Children are notorious for many things: short attention spans, obsessions with cartoon characters that quickly drive parents crazy, and a demand for instant gratification. I see it, I want it, I must have it now! So my students are never surprised but always vastly entertained when I show them a very popular video recreation of the classic Marshmallow Experiment:
There are few videos I show that get more laughter or more student interest; after all, children’s difficulty waiting for something they want is vastly entertaining when carefully edited with peppy backgrounds music (instead of the more realistic scenario, in the middle of the grocery store and accompanied by a temper tantrum).
But the video is more than just entertainment; it’s based on a famous experiment created by Walter Mischel, who many years ago placed marshmallows in front of children to investigate why some seem to resist temptation with ease while others swallow the marshmallow whole immediately. And it’s not just a child’s cute success or failure here-and-now, but potentially their success or failure throughout their lives: The marshmallow task has been heralded as a surprisingly powerful predictive tool about how children’s futures might turn out. It might not be too surprising that resisting a marshmallow as a preschooler can predict how much you weigh as an adult. But you might not have thought that plopping a marshmallow in front of a child could give you some idea of how well they’d do in high school – their academics, their ability to deal with stress, even their social competence.
Spend an hour or so reading about all the studies that have been done with preschoolers and marshmallows and two things will happen. First, you’ll either have an urge to consume some Jet-Puffed marshmallows, possibly with some idle musings about whether Kraft should get on board sponsoring this research. Second, you’ll get the impression that all those fancy intelligence tests and end-of-year achievement tests should be scrapped, in favor of placing kids at a table with a fluffy sugar concoction for five minutes.
This second reaction is exactly what Po Bronson and Ashely Merryman, co-authors of NurtureShock, warn against. These science writers point out that the original marshmallow experiment, featuring long delays with no help or special instructions as shown in that video, included only three dozen preschoolers, not all of whom could be tracked all the way to adolescence and beyond. How much do you really want to conclude from just one group of kids, who all happened to go to the same Stanford preschool? After all, they say, another study, with a different batch of three dozen kids, suggested that a marshmallow task (now a cookie task) predicted absolutely nothing about intelligence fourteen years later. In other words, “just let them eat the marshmallow“.
Of course, everyone has some bias in interpreting the evidence. The same study that Bronson and Merryman use to show that the marshmallow task doesn’t really predict teenage anything was actually published for showing that delay of gratification at 4 years old predicted some specific cognitive abilities at 18 years old. Both claims are “true”, depending on what part of children’s behavior you want to look at. As Bronson and Merryman argue, the amount of time that children resisted the marshmallow didn’t predict anything – and with over a third of the kids waiting the full 15 minutes, the task might be a lot easier than the video makes it seem. But lead researcher Eigsti and her colleagues (including Walter Mischel himself) emphasized that children’s “temptation focus”, or how much time they spent fiddling with the treat or the bell, did predict something of future: Preschoolers who spent less time starting temptation in the face became adolescents who were faster at stopping reactions on demand.
In other words, looking back at the video of those cute kids, the little boy who spent so much time inhaling the delicious scent of the marshmallow with such a pained expression on his face has the same predicted future as the little girl whose mouth was full before the adult even left the room. The power of the marshmallow task is not how long you wait, but whether you can get that marshmallow out of your mind when you need to.
This leads me to two bigger truths. First: In science as in politics, people will focus on the evidence that supports what they want to believe. Second, they can do this because the evidence is rarely dramatic and wholly in support of just one view. Putting a child in an empty room with a marshmallow can tell you something important about that child, but what it tells you depends on what you look at, and you shouldn’t expect it to be a definitive view of the future.
That being said, I’d be all in favor of making the marshmallow task a standard early childhood experience. We’d learn a lot more about how much it can tell us…and we’d get an awful lot of entertaining videos along the way.
Eigsti, I-M., et al. (2006). Predicting cognitive control from preschool to late adolescence and young adulthood. Psychological Science, 17, 478-484 DOI: 10.1111/j.1467-9280.2006.01732.x