Time and Marshmallows

By Sean Carroll | March 12, 2012 5:42 am

“Perhaps no one comprehends the roots of depravity and cruelty better than Philip Zimbardo.” At least, that’s what it says here. They’re referring to the fact that Zimbardo — a psychologist who long ago supervised the notorious Stanford Prison Experiment (chilling video here) — is an expert on the psychology of “evil” behavior. But he’s also an expert on the psychology of time, which we can all agree is much more interesting.

I recently got to hear a talk by Zimbardo, in which among other things he discussed the Stanford Marshmallow Experiment — a rather more adorable experience than the prison experiment, from what I understand. The Marshmallow experiment, originally conducted by Walter Mischel in 1972, was aimed at understanding how we think about different times — the future vs. the present. Children were asked to do some easy tasks, and then were rewarded by being given a marshmallow. But! They were told that the experimenter had to step outside for a few minutes, and if they could just sit tight and not eat their marshmallow until he came back, they could have that and also an additional marshmallow.

It’s a matter of future vs. present rewards. It’s natural (and totally rational) to discount rewards that are promised in the future — after all, the future is hard to predict, and anything can happen. If I offered you a choice between $4 today and $5 ten years from now, you’d be sensible to take the lower amount today — depending on how much you trusted me, of course. But if there is a good reason to trust, and the future isn’t that far off, it makes sense to delay gratification a bit. So what happens when some four-year-olds are put to the test?

As you see, the results were mixed. But most importantly, Mischel followed up years later, looking into how the kids who participated in the study ultimately turned out. There was a remarkable amount of correlation with this simple test and success later in life — kids who were able to hold off at age 4 for the second marshmallow turned out years later to have higher SAT scores and generally seem more competent. The hypothetical explanation is that our personalities are strongly influenced by our attitude toward time — whether we are focused primarily on the past, the present, or the future.

Zimbardo has learned a lot about how people treat different times differently, depending on personality and culture and numerous other factors. Here’s a great animated version of a talk he gave on the subject.

CATEGORIZED UNDER: Science, Time, Top Posts
  • http://wbmh.blogspot.com/2010/09/new-proof-for-existence-of-god.html wolfgang

    >> I offered you a choice between $4 today and $5 ten years from now, you’d be sensible to take the lower amount today

    Not really. A 25% return over 10 years is more than you currently get with treasuries (yielding less than 2% annually and thus returning about 22% over 10 years).

  • Naked Bunny with a Whip

    I’d have to research inflation and interest rates, and keep track of the agreement and the person offering the money, and by now I’ve already expended more than $1 worth of my time.

  • Gizelle Janine

    *starts laughing*

  • Gizelle Janine

    How about this one:

    Fourth law of thermodynamics (lowercased): Quantum mechanical laws apply to the zeroth, first, second third and fourth law of thermodynamics.

    …if a fourth one ever pops up anywhere anytime soon. I hear David Albert somewhere in NYC calling me crazy….

  • michael

    The little boy who waited for the piece of chocolate was doing everything right. He was given permission to eat the chocolate as a reward for playing the game. Then, with that settled in his mind, and now on a new issue, he decided it would also be profitable to do some waiting for a while, to see if that promised second piece of chocolate was actually forthcoming. No where did I hear him being reminded of, or even introduced to, the consequences of eating the first piece before the man returns. He was acting in good faith, and his simplistic understanding of the words that the authoritative giant in the room was using, was misinterpreted as knowledge of the situation.

  • Colugo

    Let’s be very careful with how we interpret such studies and their implications. What causes variation in impulse control? Can children with less control be helped in controlling their impulses?

    Then there’s bioethicist Julian Saveluscu:


    “I think people should test to see and to see that they provide their child with the least risk of developing serious diseases later in life. But other genes also affect how well our lives go. So take an example and there’s famous set of experiments done in the 60s. This guy called Water Michelle put a marshmallow in front of three year old children. And he said “I’m going to go out of the room, don’t eat the marshmallow and if you don’t eat the marshmallow I’ll give you two when I come back.” So he went out of the room, came back, some kids had just eaten the marshmallow, some kids hadn’t. And he followed these kids up later, ten years time, he found those who didn’t eat the marshmallow, who were able to control their impulses delayed gratification had more friends, more motivation to succeed and better academic performance, and this was more highly correlated with their university entrance than their IQ was. So if you had a range of embryos, and you were able to test for the genes that were related to impulse control, as one day we will be able to, I think those genes are pretty significant and I’d certainly want to have a child that was able to have some degree of impulse control.”

  • Chris

    What if we did a combination of the two Stanford experiments? If you eat the marshmallow, you go to jail. This is why I’m in chemistry, I don’t have people skills!

    Also I do have extremely high impulse control. Birthday/Christmas money in the bank, candy sitting in the fridge for years. And this was when I was a little kid.

  • JR

    I get less and less future oriented the more I get beaten down by the idea that even if I work hard and plan, the element of luck will decide whether I or that other future oriented person gets a return for their planning.

    Ergo, may as well be present oriented.

  • Tony Mach

    Well, the proposition he makes to these kid is misleading. He should have said something like that: “You can eat this one now OR if you leave this one alone until I come back in a few minutes, then you will get another one.” One should put much more thought into how to phrase this. Much, much more thought.

    This has nothing to do with time, but some kids simply didn’t understand his “if/then” proposition.

    And these damn psych* researchers can get away with such sloppy study design and even receive praise for it.

  • Colugo

    After Hauser and Stapel I’m much more skeptical of cogntive psychology experimental results, interpretations, and putative implications.

