Average scores on intelligence tests have risen substantially across generations worldwide¶

Claim · All claims
Evidence · 3 studies · 3 review r? · q2–q3
Currency · newest study 1999 · not yet checked against its sources · not reviewed by a person · overdue for a recheck since 2004: baselines and conditions change

Known as the Flynn effect: raw scores on standardized intelligence tests have increased across successive generations in most countries where they have been tracked, at an average rate of roughly 3 IQ points per decade — a rate large enough that intelligence tests must be periodically re-standardized against a new population sample, or older norms will systematically overstate a test-taker's relative intelligence.

Subclaims¶

q3 i? Average raw scores on intelligence tests rose substantially across the 20th century in every country with adequate data, at an average rate of about 3 points per decade, though the size of the increase varies by country. → Flynn 1999

q2 i? Plausible contributors to the rise include better childhood nutrition, greater access to information and formal schooling, and increased familiarity with the standardized multiple-choice test format itself — but whether the rise reflects a genuine increase in underlying intelligence, rather than better test-taking skill or environmental advantage, remains an open, debated question. → Neisser 1998

Evidence¶

Flynn 1999¶

Flynn, J. R. (1999). Searching for justice: The discovery of IQ gains. American Psychologist, 54(1), 5-20. doi:10.1037/0003-066X.54.1.5

q3 · synthesis of standardization data across many countries · i? · the abstract prints no effect size; the full text may · n=very large (aggregated national standardization samples) · review · r?

Flynn's synthesis of intelligence-test standardization data from numerous countries found a consistent pattern: each successive generation's raw scores outperform the previous generation's on the same test items, at a remarkably stable average rate across very different populations and testing traditions, forcing test publishers to periodically renorm tests against a new representative sample to keep IQ=100 meaningful as "average for this population, now."

Neisser 1998¶

Neisser, U. (Ed.). (1998). The rising curve: Long-term gains in IQ and related measures. American Psychological Association.

q2 · edited volume synthesizing multiple research programs · i? · no source text available to check; the entry prints no effect size · n=large (aggregated across contributing studies) · review · r?

This edited volume brings together researchers proposing and evaluating competing explanations for the Flynn effect — improved nutrition, expanded schooling, greater exposure to abstract visual reasoning tasks in everyday life (television, video games), and growing test-taking sophistication — without settling on a single confirmed mechanism.

Neisser 1997¶

Neisser, U. (1997). Rising scores on intelligence tests. American Scientist, 85(5), 440-447.

q2 · review/commentary · i? · no source text available to check; the entry prints no effect size · n=not applicable (synthesis piece) · review · r?

Neisser argues explicitly that a rising test score is not proof of rising underlying intelligence: the gains are concentrated on certain test subtypes (particularly abstract, non-verbal reasoning items) rather than uniformly across all measured abilities, which is more consistent with a shift in test-relevant skills and exposure than with a genuine, general increase in cognitive capacity.

Discussion¶

The Flynn effect is one of the best-replicated findings in intelligence research, but its interpretation remains unsettled — the same data can be read either as evidence that humanity is measurably "getting smarter," or as evidence that the tests are increasingly measuring test-specific skills (familiarity with abstract-reasoning formats, multiple-choice conventions) rather than a stable underlying construct. Practically, the clearest implication for anyone using standardized intelligence or ability tests is that norms decay: a test standardized decades ago will make today's test-takers look smarter than they would against a contemporary sample, which is why periodic re-standardization (as with the WAIS-IV, WISC-V, and similar instruments) is a scientific necessity rather than a formality.

Related Claims¶

• A century of psychometric data consistently shows performances on intellectual tests are correlated, which MI must account for — related
• US cohort achievement trends per decade are steeper in math (median 0.19sd) than in reading (median 0.04sd) across linked surveys — related