Let Them Eat Marshmallows : The Indicator from Planet Money The marshmallow test is one of the most famous social experiments of all time, but we may be thinking about it all wrong.

Let Them Eat Marshmallows

  • Download
  • <iframe src="https://www.npr.org/player/embed/618954014/618956637" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript


There is a test that you can give a kid, a really little kid, like, 4 years old. And if the kid passes the test, he or she has a much higher chance of doing well in school, of staying out of trouble, of just having more success in life.


It's called the marshmallow test. And it's one of the most famous experiments ever in any field in part because it's so simple.


UNIDENTIFIED PERSON #1: OK, so you guys each get one marshmallow. And if...

GARCIA: Put a marshmallow or some other treat in front of a kid. You tell the kid that you're going to leave the room for five minutes.


UNIDENTIFIED PERSON #1: And if when we come back your marshmallow is still there, then you get to have two.



VANEK SMITH: By the time the researcher gets back or the parent, the kid who has not eaten the marshmallow will get a second marshmallow. Now, this is really hard for a kid, right? They know they will get more candy if they wait. But that is a concept, and there is this delicious piece of candy sitting right in front of them right there.


UNIDENTIFIED CHILD #2: I won't eat my marshmallow.

GARCIA: I believe him.

VANEK SMITH: (Laughter).

GARCIA: This test was originally developed by a man named Walter Mischel back in the 1960s. But YouTube is just filled with kids all over the world taking the marshmallow test. And what the kids try to do to pass the time and not eat the marshmallow is amazing and hilarious and represents all that is good in life.

VANEK SMITH: (Laughter).


UNIDENTIFIED CHILD #2: Daddy, can I eat it now? Daddy? (Singing, unintelligible).

GARCIA: Incidentally, that's also how I procrastinate from writing scripts for the show.

VANEK SMITH: Singing the "Jeopardy!" song?

GARCIA: Oh, yeah.

VANEK SMITH: (Laughter). I do have a favorite among all the videos that I watched. And it's one where the dad gives his sweet little boy a Rolo. It's a Rolo test, in this case. The kid is staring at this Rolo. He sings. He talks to himself. And then at one point he just starts sobbing.


UNIDENTIFIED CHILD #3: I want to eat it now (sobbing).

VANEK SMITH: It's terrible (laughter). This poor kid. But he did not eat the Rolo, which means, apparently, that he is bound for great things because researchers followed many of the marshmallow test-takers for years after the initial test in the '60s, and the results were just stunning. At least that's what everybody thought. This is THE INDICATOR. I'm Stacey Vanek Smith.

GARCIA: And I'm Cardiff Garcia. Today on the show, the marshmallow test. One of the most famous and apparently predictive experiments in history has just been rerun by a new group of experimenters. And we would tell you what they found, but we're going to actually make you wait. Trust us; it will be worth it.


UNIDENTIFIED PERSON #2: Oh, my gosh. Guess what? You won the game. You didn't eat them.


VANEK SMITH: Oh, Cardiff, before we go any further, I have something for you.

GARCIA: Ooh (ph).

VANEK SMITH: It's a chocolate bar.

GARCIA: Yeah, but it's not just any chocolate bar.


GARCIA: This is my favorite. This is Lily's...


GARCIA: ...Forty percent creamy milk chocolate bar.

VANEK SMITH: I know. Well, I just thought, you know, since we're doing the marshmallow test, we should just open it. There it is.

GARCIA: Have some?

VANEK SMITH: Well, no, we're supposed to wait till the end of the podcast. That's the whole point, right?

GARCIA: I have to stare at this until we finish the podcast?

VANEK SMITH: I mean, this poor 4-year-old had to do it for five minutes.

GARCIA: I don't like this.

VANEK SMITH: Tyler Watts is with NYU's applied developmental psych department. He and a group of colleagues just published a paper about the famous marshmallow test where they looked at a lot of the results that the original test done in the '60s had found.

TYLER WATTS: The kids who waited longer initially at age 4 had higher SAT scores - much higher SAT scores. They found that the kids were sort of better socially adjusted and they did their homework. And they've even found that they have lower body BMI as adults. So they're sort of...

VANEK SMITH: They're more fit?

WATTS: They're more fit. There's indications that the kids that waited longer are healthier adults.

GARCIA: And in a way, this makes some intuitive sense, right? I mean, this is, like, an elegant, simple way to test a person's ability to delay gratification. And delaying gratification we know leads often to success in life. So like, studying for a test instead of playing video games or going to the gym or finishing a podcast instead of eating chocolate bars.

VANEK SMITH: (Laughter) It smells good.

GARCIA: You get the point.

VANEK SMITH: I can smell it. It smells good.

GARCIA: Very tempting.

VANEK SMITH: But it's also sort of chilling, right? It makes it seem like this ability to delay gratification is just baked in when you're so tiny, and there's nothing you can do about it. And it kind of maps out your whole life. That doesn't feel great. And policymakers and educators have drawn all kinds of conclusions from the marshmallow test over the years.

