Mar 8, 2009

Mischel’s Marshmallows

How are your New Year's resolutions holding out? This might at least help you feel better about them.

Psychologist Walter Mischel explains how one little test involving a marshmallow might tell you a frightening amount about what kind of person you are. And Radiolab favorite Jonah Lehrer helps us make sense of the results. This one's all about our will power (or lack thereof).


Correction: An earlier version of this piece incorrectly stated that the kids who performed better on the marshmallow test had higher GPAs in high school and went to better colleges. Those elements were not a part of Mischel’s original study. The audio has been adjusted to reflect this fact.

THE LAB sticker

Unlock member-only exclusives and support the show

Exclusive Podcast Extras
Entire Podcast Archive
Listen Ad-Free
Behind-the-Scenes Content
Video Extras
Original Music & Playlists

Speaker 1: (laughing)

Speaker 2: Alright, sh! Be quiet!

Speaker 1: You're-

Speaker 2: Listening-

Speaker 1: -to Radiolab.

Speaker 2: The podcast-

Speaker 1: From-

Speaker 2: New York Public Radio-

Speaker 1: -Public Radio, WNYC [crosstalk 00:00:09]

Speaker 2: -WNYC.

Speaker 1: -and NPR.

Jad Abumrad: Hey, I'm Jad Abumrad.

Robert Krulwich: And I'm Robert Krulwich.

Jad Abumrad: This is Radiolab the podcast, and let me just throw something out here to set up what is going to be an awesome podcast. New Year's resolutions, okay?

Robert Krulwich: Yes?

Jad Abumrad: Now I know that's a few months out of date, but that's exactly my point, by now most of us are the same old schmucks that we were last year. We've completely forgotten about all the promises we made to ourselves, we've slipped back into old patterns. 

How are you doing, out of curiosity?

Robert Krulwich: I didn't make any, so I didn't have any problem with them at all. But had I made  previous years I did...I would have long since failed. 

Jad Abumrad: Long since failed. Do you ever wonder why with things like New Years resolutions, which require willpower, why none of us are any good?

Robert Krulwich: Well, I'm not sure none of us. I know people, I'm married to a person who seems to be able to make a resolution and fiercely and determinedly stick to it.

Jad Abumrad: See, there you go. See, some of us can do it, some of us can't. 

Robert Krulwich: Because I was born under a dark star.

Jad Abumrad: (laughing) There you go, there you go. See, the real question is: are some of us just doomed because we're born under a dark star to not have willpower while others do have it? Or is it something that can be taught? Do you ever wonder about that?

Robert Krulwich: Yah.

Speaker 3: [crosstalk 00:01:23] Hello, test, test, test, test...

Walter Mischel: So you're monitoring the sound...

Speaker 3: Yah, as you can see, that's [inaudible 00:01:27]

Jad Abumrad: Well, I want to introduce you to somebody, his name's Walter.

Walter Mischel: Walter Mischel.

Jad Abumrad: He's a psychologist.

Walter Mischel: Columbia University in the Psychology Department.

Speaker 3: So, what's going to happen is...[inaudible 00:01:38]

Jad Abumrad: And if you ever asked yourself questions about willpower, like what is it exactly, how does it work? If I don't have it, can I get it? Well, he has done a remarkable study.

So are we ready to go? On the Marshmallow Study.

Walter Mischel: Ah, the Marshmallow Study.

Jad Abumrad: At the start of the study, how did you come up with the idea?

Walter Mischel: Well, I had three little girls, who at that time...

Jad Abumrad: We're talking, now, late 60s.

Walter Mischel: Were going from being these sweet sometimes, and yappy often, little creatures.

Jad Abumrad: How old were they at that point?

Walter Mischel: They ranged in age from 2 to 5. And what I was...what any parent immediately knows is that when kids start going into the fourth year of life a lot happens to them.

Jad Abumrad: Like he noticed that one of his daughters, who had just hit 4, suddenly she could delay gratification, almost all of a sudden. If he told her "Look, I know you want that...I don't know, pack of bubblegum, but don't whine, don't complain, when we get home I'll give you something better." Suddenly, she could wait.

Walter Mischel: This was in the late 1960s, and there was almost no work on quote "willpower" or self-control or self-regulation. It just wasn't a topic.

Jad Abumrad: So he thought he would jump in. And here's what he did: to test this hypothesis, something happens to kids around the age of 4, he went to a nursery school-

Walter Mischel: -a big nursery school at Stanford University.

