Zeiss AR (1972). "Cognitive and attentional mechanisms in delay of gratification." Journal of Personality and Social Psychology
• Mischel W, Shoda Y, Rodriguez ML (1989). "Delay of gratification in children." Science
• Shoda Y, Mischel W, Peake PK (1990). "Predicting adolescent cognitive and self-regulatory competencies from preschool delay of gratification: Identifying diagnostic conditions." Developmental Psychology
• Watts TW, Duncan GJ, Quan H (2018). "Revisiting the Marshmallow Test: A Conceptual Replication Investigating Links Between Early Delay of Gratification and Later Outcomes." Psychological Science
• Kidd C, Palmeri H, Aslin RN (2013). "Rational snacking: Young children's decision-making on the marshmallow task is moderated by beliefs about environmental reliability." Cognition
• Milkman KL, Minson JA, Volpp KGM (2014). "Holding the Hunger Games Hostage at the Gym: An Evaluation of Temptation Bundling." Management Science

ILTY won't grade your character by what you did at 10pm. Mr. Relentless will ask what was within reach and why, and the Stoic Advisor will help you tell a reward worth waiting for from one you only think you want. Each conversation ends with action items, usually a distance to change, and because ILTY remembers, it will ask next time whether the second marshmallow showed up. The optional mood check-in before and after a conversation makes the wanting visible instead of automatic. It isn't therapy, and it isn't for crisis. It's a place to build the track record that shame never could. Try it free.