How Much Does Intelligence Matter for Creative Achievement and Professional Success?

A brief review of the Study of Mathematically Precocious Youth (SMPY).

Longitudinal studies of the highly intelligent demonstrate the importance of IQ for creative and professional achievement. The Study of Mathematically Precocious Youth (SMPY), started in the early 1970’s by Julian Stanley and now run by David Lubinski and Camila Benbow at Vanderbilt University, has tracked multiple cohorts of high-IQ individuals, the oldest of which they’ve tracked for nearly 50 years.

SMPY identifies gifted youths by administering the SAT, a highly g-loaded test typically taken by high school seniors, to thirteen-year-olds. In the earliest SMPY cohorts, the threshold for being accepted was the top percentile of cognitive ability in one’s age group. In some of the later cohorts, the threshold was raised to top 0.5% and even top 0.01% of cognitive ability (Lubinski and Benbow, 2006)

Some have argued that cognitive ability is a threshold for success; past a certain point, it doesn’t matter (Jauk et al., 2013; Gladwell, Outliers). This is called the “threshold hypothesis” of IQ. For example, analysis of the Terman study, a program for high-IQ individuals that predates SMPY by nearly 40 years, finds that “Termanites” (the students selected for the program), while successful, weren’t as remarkably accomplished as one might expect given their high IQs. However, this interpretation ignores the fact that the base rate of extreme accomplishment—for example, winning the Nobel prize—is extremely low.

The SMPY data demonstrate that the “threshold hypothesis” is likely false; IQ predicts variation in accomplishment among the general population as well as among the extremely bright. For example, 2% of the general population has a Ph.D.; among individuals in the top percentile of cognitive ability, the rate increases to 23% (Kell et al., 2013a, p. 5). Those in the bottom quartile of SAT math scores among the first three SMPY cohorts—which puts them in the top percentile of math ability among the general population—earn doctorates at a rate of 15%. However, those in the highest quartile of the SMPY cohorts (that is, those in the top 0.01% of math ability), earn doctorates at 2.7x the rate of the bottom quartile.

This pattern of compounding gains of intelligence holds for many realms of creative and professional accomplishment. The odds ratio (OR) for the top quartile of SMPY (top 0.01% of math ability) compared to the bottom quartile (top 1% of math ability) are as follows (Robertson et al., 2010):

  • Any Peer-reviewed publication (OR = 4.5)
  • STEM publications (OR = 5.9)
  • STEM Doctorates (OR = 18.2)
  • Having more than one patent (OR = 6.1)
  • Income in the 95th percentile (OR = 3.3)
  • STEM tenure at a top 50 university (OR = 7.7)
Professional Outcomes Odds Ratios by SAT-M Quartile. The Odds Ratio is calculated as the ratio between the proportion of outcome X (e.g. having a patent) among those in the top quartile of SMPY SAT math scores vs. the proportion in the bottom quartile.

To state it plainly: in the U.S., those in the top 0.01% of mathematical ability are, on average, 18 times as likely to earn a STEM Ph.D. as someone in the top 1% of mathematical ability. Evidently, even among the highly intelligent, intelligence keeps on giving.

Intelligence differences cause variation in creative achievement among the highly accomplished, too. Among the 534 individuals with doctorates in the first three cohorts of the SMPY sample, those in the top quartile of mathematical ability are 3.07 times more likely to have at least one STEM publication and 4.89 times more likely to have a patent than those in the bottom quartile (Park, 2008). Thus, even among those with doctorates, innate differences in intelligence affect the probability of getting a patent.

This isn’t to say that mathematical ability is all that matters, or even that intelligence is all that matters. There are certain traits, like spatial ability, which aren’t tested on the SAT, yet add significant incremental validity in predicting who gets patents, peer-reviewed publications, and so on (Wai & Kell, 2017; Kell et al., 2013b).

Lest you think the Study of Mathematically Precocious Youth is a fluke, it has been longitudinally tracking individuals for upwards of fifty years, has sample sizes in the thousands, and has been replicated in its sister program, the Duke Talent Identification Program.

The main takeaway is intelligence matters. It is one of a few important determinants of creative achievement and success. Though it certainly isn’t sufficient for creative and professional success, in most cases it is necessary and, among the far-right tail, produces super-linear returns.