The Marshmallow Test — what the world’s most famous experiment on self-control actually showed

The marshmallow test is an experiment in which young children were given the choice between one sweet straight away or two if they could wait a little while on their own. It became world-famous when follow-up studies showed that children who waited longer went on to do better at school, and this was interpreted as meaning that willpower at an early age determines one’s future. Later and more extensive research has significantly undermined that interpretation. What remains valid is something more useful: the children who managed to wait did so not by gritting their teeth, but by directing their attention.

The experiment

In the 1960s and 70s, psychologist Walter Mischel and his colleagues conducted a series of experiments at Stanford. Pre-school-aged children were asked to sit in a room with a reward in front of them — sometimes a marshmallow, sometimes a biscuit or a pretzel. If they waited until the researcher returned, they would receive two. If they rang a bell, the researcher would come straight away, but then they would only get one.

What Mischel was most interested in was not which children waited, but how they went about it.

The children who looked away

The children who managed to wait the longest had often found ways not to think about the reward. They covered their eyes, turned away, sang, talked to themselves or made up games. When the researchers hid the reward, the children waited longer. When they encouraged the children to think of the sweet as something else — for example, a picture of a cloud — it also became easier.

This is the part of the marshmallow test that is most often overlooked, and the most useful. Self-control here was less about resisting a temptation right in front of you, and more about ensuring it did not take hold of your attention.

The famous follow-up

Many years later, Mischel and his colleagues followed up with some of the children. Those who had waited longer as young children scored, on average, better in university entrance exams and were described by their parents as better at coping with frustration. It was here that the interpretation emerged: that the ability to wait at the age of four predicts future success in life.

The problem was the sample. The children came largely from the same nursery at Stanford, with parents who were often academics, and the number included in the follow-up was quite small.

What a larger-scale replication showed

Tyler Watts, Greg Duncan and Haonan Quan replicated the study using a significantly larger and more diverse sample of children. They also found a link between waiting a long time and later academic performance — but it was considerably weaker than in the original study. And when they took the children’s family background into account, including their parents’ education and home environment, the link diminished further.

This suggests that much of what the marshmallow test measured was not an innate trait of the child, but rather traces of the environment in which the child had grown up.

Trust determines whether waiting is wise

Celeste Kidd, Holly Palmeri and Richard Aslin conducted a follow-up study that turned the perspective on its head. Before the test, the children were shown an adult who either kept or broke a promise — for example, to come back with better crayons. The children who had just experienced an unreliable adult waited for much shorter periods than those who had experienced a reliable one.

This isn’t a lack of self-control. It’s rational. If you can’t trust that the second piece of sweet is coming, it’s wise to take the first one. A child who has grown up in an environment where promises are often broken has learnt something true about the world. Why this means that much of what we call willpower is actually down to circumstances is explained in the article on self-control or willpower.

What an adult with a phone can take away from this

The marshmallow test does not show that your fate was decided in nursery. If you feel you have poor impulse control as an adult, this is not a definitive statement about who you are.

What that test does show, however, is which strategies work:

Do as the children do: look away. A mobile phone lying in front of you is like a marshmallow on the table. Put it in another room, in a bag, out of sight. You don’t have to resist what you can’t see.

Change how you view the temptation. The children who thought of the sweet as a cloud waited longer. Seeing a feed as a product designed to keep you hooked, rather than as something you’re missing out on, can have the same effect.

Create a reliable reward. If waiting never pays off, you’ll stop waiting. Make sure that what you give yourself instead of your phone actually happens.

You can find out how to train this over time in the article on practising impulse control.

---

The Zenith Protocol is a 30-day programme for impulse control based on what the marshmallow test actually demonstrated: that it’s about managing your attention and environment, not just gritting your teeth. It’s free to get started.