WATTS: There's a charter school in Houston that actually put on their website - and I'm going to mess up the quote, but it's something like, if you can only teach your kid one thing, teach them delay of gratification, right? And that's the point that we were really after with our study. Is that actually right?

VANEK SMITH: Tyler and his colleagues wanted to take a second look at the test. Walter Mischel's data from the original marshmallow test in the '60s had come from a group of 4-year-olds at Stanford University's preschool. So it was a pretty particular group of kids. Tyler and the team he worked with wanted a broader sample.

GARCIA: And fortunately, experimenters have been giving this test many, many times through the years. And so Tyler and his team used a different set of data taken from a test given in 1991. And this test was given to nearly a thousand kids from all over the country. These kids had really diverse socioeconomic backgrounds and were much more racially and ethnically diverse than the original group. And here's what they found.

WATTS: And this was kind of a basic finding of the paper - kids from higher socioeconomic status families waited longer on the test.

GARCIA: Specifically, kids from higher socioeconomic backgrounds had a 70 percent success rate on the marshmallow test. In other words, they waited for that second marshmallow. Kids from lower socioeconomic backgrounds only had a 45 percent success rate in turning it down.

VANEK SMITH: This seems really bleak, right? I mean, it just sort of looks like success in life is just about how much money your family has and how much education they have. And like, if you don't come from money or some kind of status group, your chances of success in life are just way lower even starting at age 4. So Tyler's team took these results, and they went a step further. So instead of just comparing all of the kids, they took kids from similar backgrounds and just compared their marshmallow test results.

WATTS: Let's take two kids who basically are the same in terms of ethnicity, gender, parenting environment, early cognitive ability. And let's take these two kids. And if one of them is able to delay gratification a little bit longer than the other, does that matter for later life outcomes? And what our results suggest is not very much.

GARCIA: In other words, take two kids from similar backgrounds, and one passes the marshmallow test and one doesn't. Their future success in life doesn't seem to be any different. So Tyler and his team started to think that maybe the marshmallow test was more reflecting something rather than predicting something.

WATTS: You have - you view that kids who wait longer or have better delay of gratification early on have many more markers of success later on. But is it the delay of gratification that's driving that, or is delay of gratification sort of symptomatic of, right? Delay of gratification is kind of a downstream effect of like you said, you know, socioeconomic factors, maybe sort of cognitive factors, other behavioral factors, parenting factors - all the things that we controlled for. And those were the things that were also really probably driving the markers of success later on in life.

VANEK SMITH: So the takeaway for Tyler and his team is the marshmallow test is clearly pointing at something, but it might just be pointing out what an amazing advantage kids have when they come from a wealthier background, a privileged background.

GARCIA: Or it could be pointing to something else.

WATTS: But that doesn't at all mean that delay of gratification isn't an important life skill that we should continue trying to understand and trying to figure out. It's just probably not something that we need to be overly concerned with teaching 4-year-olds.

GARCIA: Tyler says almost certainly, though, what the marshmallow test does reveal is how important the early childhood environment is for a kid. And those factors start to show themselves really early on.

VANEK SMITH: I, though, had a little bit of a different takeaway. So I watched dozens and dozens of these videos of kids taking marshmallow tests. It's a little bit addictive after a while. And all the kids and all their crazy mental gymnastics, the trying to resist the marshmallow - and do you remember the little sweet kid who started crying...

GARCIA: Yes, of course.

VANEK SMITH: ...Tortured by the Rolo? So at the very end of that video, the most amazing thing happens. The kid is finally eating his Rolos. His face is all red. He's, like, chewing. His nose is running. He's not crying anymore, but he still seems pretty traumatized from the emotional toll of the test. His dad is comforting him.


UNIDENTIFIED PERSON #2: You did a really good job waiting, didn't you?


UNIDENTIFIED PERSON #2: Good job, buddy.

UNIDENTIFIED CHILD #3: I want to do it again.

GARCIA: (Laughter).

VANEK SMITH: He wants to do it again. He was sobbing for, like, two minutes. I really think this is the true reveal of the marshmallow test, right? It's, like, so much in life you want something so much and you strive for it, and you go through all these hoops to get it. And all you can think about is, like, doing the whole thing over again (laughter). It's the human condition. The marshmallow test is the human condition.

GARCIA: That was lovely, Stacey.

VANEK SMITH: Thank you.

GARCIA: Can I have some of this chocolate bar now?

VANEK SMITH: You can have the chocolate. I think we've succeeded.

GARCIA: Thank you. About time. Yes.

VANEK SMITH: How do you feel about sharing?

GARCIA: I love sharing.

VANEK SMITH: OK, excellent. OK.

GARCIA: Oh, that's fantastic.


Copyright © 2018 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.

NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.