One of the goals of intelligence testing has been to help to
identify those people who are likely to be productive in life and to give them
access to resources that will allow them to maximize their potential.
The danger with these intelligence tests, though, is that
they do not perfectly predict what will happen to people in the future. In his fascinating book, Ungifted, Scott Barry Kaufman
provides both a personal story as well as data about the strengths and
weaknesses of intelligence testing. On
the personal side, he had difficulties in school early on, and performed poorly
on tests of intelligence, but he has published significant scientific work in
the field of intelligence. So, the tests
were not a good predictor of his future success. He is not alone. Intelligence tests predict only a small part
of what makes people successful.
Perhaps what is missing is other intelligence tests. This question was explored in a paper in the
September, 2013 issue of Psychological Science by Harrison Kell, David Lubinski, Camilla
Benbow, and James Steiger.
They examined data from a longitudinal study of participants
who were first given the SAT as 13-year-olds in the 1970s. The participants all scored in the highest
one-half of one percent on either the math or verbal SAT. Later, they were given other tests, including
two tests of spatial reasoning ability.
One of these tests, (the Mechanical Reasoning portion of the
Differential Aptitude Test explores people’s ability to reason about
configurations of devices like gears and pulleys. Another (the Spatial Reasoning portion of the
DAT) looks at people’s ability to construct boxes and shapes from segments.
The researchers examined whether the 563 individuals in this
sample had published papers in research journals and had gotten patents for
inventions. Overall, 160 individuals
from this sample had published journal articles or had a patent.
A key finding from this research was that the combination of
math and verbal SAT score could be used to predict which people in the same
were likely to publish in research journals.
Perhaps unsurprisingly, people who published in science and technology
journals tended to have higher Math SAT scores as children than those who
published in Arts, Humanities and Social Science Journals. Those who published in journals tended to have
higher verbal SAT scores as children than those who just had patents.
Overall, this analysis predicted about 11% of the difference
among people in their types of publications.
Spatial ability added to this prediction. Those people who published in the sciences
and who held patents had higher scores on the tests of spatial reasoning as
children than those who published in arts and humanities journals or in biology
and medical journals. The addition of
the spatial ability scores added the ability to predict an additional 7% of the
differences among people.
How should these results be interpreted?
On the one hand, it is clear that extraordinary intelligence
scores obtained as children do have some influence on later performance. Over 25% of the people in this sample
ultimately published papers in research journals or had patents. That is a higher percentage than you would
expect in the population of the United States at large.
At the same time, that means that most of the people in this
sample did not publish or have patents.
In addition, though SAT scores and spatial ability scores help to
predict the fields people will enter to some degree, most of the differences
among people involve factors beyond these tests of ability.
Ultimately, tests of intelligence or ability are a
relatively small predictor of people’s success.
As I discuss in Smart Thinking, raw ability is
useless on its own. Success is built on
a basis of hard work acquiring an area of expertise and honing that knowledge
to use it to solve new and interesting problems. Intelligence may make it a little easier to
acquire that knowledge and to use it, but it does not substitute for years of
hard work.