Jad Abumrad: -gathered up a bunch of kids of different ages, and one by one he put them in a room.

Walter Mischel: A small room off the play room-

Jad Abumrad: He'd bring the kids in, sit them in a chair, and then he would give them a choice which had to do with marshmallows.

Walter Mischel: Yummy.

Jad Abumrad: What kid doesn't love marshmallows? And he would say to them, "Here's your choice, you can either have one-"

Walter Mischel: -one marshmallow-

Jad Abumrad: -now, or if you wait, you can have two-

Walter Mischel: -two marshmallows-

Jad Abumrad: Later."

Walter Mischel: Later. Now that creates a lot of dilemmas.

Jad Abumrad: Yah, it does. Especially because, you know, the researchers leave the room to give the kids a chance to think, and right there on the table, in front of the child, were the marshmallows calling them. 

Walter Mischel: There are no distractions in the room, there are no pictures, there are no toys, there's no anything.

Jad Abumrad: [crosstalk 00:03:46]The child is left to stare at the marshmallows...

Walter Mischel: So it's basically, it's an isolation chamber. 

Jad Abumrad: Pure agony.

Walter Mischel: I can show you video of what they're doing, and you can see it.

Jad Abumrad: Could we watch?

Walter Mischel: Sure, let's turn it on and...[inaudible 00:03:57]

Jad Abumrad: He walked over to a TV and pulled up a video of the Marshmallow Experiment, and I must tell you that it is some of the greatest video ever shot...particularly because the round he showed us, the researchers had replaced the marshmallows with something even more enticing.

Walter Mischel: Oreo cookies. 

Jad Abumrad: Mmmm, Oreos.

Walter Mischel: So, what you're seeing over here is a tortured looking, adorable little girl wearing a blue sweatshirt.

Jad Abumrad: About 5 years old, pigtails...

Walter Mischel: -Kind of sniffing at it.

Jad Abumrad: She puts her face right up to the cookie...she's weighing "Do I want this badly enough?"

Walter Mischel: That's exactly what she's doing.

Jad Abumrad: She picks up the cookie, puts it down...

Walter Mischel: Weighing and reevaluating her choice. 

Jad Abumrad: She knows that she can have this one now, but if she waits she can have more, but...[crosstalk 00:04:38] she gives in after just a few minutes.

Okay, now we have a sort of doughy faced boy in a yellow t-shirt who's kicking because he's so antsy. This kid's strategy is not to confront the cookie directly, but to kick the table that holds the cookie. Kick it, kick it, kick it!

Walter Mischel: And it's a very male response.

Jad Abumrad: But remarkably, he holds out way longer than the first girl. 

And of course there were the cheaters.

Walter Mischel: Here we see a guy looking at the two cookies, he's...

Jad Abumrad: He's picked one up...

Walter Mischel: ...Taken one, which is against the rules. He's licking the inside, he's licking out the cream, replacing the licked cookie and putting it back on the tray.

Jad Abumrad: Anyhow, we watch kid after kid after kid being tortured by the gravitational pull of Oreo cookies: some, to avoid the pull, went under the table, some turned their backs and started singing a song. All in all, Mischel and his team tested 500 kids in that initial study. And they found that yes, 4 year olds are dramatically better than younger kids at resisting temptation. 

So something does happen at around 4 years old. But within that...and here's where the plot thickens...there was a huge range. Some kids gave up after a minute. Others could last twenty minutes. And most fell somewhere in the middle, where they could resist the cookies for about-

Walter Mischel: Seven minutes. Eight minutes.

Jad Abumrad: That's the average?

Walter Mischel: All depending on age and what the goodies are and so on.

Jad Abumrad: Now, in and of itself, there's nothing too surprising here. You basically confirmed this hypothesis, but what is amazing, truly amazing, is what happens next.

Walter Mischel: I just was sitting around the kitchen table with my kids about 5-6 years after the studies began-

Jad Abumrad: He was talking with his kids, and he knew that they still went to school, this is now 5 or 6 years later, with some of those kids from the initial study, so he's asking them-

Walter Mischel: You know, " How is Jenny doing? And how's Cecily doing?" This was just totally informal.

Jad Abumrad: His kids told him, "Well, Cecily's doing fine. Jenny not so good." And he began to realize a weird pattern: the kids like Cecily who, he remembered, had been good at waiting for the marshmallow back in the study, they-

Walter Mischel: -Seemed to be the ones who were doing better.

