Can a Marshmallow Predict Your Future?

I See...A MarshmallowI See...A MarshmallowSome 40 years ago, in the early 1970s, Stanford University psychology professor Walter Mischel published a groundbreaking study testing children’s self-control and delayed gratification. In the study, children were presented with a treat (sometimes a marshmallow, but pretzels and chocolate were also used) and told that they could eat the treat now, or wait until the adult returned, at which point they would be given two treats. The original purpose of the study was to determine the age at which children develop the ability to delay gratification, but follow-up studies revealed something much more meaningful and shocking.

The impetus for Mischel’s study was one he published in 1958, involving children in Trinidad. On the island, stereotypes existed about different ethnic groups' recklessness and other behavioral tendencies. Mischel performed a small study involving about 50 children from Black and East Indian families, giving them a small treat, but promising a much bigger treat if they had not eaten theirs when he returned in a week. He found that a significant ethnic difference did, indeed, exist; Black children were significantly more likely to eat their treat than those from East Indian families.

The Stanford Marshmallow Experiment, which took place at the Bing Nursery School on the university campus, was much larger, eventually including three experiments, with a total of 600 participants. The setup for the first test The Marshmallow TestThe Marshmallow Testwas simple. Children between the ages of 3 and 5 years were brought in one at a time with a research assistant, and given a treat. Each was told that the treat was theirs to do with as they wished, but that they would receive a second treat (e.g., two marshmallows) if they waited and did not eat the treat until the research assistant returned as much as 15 minutes later. Subjects could call for the researcher to return if they decided to wait no longer and simply take the single reward. Each subject was timed, filmed, and watched through a two-way mirror. Variations were also performed in which the children could not see the treat they were resisting, or with other less (or more) desirable treats present, etc.

Very few of the children were unable to hold out at all, but there were significant variations in children’s ability to delay gratification for a larger reward later. The children engaged in adorable, yet somewhat sophisticated, strategies for helping themselves resist. Some sang songs aloud to distract themselves. Others repeated the rules aloud. Some turned their chairs away for long stretches of time, or covered their eyes.

Many years later, this simple study became more than a mildly interesting examination of self-control. Mischel’s first follow-up, published in 1990, revealed a shocking result. Those children who were most able to exhibit self-control were described by their parents as significantly more competent. Two years later, Mischel found that the children’s ability to delay gratification even correlated with higher SAT (Scholastic Achievement Test) scores. Later correlations have even been found with teen pregnancy, criminal behavior, and BMI (Body Mass Index) ratings of overall physical health. It seems that, for whatever reason, subjects as early as 3 years of age showed a pattern of self-control (or lack thereof) that tended to predict their overall success in life. Many psychologists suggested that the most likely culprit was a home environment that either encouraged or discouraged (perhaps even implicitly) self-control, creating a personality type that followed these children into adulthood, leading them to make life-changing decisions that fit the pattern.

Recently, a group of researchers at the University of Rochester in New York devised a test that may provide an even more complete picture of the results of Mischel’s study. In it, they replicated the now famous Marshmallow Experiment, but with one twist. Before children were presented with the self-control portion of the experiment, they were betrayed. Specifically, they were given old, used, broken crayons in a mason jar and invited to color a picture. The research assistant then promised to return in a few minutes with a set of brand new, beautiful coloring pens, pencils, and crayons for the child to use. In half of cases, the research assistant followed through on that promise; in the other half, she returned empty handed, apologized, and made an excuse for not having the better set of coloring tools.

This time, children who had been “betrayed” showed the least apparent ability to resist temptation and delay gratification, leading the researchers to conclude that the causal factor may be a home life in which adults can Mean Wait TimeMean Wait Timebe trusted to follow through on their promises (or where good things actually do come to those who wait). This is important because it replaces Mischel’s conclusions about “good” and “bad” children, with one that suggests children who choose not to delay gratification may actually have been making a logical decision; based on their past experiences, a bird in the hand is, indeed, worth two in the bush. In other words, take the sure thing now because you can’t be sure what will happen in the future.

I will leave it to my fellow skeptics to determine the myriad of areas where these findings apply, but for me it's a great example of the self-correcting nature of science. I learned about the Marshmallow Test in graduate school, and became quite attached to its findings. However, rather than being "offended" or rejecting these new findings, I feel a renewed sense of wonder and a desire to learn more about the subject it attempts to understand. That reaction is actually quite common in science; would that it were so in other realms of understanding the world we live in.