  • Jason

    I think this correlation has less to do with the success of individuals and more to do with how society rewards obedience and/or blind trust. The 1st girl in this video he told her the truth when he just said he was going out of the room. The others when he said he was going to check his car, he lied to them. How do we know they didn’t pick up on that lie?

  • Mark Erickson


  • Mark Erickson

    I call BS on prof. Z. Brains are not rewired by video games. Every generation says that today’s kids are not like them. Distrust anyone who says human nature has changed and scorn anyone who says it has fundamentally changed.

  • Albert Einstein

    Right, so people who fail in life do so not because of “oppression” by imperialist paternalist blah blah blah but because they lack self-restraint. Another liberal superstition exploded.

  • Low Math, Meekly Interacting

    The intro puts a lot of emphasis on temporal perception and tendency to trust. My understanding of this experiment is that the children all likely believed a second marshmallow would soon be theirs for the taking. What was being tested was whether or not they could control their impulse to eat the marshmallow immediately in exchange for a reward. It was a a test of self-control, about the need for immediate gratification quite independent of future expectations. The most interesting aspect of this experiment is that some kids eat the marshmallow even though on some level they know it will cost them a marshmallow in the future.

  • http://www.rebeccastanek.com/blog/ becky

    An interpretation of the marshmallow test that does not get enough consideration is that it proves the hypothesis that children who come from homes where promises are routinely kept do better in life.

    I have a toddler, and my partner and I work very hard to make sure that we are honest and reliable. We don’t say “maybe” or “later” to him if we mean “no”.

    And fwiw, while parenting books should be taken with many grains of salt, the ones that have worked well for us also emphasize that.

  • http://wbmh.blogspot.com wolfgang

    One more possible alternative explanation:

    There were studies which showed that consumption of refined sugar is bad for your brain. It seems natural that kids who cannot resist candies will consume more sugar in general and therefore will do worse in the long run (on average).

  • Clausentum

    Quite apart from the usual caveats about soft-science, the massive plug at the end of the clip for the latest flavour-of-the-month self-help book doesn’t look like dispassionate science as we know it.

    As others point out above, all sorts of interpretations of this experiment are possible. The time one looks like over-interpretation.

  • marshall

    Sometimes human societies are stable and rational, sometimes they are not. Kids are brought up in both times, and need to be able to adapt to both. In stable times, it may be rational to accept promises for the future. In unstable times, it may not be. If Sean Carroll offered me, today, a choice between $ 4 now and $5 in 10 years, I might take the $5. If he made the same offer, today, to a child in Afghanistan, I would advise that kid to take the $4 and run.

  • Pingback: Your Outlook on Time Could Affect Your Life (Discover) « Bowdoin Daily Sun()

  • CJSF

    In 10 years, that $1 difference might not mean a heck of a lot. Also, 10 years is a long time. If I’m toiling away on a mine on the Moon and the person offering me the money is dying in a hospital in Omaha, what good does that do me? Even if the offer was to double it to 8 dollars, that’s just not worth it to me to wait 10 years.


  • Christian Takacs

    Did anyone consider that if the kid was hungry “now” , he/she would naturally eat now. Kids being as active as they are , often are hungry. If I’m hungry now, you telling me you will buy me dinner ‘later’ without being more definite when that is will not slow me in the least from eating. Like most psychology pseudo-science, the real understanding is to be gained by watching the clowns in the white lab coats look like they know something.

  • Kaleberg

    The experiments were performed at the recently formed Bing Nursery School, an experimental day care facility at Stanford University and relied on the researcher’s ability to track about 100 of the over 600 original participants. All told, it was a pretty good study, but the data collection was rather minimal. I don’t now if the data was available, but it would have been even more interesting if the study had controlled for social class, access to societal resources.

    If you live in a world of layoffs, hour cuts, evictions, crime, lousy cars, poor medical care, minimal savings and the like, then you learn it makes a lot more sense to take the marshmallow and enjoy it now. If you live in a world with secure incomes, some savings, a safe neighborhood, good medical care and the like, then you learn it makes a lot more sense to wait for another marshmallow.

  • http://begreen.botw.org Be Green

    Behavioral economists refer to this study often to show the correlation between personal success and the ability to defer personal gratification. A follow up study in 2011 indicates that the trait sticks with you for life.

  • http://www.fieldopticsresearch.com/ Julie Warner

    I’m wondering if this is entirely due to the child’s own personal makeup or if the influence of home (nature vs nurture) might play a part. If a child is given everything immediately when they want it, does this foster a present oriented person?

  • Seth

    This topic was covered in an equally entertaining/informative fashion on the public radio program RadioLab. Link: http://www.radiolab.org/blogs/radiolab-blog/2009/mar/09/mischels-marshmallows/

  • Pingback: ATLAS Peeped | Cosmic Variance | Discover Magazine()

  • Pingback: ATLAS Peeped – - ScienceNewsX - Science News AggregatorScienceNewsX – Science News Aggregator()

  • jansuch

    >>Not really. A 25% return over 10 years is more than you currently get with treasuries (yielding less than 2% annually and thus returning about 22% over 10 years).

    Nitpick: You are using 1-year rates and you should be using 10-year rates, which actually do return more than $5. Never mind that you should be comparing to the individual’s next best use of the money, which is certainly not to invest them in Treasury Bills.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Cosmic Variance

Random samplings from a universe of ideas.

About Sean Carroll

Sean Carroll is a Senior Research Associate in the Department of Physics at the California Institute of Technology. His research interests include theoretical aspects of cosmology, field theory, and gravitation. His most recent book is The Particle at the End of the Universe, about the Large Hadron Collider and the search for the Higgs boson. Here are some of his favorite blog posts, home page, and email: carroll [at] cosmicvariance.com .


See More

Collapse bottom bar