Jad Abumrad: They were doing better at school.

Walter Mischel: And suddenly on a very small n...a very small sample, it looked like there were differences here. Could there be a relationship between the number of seconds these kids waited when they were 4 and how they're doing when they're 10 and 11 and 12? So, what we did beginning when these kids were 15 years old, 14 to 17 years old, was the first follow up wave. 

Jad Abumrad: So he tracked down as many of those original kids as he could find and just for starters, he looked at their SAT scores. Now, keep in mind, this is 10 years later, and some stupid little test that a kid takes when they're 4 that has to do with resisting a marshmallow and how many seconds they can resist, should not...I repeat, not...have anything to do with how they do on their SAT scores, which is one of the most important tests they're ever going to take. But-

Walter Mischel: We found remarkable correlations between the actual SAT scores and the seconds of delayed time.

Jad Abumrad: In other words, the kids who waited the longest when they were 4, staring at the Oreos, did better on the SATs than the kids who just gave up immediately?

Walter Mischel: Yes.

Jad Abumrad: How much better?

Walter Mischel: Significantly.

Jonah Lehrer: The differences are so big.

Jad Abumrad: That's Jonah Lehrer, science writer, often a guest on this show, who's also reporting about these follow up studies.

Jonah Lehrer: Yah, the difference between a kid who can wait one minute for the marshmallow and a kid who can wait twenty minutes...the difference in their SAT scores is 210 points.

Jad Abumrad: No way!

Jonah Lehrer: Yah, this isn't fiddling at the margins, this is a profound difference.

Walter Mischel: Well above what one would expect by chance.

Jad Abumrad: And as they dug into the data, it turned out it went way beyond SATs: the kids who waited back when they were 4, now in their late teens-

Jonah Lehrer: Parents reported that they were better behaved.

Jad Abumrad: And on the flip side, the kids who hadn't been able to wait, back in the initial study, they were more likely now to be suspended from school, to be classified as "problem kids," to know, the kind of kids who-

Walter Mischel: -were most likely to wind up as the bullies. 

Jad Abumrad: The results were so odd and strong that Mischel and his team decided to keep following those kids up through their teens, into their twenties and thirties and beyond.

Walter Mischel: So, this is now roughly 40 years, because they're about 44, 45 at this point. We're still in touch with all the 250.

Jad Abumrad: And they've expanded the kind of data that they're keeping track of. Everything from-

Walter Mischel: How well they're able to stick to goals at work, how far they go in school.

Jad Abumrad: Even their health.

Walter Mischel: For example, body mass index. 

Jonah Lehrer: A huge amount of data, they really want to amass an FBI file, so to speak.

Jad Abumrad: And the more data they collect, the worse it gets. The kids who could wait back when they were 4, now in their 40s have better jobs, they've gone farther in their education, they're even skinnier-

Walter Mischel: Yah, yah.

Jad Abumrad: -Than the kids who couldn't wait. Now think about this: how much willpower you exhibit as a 4 year old in something so insignificant as trying to resist the pull of a marshmallow could predict this frightening amount about your life?

Walter Mischel: There's no question that there was something predictive about it, and that it wasn't a fluke. 

Jad Abumrad: Here's what I'm wondering...if I have a 4 year old, and I informally give them this test, and they fail-

Jonah Lehrer: (laughs) Yah.

Jad Abumrad: Should I be worried? Do you know what I mean?

Jonah Lehrer: Yah, I mean, you know me, I've got such a bias against these kinds of rigid predictive variables when it comes to the brain.

Jad Abumrad: Yah.

Jonah Lehrer: But I think these experiments have to give you pause.

Jad Abumrad: Because one interpretation of Mischel's work, and it's one that we unfortunately cannot rule out, is that maybe self-control is hardwired. You either have it by the age of 4 or you don't, and if you don't have it, statistically, you're screwed.

There's another way to interpret this, thankfully. If willpower gets its start when you're 4, which is a big "if," but let's just assume for the moment it does, then you got to ask: what exactly happens at age 4 that separates the willful from the rest of us?

Walter Mischel: So what you're seeing over here is a tortured looking, adorable little girl wearing a blue sweatshirt...

Jad Abumrad: Walter Mischel would say go back to those videos, really look at them. If you do, much of what you see has very little to do with nature.

