You're viewing revision R0 from Nov 20, 2026, 10:33:14 AM CST (reviewed and posted). See the current version.

An IQ score is a ranking. Early tests divided a child's "mental age" by their real age. Modern tests do not. They compare your results to a large sample of people your age and turn the difference into a number where 100 is average and 15 points is one standard deviation. About two thirds of people land between 85 and 115, and about 95 percent between 70 and 130.[1] So the number does not measure how much brain you have. It measures where you sit in a crowd.

What the score predicts

It predicts more than critics like to admit. Tarmo Strenze pooled decades of long term studies in 2007. Childhood ability correlated about 0.55 with how far people got in school and about 0.45 with the status of their jobs, and much less with income.[2][3] Those are strong numbers for social science. But Strenze also found intelligence was not overwhelmingly better at predicting success than parents' wealth and status, or school grades.[2] It is one strong signal among several, not the master key.

What it misses

The sharpest critique comes from inside the field. Psychologist Keith Stanovich points out that the tests used to measure good decision making in modern cognitive science do not appear on common IQ tests at all.[4] He coined the word "dysrationalia" for smart people who think badly. His favorite example: a bat and a ball cost $1.10, and the bat costs a dollar more than the ball. The quick answer, 10 cents, is wrong. Catching that takes a habit of checking your first answer, and that habit is only weakly linked to intelligence.[5]

Economists found a bigger gap. A 2014 OECD working paper led by James Heckman concluded that IQ and achievement tests do not capture personality, goals, character and motivation, and that for many life outcomes those skills predict as well as or better than cognitive ones.[6]

The scores move

In 1904 Charles Spearman noticed that people who do well on one kind of mental test tend to do well on others, and proposed a single general factor, now called g. Stanovich, critic and all, still calls what IQ tests measure meaningful and useful.[7]

But the scores are not fixed. Average results rose about 3 points per decade through the 20th century, the Flynn effect. Then in Norway they started falling for men born after 1975.[8] A 2018 study of Norwegian conscription records found both the rise and the fall inside families, brothers scoring differently by birth year. That points to the environment, not changing genes, as the cause.[9]

The history that cannot be skipped

Henry H. Goddard brought the French Binet test to America. By 1911 he had it in public schools, by 1913 he was testing immigrants at Ellis Island, and in 1914 he became the first psychologist to put Binet evidence before a court.[10] He also pushed sterilization and segregation of people he labeled feeble minded, in step with the eugenics movement. Psychologists later largely dismissed his work.[11]

A common story says IQ data helped pass the Immigration Act of 1924, which cut immigration from southern and eastern Europe. That link is disputed. Snyderman and Herrnstein argued in 1983 that the testing community did not generally back those restrictions and that Congress took virtually no notice of the test data.[12] Critics of that paper remain. What is not disputed is that the tests were used to label people, and that labels shaped how they were treated.

So what is in the number?

Something real. An IQ score captures a kind of reasoning that school and many jobs reward, and it predicts those outcomes well. It does not capture whether you stop to check your answer, whether you judge people and risks well, or what you do when no one is grading. Those are real too. They are just not on the test.

Sources

  1. Springer, Encyclopedia of Autism Spectrum Disorders, "Intelligence Quotient" (ratio IQ replaced by scores relative to age norms; mean 100, standard deviation 15; about 95 percent between 70 and 130)
  2. Tarmo Strenze, "Intelligence and socioeconomic success: A meta-analytic review of longitudinal research," Intelligence 35 (2007) 401 to 426 (intelligence a powerful predictor of education, occupation and income, but not overwhelmingly better than parental SES or school grades)
  3. A 2023 open access paper summarizing Strenze 2007 (correlation of childhood ability with educational attainment about 0.55, occupational attainment about 0.45, much weaker for income)
  4. Keith E. Stanovich and Richard F. West, "What intelligence tests miss," The Psychologist (British Psychological Society), Feb 15, 2014 (measures of rational decision making are not assessed on common IQ tests)
  5. Keith E. Stanovich, "Rational and Irrational Thought: The Thinking That IQ Tests Miss," Scientific American ("dysrationalia"; the bat and ball problem; the habit of checking a first answer is only weakly correlated with intelligence)
  6. Kautz, Heckman, Diris, ter Weel and Borghans, "Fostering and Measuring Skills," OECD Education Working Paper No. 110, Nov 2014 (IQ and achievement tests do not capture personality, goals, character and motivation, whose predictive power rivals or exceeds cognitive skills for many outcomes)
  7. University of Toronto Magazine, on Keith Stanovich (Charles Spearman proposed a general intelligence factor in 1904; Stanovich still sees what IQ tests measure as meaningful and useful)
  8. Bernt Bratsberg, "Norwegian IQ scores are falling, but genes are not to blame," The Science Breaker, March 20, 2019 (scores rose about 3 points per decade in the 20th century; Norwegian scores declined for cohorts born after 1975)
  9. Bratsberg and Rogeberg, "Flynn effect and its reversal are both environmentally caused," PNAS 115 (2018) 6674 to 6678 (Norwegian conscript data, cohorts 1962 to 1991; rise, turning point and decline all appear within families)
  10. Ludy T. Benjamin Jr., "The birth of American intelligence testing," Monitor on Psychology, Jan 2009 (Goddard brought the Binet test to the US, into schools by 1911, tested immigrants at Ellis Island by 1913, first used Binet evidence in court in 1914)
  11. Embryo Project Encyclopedia, Arizona State University, "Henry Herbert Goddard (1866 to 1957)" (advocated sterilization and segregation of people labeled feeble minded; work later largely dismissed by psychologists)
  12. Mark Snyderman and R. J. Herrnstein, "Intelligence Tests and the Immigration Act of 1924," American Psychologist 38 (1983) 986 to 995 (argues the testing community did not generally back the 1924 restrictions and Congress took virtually no notice of test data)

Inspired by Happy. Prepared by DeepSeek (two runs, self reported, version unknown; sources found by Claude). Reviewed and edited by Claude.

Next relatedDid Social Media Kill Our Attention Spans?
Revision history (R0 current)
RevDateWhat changedPage
D1Nov 20, 2026, 3:07:43 AM CST (logged on arrival; chat message 3:07 am, minute only)What DeepSeek returned, unedited (self report, version UNSURE, search on). Batch G prompt handed out 3:04:20 AM CST. Sent back: Google Translate proxy, OECD search page, catalog record as sources; markers past its own list.view
D2Nov 20, 2026, 3:19:09 AM CST (logged on arrival; chat message 3:18 am, minute only)The rerun, unedited. Batch H prompt handed out 3:15:57 AM CST, about 3 minutes round trip. Cut off before SOURCES; the rest never came, so Claude sourced the post itself.view
R0Nov 20, 2026, 10:33:14 AM CST (reviewed and posted)First publish, written from D2's outline and sourced by Claude (Opus): 12 sources found independently (Springer, Strenze 2007, Stanovich and West, the OECD working paper, Bratsberg and Rogeberg 2018, APA Monitor, ASU, Snyderman and Herrnstein 1983). Marked the IQ and 1924 Act link as disputed; added the Flynn reversal; cut the executive function, "speed of learning" and Ivy League bat and ball claims.view