The Marshmallow Test and Trust
The Stanford “marshmallow” test is a famous experiment conducted in the 1960s and 1970s. Its purpose was to measure preschoolers’ ability to delay gratification. The experiment, conducted by psychologists Walter Mischel and Ebbe Ebbesen, consisted of presenting a child with two options: get a reward immediately or a get a larger one later.
A child could receive one marshmallow (or another favorite treat — a cookie or pretzel) immediately, or, the child could get two marshmallows later. The two marshmallows were given only if the child waited for 15–20 minutes to pass, seated alone in a room with a table on which stood a plate with the two marshmallows and a bell. The child could ring the bell to call the researcher back into the room before the allotted time. If the child managed to wait the entire time, they got the two marshmallows. If they called the researcher, they could have just one.
Most children chose to delay the reward and not receive the immediate one marshmallow. Of the children that chose to delay the reward, about one-third managed to wait the time required for eligibility for the full two marshmallows.
Originally, the purpose of the experiment was to gain a better understanding of how children develop deferred gratification and their strategies for maintaining the ability to wait for a greater reward. Children’s strategies varied — self distraction, imagining the two marshmallows are “just a picture”, humming, singing, kicking furniture, rocking on chairs and even turning their back on the marshmallow.
The full significance of the marshmallow test was realized in its follow-on studies. In 1988 Mischel found that preschoolers that did well on the original test, were described by parents, ten years later as more competent adolescents. Later on success in the test was correlated with higher SAT scores (210 points more compared to the most impatient children), educational attainment and even lower body mass index measurements. Successful delay gratification also correlated with better mental health outcomes.
Many people first encountered the marshmallow test in Daniel Goleman’s bestseller, “Emotional Intelligence”. Since then it has become a popular, encouraging Cookie Monster to delay its cookie gluttony, getting favorable mentions in The Atlantic and The New Yorker and even inspiring this great TED talk (complete with videos of children trying hard to resist temptation).
Most notably, Mischel just published “The Marshmallow Test: Mastering Self Control.”. You can read an interview with him about the book here.
But sometimes eating the marshmallow makes a lot of sense.
David Ogilvy, the famous advertising man, commented on one such instance:
“When I was a boy, I always saved the cherry on my pudding for last. Then, one day, my sister stole it. From then on, I always ate the cherry first.”
Mischel has a similar thought in this well-written book.
“When preschoolers have an experience with a promise maker who fails to keep his promise, not surprisingly they are much less likely to be willing to wait for two marshmallows than to take one now”
Working in Trinidad and administering a similar test to adolescents there, Mischel started to think maybe there was an additional element to consider: trust.
“Perhaps those that came from homes with absent fathers … had fewer experiences with men who kept their promises. If so, they would have less trust that the stranger — me — would ever really show up later with the promised delayed reward. There’s no good reason for anyone to forgo the “now” unless there is trust that the “later” will materialize. In fact, when I compared the two … groups by looking only at the children who had a man living in the household, the differences between the groups disappeared”.
So, trust is important if you want to reward people for delayed gratification.
Lack of trust (or fairness) can induce adverse outcomes.
Why am I writing this? I am the CEO of an enterprise gamification company. I believe that enterprise gamification (and indeed any dialogue with employees) should be based on trust and fairness.
Gamification, the practice of using game mechanics to encourage employee behavior, requires trust too. Reading Mischel’s book I was struck by how important it is to keep the game rules straight, fair and achievable. After all, we are requesting employees to defer their gratification, modify their behavior, doing what they perhaps would rather do less or not do (completing CRM information, working better at customer service, elearning and more).
In his book Mischel describes another experiment, to answer the question “would those who delayed more in the first session be less likely to give in to a strong temptation in a different situation — one in which cheating was the only way to succeed?” Thia experiment uses badges to reward children for a game of skill (gamification reminder, anyone?). There’s one catch: the game rules are not fair.
To do this, children were introduced to a game of skill — using a “ray gun” to destroy a “rocket” target (this being the 1960s). “Above the target, a row of five lights illuminated the number of points earned after each shot. Three brightly colored sportsmen badges ( marksman, sharpshooter and expert) were flashed and offered as prizes, to be awarded on the basis of the total number of points obtained.”
But the game rules were “wrong”. The number of points a child could get for each shot was random and had no connection to their skill level. The score they received did not make them eligible for any badge. The only way to get a badge was to falsify their scores. As the boys played and kept their scores (and cheated) researchers tracked it all. The results were correlated with the delayed rewards: those that weren’t good at delay gratification cheated earlier, but“if the boys who had preferred the delayed rewards did cheat, they waited much longer”.
What’s the takeaway for employers? Keep employee trust high — it translates into “willpower” (the synonym many use when asked what the marshmallow test tests). When using rewards, gamification, anything, keep the rules of the game honest and fair. Otherwise, you’ll create a game that rewards cheaters.
Originally published at www.gameffective.com.