Walter Mischel: She's looking at it, she's withdrawing from it, and she's holding her breath...

Jad Abumrad: The first thing you notice he says is that every kid is in agony, even the good kids. They are not models of strength, they are suffering too, because Oreo cookies, man...they're good.

Second, the very simple difference between the good kids and the bad kids seems to be that the good kids just found ways to distract themselves. They had strategies. Like this girl:

Walter Mischel: She's saying "shhh" to herself, 

Jad Abumrad: Shushing, that seemed to work. Or-

Walter Mischel: Kicking.

Jad Abumrad: -Kicking the table. That's actually a good strategy, he says. Or making up a tune.

Walter Mischel: Kids composing new songs.

Jad Abumrad: Or just turning around in the chair.

Okay, we're back to girl number one now, she has her back completely turned to the cookies.

Jonah Lehrer: Some kids actually pretended the marshmallow was a cloud.

Walter Mischel: I mean, the kinds of things one sees are extraordinary.

Jonah Lehrer: Or the cookie was a UFO. They turned this hot stimulus, Mischel puts it, a cold stimulus. 

Jad Abumrad: And that might be all that separated the kids who could wait from the kids who caved.

Jonah Lehrer: The best kids simply had a better bag of tricks.

Jad Abumrad: And if that's all we're talking about...tricks...well, you can teach a trick. 

Jonah Lehrer: I think one of my favorite twists on his experiment that he did was: he found that you could take kids who had trouble delaying gratification, so kids who had really had a tough time waiting for the marshmallow, and you simply say to them, "Put a picture frame around the marshmallow, pretend it's a picture and not a delicious piece of candy." All of a sudden you gave kids this little trick, and you could turn low delayers into much higher delayers.

Jad Abumrad: Oh, so he could help these kids?

Jonah Lehrer: Absolutely, you can teach, just as a simple suggestion, "Why don't you just pretend this is a picture frame?" He's got these bar graphs and it's dramatic. All of a sudden the low delayers are performing just like the high delayers. 

Jad Abumrad: Really? And is there any evidence that if you teach these kids, these low delayers, and you give them these tools, that you can send their life in the right direction in some way?

Jonah Lehrer: There's no evidence of that yet. He hasn't shown that simply teaching kids how to draw a picture frame around a marshmallow leads to a balloon in SAT scores, and that's where, I think even Mischel admits, that it remains unclear how viable these tricks are. 

Because you can teach, obviously, a kid a trick to deal with the marshmallow in a lab setting, but that doesn't mean they'll necessarily be able to go home and...and you can't go around drawing picture frames all day long. 

Jad Abumrad: But that's kind of the crux of it, now, I kind of have to know that these tricks are useful.

Jonah Lehrer: Yah.

Jad Abumrad: Otherwise, you're just picking your interpretation, you could say, "Oh, it's learned tricks," or "It's hardwired."

Jonah Lehrer: Yah, so I think that the answer to that remains completely unknown, that they're just beginning to fly all these subjects a couple days worth of brain scans. I mean, this is a big, big project. No one quite knows what they'll find. 

Walter Mischel: I think it's highly likely to be like most things in life are turning out to be...which is: yes, the wiring makes a difference, and yes, the experience makes a difference, and the wiring and the experience are interacting and changing each other. 

Robert Krulwich: So, Jad?

Walter Mischel: Mm-hmm (affirmative)

Robert Krulwich: I, who have never been able to withhold my fierce desire to eat all the chicken when seven pieces of chicken are served to three people, I will eat four pieces of chicken...

Walter Mischel: Mm-hmm (affirmative)

Robert Krulwich: And yet, somehow I have gone through most of my life holding a job.

Jad Abumrad: Doing quite well.

Robert Krulwich: So I think, before people take Mr. Mischel's views too close to heart, remember there are outliers, there are people who have grabbed the marshmallow early, and yet who have somehow thrived.

Jad Abumrad: That's right.

Robert Krulwich: I want them to know that.

Jad Abumrad: Anyhow, Radiolab is funded by the Alfred P. Sloan Foundation,

Robert Krulwich: Yay!

Jad Abumrad: The Corporation for Public Broadcasting,

Robert Krulwich: Yay!

Jad Abumrad: and the National Science Foundation.

Robert Krulwich: Yay!

Jad Abumrad: I'm Jad Abumrad.

Robert Krulwich: And I'm Robert Krulwich.

Jad Abumrad: See you later.