John Carney of CNBC
is suspicious of the explanation for the results of
the Stanford marshmallow experiment - which is also getting some
play today by Joe Wiesenthal at Business Insider. Here's Carney - who, by the way, once worked on one of Pat Buchanan's presidential campaigns:
Economists tend to see this as a measure of time preference. Some say it is an indicator of general intelligence. I’ve heard investors talk about this as a test of whether someone has the character to be an investor for the long-term. Even short-term traders have told me it is a good test of the ability to see a strategy through.
I’m not convinced. The first thing that raises a red flag with me is that the marshmallow test better predicts SAT scores than IQ tests. Now, SATs are basically IQ tests. But IQ tests are actually better predictors of general intelligence than SAT scores. This means that both the marshmallow test and the SAT are testing for something in addition to intelligence.
What is it that they are testing? I think it’s compliance with rules set by authorities. The kids who do not eat the marshmallow are not exercising "self-control." They are submitting to outside control. Similarly, I suspect that those who follow instructions carefully probably are able to outperform their IQs on the SATs.
It actually makes sense for the SAT to test for compliant personalities. The SAT is supposed to test for college aptitude, not just general intelligence. Following rules and obeying authority is a critical skill for success not just in educational settings, but in our increasingly rigid society. Conformity to the rules set by authorities and trust that promised rewards will be delivered are probably very important to success.
To put it slightly differently, the SAT seems to be prejudiced against high IQ non-conformists and those skeptical of authority—the kids who didn’t wait for the marshamallow.
Since the marshmallow experiment gave children a choice as to whether to eat the marshmallow soon or late, it would seem that whether to obey authority or not doesn't enter into it. Of course, the experimenter could have given off subtle clues and hints that told the child that he wanted him to refrain from eating the marshmallow, but we don't know that.
The SAT can be studied for, while IQ tests cannot; therefore, the other main quality which the SAT probably determines is precisely what the marshmallow experiment purportedly measures, i.e. time preference and (lack of) impulsivity. Both are correlated with IQ but separate from it, which would explain why the marshmallow experiment is even more correlated with SAT results than IQ tests are.
But, whether the test measures impulsivity, obedience to authority, IQ, or some combination thereof, the important takeaway here is that a test given to four-year-olds can more or less accurately predict future success - or lack of it. The reigning ideology has it that people, and especially children, are well nigh infinitely malleable - Head Start! Improve our schools! Leave no child behind! - yet the experiment shows that notion to be seriously flawed at best, totally invalid at worst. (My bet is on "totally invalid".)
In his book,
We Are Doomed, (which you should buy and read, it's excellent), John Derbyshire quotes Deborah Solomon of the NY Times, who in the course of an interview with Charles Murray, said, "I believe that given the opportunity, most people could do most anything." (To which Murray replied, "You're out of touch with reality in that regard.")
Solomon couldn't have given a better precis of the reigning ideology in education. The significance of the Stanford marshmallow experiment is that it shows how profoundly wrong and damaging that ideology is.