When to eat the marshmallow: new perspectives on impulse control

In light of the fact that many readers will have an assortment of Christmas treats tempting them, I thought a post on impulse control would be timely.

In the now paradigmatic Stanford marshmallow experiment, children were given an option – one marshmallow which they could have immediately, or two marshmallows, provided they could wait 15 minutes. This option presents a problem of sorts. Is it better to have a small reward immediately, or a larger one after some delay? Common sense says that waiting is the better option. Doubling your reward whilst only paying a marginal cost of your time seems like the rational thing to do. Children who fail to wait are, therefore, seen as succumbing to temptation. A deficiency in self control leads them to make a poor decision.

Whether or not individuals are able to wait for the extra marshmallow at age 5 seems to be indicative of their long term capacity to make good decisions. Children who wait for the extra marshmallow go on to be more successful as adolescents across a broad range of areas. They perform better academically, are more socially competent, and are better able to deal with frustration and stress.  The reason for the difference in long term outcomes between individuals can, in large part, be attributed to self control. Those who are more able to resist temptations will better be able to pursue their long term goals.

However a recent study indicates that it may be more than just self control that explains why some children wait for the extra marshmallow and some don’t. Whether the children have reason to trust the person offering them the marshmallow is a large factor in determining whether they will wait for the second marshmallow. In the study children performed a simple art task. All children were offered a set of high quality art supplies with which to do the task. For half the participants (in the reliable condition) the experimenter returned with the promised supplies. For the other participants (in the unreliable condition) the experimenter returned without the supplies and explained she had been wrong and that the children had to instead perform the task with some old crayons. All of the children were then given the classic marshmallow option – one marshmallow now, or two if they can wait 15 minutes.  The broken promise made a large difference to which option the children took. Only 7% of the children in the unreliable condition waited the full 15 minutes for the second marshmallow, as opposed to 64% in the reliable condition.

The difference in performance between these groups of children seems completely rational. If you can’t trust someone to bring you art supplies, why risk a guaranteed marshmallow on their word? This makes explicit the fact that under certain conditions, waiting the extra 15 minutes for the second marshmallow is not the best decision. Delaying gratification is a strategy that only pays off in certain, stable environments. If the environment is unstable – if you can’t trust the offer of the second marshmallow – it is better to take your opportunity while you can.

This type of research has implications for debates regarding embryo selection.  If we accept that we have good reasons to select embryos that will be the most likely to have good lives then we will have reasons to select embryos who are predisposed to delay gratification in the classic marshmallow test. As stated earlier, performance in this task indicates children will end up doing better in a range of measures that effect well-being.

Choosing embryos who are predisposed to delay gratification in the classic marshmallow experiment will primarily involve looking for genes associated with impulse control.  Having good impulse control means that if individuals choose to delay gratification they will be able to do so. However it may also involve looking for genes which influence an individual’s ability to recognise when delaying gratification is a good decision and when it is poor one. In other words we need to look for factors that influence a person’s ability to know when it is best to wait, and when it is best to just eat the marshmallow.

I trust you will all make the right decision about when to eat your Christmas goodies!

  • Facebook
  • Twitter
  • Reddit

2 Responses to When to eat the marshmallow: new perspectives on impulse control

  • Anders Sandberg says:

    Generally, I would suspect enhancing children’s ability to 1) accurately judge character, 2) predict the consequences of their and other’s actions, 3) control their actions would be a pretty good moral and cognitive enhancement. These abilities are also to some extent learnable, so enhancing the ability to learn them would also be a good idea.

  • Lizabeth says:

    Learning to distrust and therefore taking precautions to protect oneself from being duped sheds a new and interesting light on marshmallow test. Children adapt and hold those lessons close.

    It will be fascinating to see what can be extrapolated from similar studies over time. Will a reasonable connection be made to the horrifying acts of violence we have been witnessing as of late? Will the general public wait for sound science to prove or disprove any connection? History would suggest that the public will not.

Recent Comments

Authors

Affiliations