The science behind IQ tests
A short, honest tour of what cognitive testing actually measures — and what it does not.
A century of converging evidence
Intelligence testing began with Alfred Binet and Théodore Simon in 1905 Paris, as a school-screening tool. Since then, the field has accumulated more empirical replications than almost any other area of psychology. The central finding — that performance on cognitive tasks tends to correlate positively across domains, producing a single 'general factor' often called g — has held up across cultures, decades, and dozens of distinct instruments.
What the score predicts
Composite IQ scores are among the strongest single predictors of academic achievement, occupational training success, and several health outcomes. The effect sizes are large at the population level and modest at the individual level, which is the typical signature of a real but probabilistic relationship. A higher score makes a wide range of outcomes more likely; it does not determine them.
Fluid and crystallized intelligence
Modern theory distinguishes fluid intelligence (the ability to reason about novel material, peaking in the early twenties and declining slowly thereafter) from crystallized intelligence (accumulated knowledge and vocabulary, which keeps growing into the seventies in healthy adults). Most IQ tests, including ours, sample both. The five categories on this test were chosen specifically to cover the fluid-crystallized spectrum.
What IQ tests do not measure
Cognitive tests do not measure motivation, conscientiousness, social skill, creativity, ethical judgment, or wisdom. They do not measure your capacity for love, your taste in art, your sense of humor, or your willingness to take a risk. These are independently important traits, and the IQ literature has been clear for decades that they are not subsumed by g. A test score is a narrow signal about a narrow construct.
The Flynn effect and its slowdown
Average IQ scores in industrialized countries rose by roughly three points per decade for most of the 20th century — the so-called Flynn effect. The rise has slowed and in some samples reversed since the 1990s. The dominant explanation is environmental: better nutrition, more schooling, and richer symbolic environments push fluid scores up; their plateau and recent decline likely reflect the maturation of those environmental gains.
Heritability and what it means
Heritability estimates for adult IQ in stable Western populations cluster around 0.5 to 0.8. This number is widely misunderstood. Heritability is a population-level statistic that describes the share of variance in a trait that tracks variance in genes, in a particular environment. It does not mean a given person's score is 'mostly genetic,' and it does not place a hard ceiling on environmental intervention. The same trait can be highly heritable and highly responsive to environment, simultaneously.