Wired For Success: Brain Scans Predict Mathematical Ability

This is a huge story on Reddit right now, getting a staggering 4000+ up-votes in less than eight hours: Brain scans better forecast math learning in kids than do skill tests, study finds: Gray matter volume and connections between several brain regions better forecast 8-year-olds’ acquisition of math skills than their performance on standard math tests.

The viral response on Reddit is evidence that the large millennial demographic has an interest in HBD-related matters, whereas with older generations you don’t see as much enthusiasm. Millions of Young people are willing to give HBD the benefit of the doubt instead of dismissing it a racist, sexist, or whatever. Maybe this is because millennials, many of whom are well-educated and have superior critical thinking skills, are realizing that the left’s environment-based explanations (not enough public school funding, rich people, poverty, etc) for education and socioeconomic under-achievement among certain groups has fallen short, and that biology, not environment, is the primary cause of underachievement. That means people are falling behind because of low IQs or ‘poor brain wiring’ in an economy where intellect is increasingly important. People who are wired for math success are also wired for socioeconomic success, as IQ and income are highly correlated, and in the post-2008 economy this trend has accelerated. The hyper-competitive economy has magnified the economic consequences of individual cognitive differences, such that a person with an IQ 120, for example, is at a much greater advantage than, say, 100 years ago. Until the 2000′s, people who were bad at math still had a lot of good career options, but nowadays with manufacturing jobs going away and everything being outsourced and automated, math, and intellect, and wealth are becoming increasingly intertwined and inseparable. That article hits close to home – maybe those kids who were whizzes at math in elementary school and are now making six-figures in STEM jobs didn’t just make better life-decisions, but are actually biologically better than everyone else, too. That means you have much less free will than you want to believe, and that biology preordains some to success but all too many to mediocrity or failure, and there’s little that anyone can do about it.

As shown below, the income disparity between advanced degree holders and everyone else has widened markedly since 2010. Getting an advanced degree typically requires an IQ of at least 115, hence being smart has never been so important for socioeconomic advancement as it is today. If you don’t have an IQ that is at least one standard deviation above average, you life may suck.

The research tracked 43 children longitudinally for six years, starting at age 8, and showed that while brain characteristics strongly indicated which children would be the best math learners over the following six years, the children’s performance on math, reading, IQ and memory tests at age 8 did not.

Hmmm….but I also wonder among the sample how many poor test performers had unfavorable brain scans, versus good test takers? It could be a small faction of poor test takers have good scans , whereas a larger percentage of good test takers do. The study does not explain why the relationship holds in the future but not at the time of the initial test.

This was the most up-voted comment on Reddit:

There is a difference between aptitude and achievement. Psychometric (standardized) tests are most often tests of achievement, not aptitude. The measure of connectivity and cortical thickness is a measure of aptitude, not achievement.
The brain scans will not tell you if a child understands complex fractions. An achievement test will tell you this. An achievement test will not tell you if the same child could one day master tensor calculus. The article suggests this may be possible.
Neuroimaging will not replace achievement tests, and achievement tests are not good measures of aptitude.
Note that the researchers used neuroimaging to measure aptitude, and then psychometric tests to assess achievement later on.

However, although millennials show a keen interest in IQ and HBD related topics, confusions still remain. A psychometric test is not always the same as a standardized test. A psychometric test doesn’t even have to be an IQ test. Some psychometric tests test for personality or mental disorders, etc. A well-constructed IQ test should not be an achievement test, meaning that it’s possible for someone who gets low grades or is even illiterate to perform well on an IQ test. IQ tests or the pre-1995 SAT are good aptitude tests, and the later is also effective as a achievement test. As for cortical thickness being a proxy for aptitude and IQ, the relationship isn’t actually so clear. Counterintuitively, according to a study, high-IQ individuals early in life have less cortical thickness than average-IQ subjects:

Later in life, cortical thickness tends to be the same for all IQ levels.

Comments are closed.