Marshmallows Revisited, Again

Marshmallows Revisited, Again Photo Credit: Clipart.com

Children in the classic “Stanford Marshmallow Study” may have been more strategic than we thought.


A new take on a landmark study. I’m Bob Hirshon and this is Science Update.

The “Stanford Marshmallow Experiment” was a classic study of delayed gratification: Preschoolers tried to resist eating one marshmallow, in order to earn two later. The children’s success was largely attributed to impulse control. But University of Rochester cognitive scientist Celeste Kidd wasn’t so sure.

It could’ve been the case that kids were simply incapable of waiting. Alternatively, young children could be capable of making a decision about whether waiting or not would be worth it.

To find out, she and her colleagues repeated the marshmallow experiment. But first, the kids got to color, while a researcher either delivered or reneged on a promise to bring them nicer crayons. When the adult kept his or her word, the child was much more patient in the subsequent marshmallow test. The results suggest that the reliability of the adults around them can strongly influence children’s self-control. I’m Bob Hirshon for AAAS, the Science Society.

Making Sense of the Research

First, a little more on the original Stanford Marshmallow Experiment. In 1972, Stanford psychologist Walter Mischel ran the first set of experiments with four- to six-year-old children at Bing Nursery School. Each child was brought into a room free of distractions and offered a marshmallow (or, if they preferred, a cookie or a pretzel stick). They were told that they could eat the marshmallow right away, but if they waited fifteen minutes, they would get a second marshmallow. 

Being able to postpone a reward like that is called delaying gratification, and it's considered an important part of emotional and social maturity. (Imagine if adults all went around doing exactly what they felt like doing at all times!) In the original study, older children were generally more likely to delay the marshmallow reward, but some children were better at it than others. 

What's fascinating is that when those same children were tested years later, first as teenagers and later as middle-aged adults, the differences in ability to delay gratification remained very stable. In other words, kids who could wait for a second marshmallow grew up to be teens and adults who showed other age-appropriate hallmarks of delaying gratification. As teenagers, for example, the successful delayers had higher SAT scores than the non-delayers, and were generally rated by their parents as more mature and competent.

These findings have often been used to support the hypothesis that our ability to delay gratification is part of our innate personality, and may be hard to change. But Kidd's new experiment suggests that it may not be so rigid after all. In her study, children performed the same waiting-for-a-marshmallow test that the kids did back in 1972. But beforehand, an adult gave them some crayons to color with. The crayons were in bad shape: old, broken in half, worn down, or papers peeled off. 

After a few minutes, the adult in the room left and promised to come back with better crayons. When the adult returned, sometimes he/she had the crayons and sometimes he/she didn't. In the reliable-adult condition (where the adult kept her promise), children performed better on the marshmallow test than in the unreliable-adult condition. That suggests that the children used the adult's trustworthiness about the crayons to judge whether or not it was worth waiting for another marshmallow. 

Of course, in the original experiment, there was no such task beforehand, and the adult experimenters would have tried to treat all the children the same. So the new study doesn't contradict the idea that people have innate differences in the ability to delay gratification. It does suggest, though, that this ability can be adjusted quickly to suit the environment, and that chronic exposure to unreliable, unstable adults may deal a serious blow to a child's sense of self-control.

Now try and answer these questions:

  1. What was the original Stanford Marshmallow Experiment?
  2. What were the follow-up experiments since then, aside from this one? Why did they suggest that self-control is part of our innate personality?
  3. What does this study add to our understanding of the original experiment?
  4. Can you think of situations in which children are exposed to “unreliable adults” in real life? Do you agree that this can affect self-control in everyday situations? Why or why not? 

You may want to check out the related podcast, Kids and Marshmallows Revisited, which is an earlier follow-up on this study.

Going Further

For Educators

Check out the related series of lessons:

Related Resources

Pennyless Pricing
6-12 | Audio
Antibiotics and Asthma
6-12 | Audio
Robotic Dog
6-12 | Audio

Did you find this resource helpful?

Science Update Details

Grades Themes Project 2061 Benchmarks