— IQ reference guide

Why IQ scores come with confidence intervals

Standard error of measurement and why a single IQ score is a range, not a point.

This guide covers the published research, methodological considerations, and practical implications of the topic. It is intended as a reference for anyone seeking a grounded, evidence-based understanding rather than the popular-media version.

What the research actually shows

The published literature on this question spans clinical psychometrics, behavioral genetics, education research, and increasingly molecular genetics. Strong findings include: (1) substantial heritability of IQ in adulthood (~0.7-0.8 in twin studies); (2) the Flynn effect of rising measured IQ over the 20th century; (3) IQ's status as the single best general-purpose predictor of academic and complex-job performance; (4) the limited evidence for trainable IQ improvements beyond schooling and early-life nutrition.

The mechanisms

The biological substrate of IQ involves brain volume (modest correlation ~0.3), white matter integrity (DTI metrics correlate ~0.2-0.4), neural efficiency (lower energy consumption on cognitive tasks in higher-IQ subjects), and gene expression patterns identified through GWAS. The behavioral substrate involves working memory capacity, processing speed, and the ability to inhibit irrelevant information.

What this means in practice

For an individual, the practical implications depend on context: educators care about predicting school achievement; clinicians care about diagnostic categorization; employers care about job-performance prediction; researchers care about construct validity. Each use case interacts with IQ data differently, and the same score can be appropriately weighted differently across these contexts.

Common misconceptions

IQ is not destiny. Heritability estimates apply to populations, not individuals. A high-IQ child raised in deprivation can score below a low-IQ child raised with rich environmental support.
IQ is not fixed at birth. Early-life nutrition, schooling, and stimulation all shift measured IQ within meaningful ranges.
A single IQ score has substantial measurement error. Standard error is typically ±5 points on WAIS; a "true" IQ score is best understood as a range, not a point.
IQ measures only specific cognitive abilities. Creativity, wisdom, social cognition, and practical street-smarts are partially distinct from IQ — though usually correlated.

Get your own IQ — calibrated and decomposed across six axes

Core Brain returns a composite IQ plus per-axis scores in 20 minutes.

Practical implications

For most people, the most useful single fact about IQ is that it sets a soft ceiling on rate of acquisition for cognitively-demanding skills — but it does not determine ultimate achievement. Conscientiousness, deliberate practice, sustained focus, mentorship, and circumstance dominate the variance in real-world outcomes once you're past the threshold of IQ ~115. The 130+ tier opens additional doors but does not close any.

Frequently asked

Why IQ scores come with confidence intervals

Standard error of measurement and why a single IQ score is a range, not a point.

Is this based on real research?

Yes — the content draws on peer-reviewed cognitive psychology, psychometric, and behavioral genetics research. References include works by Carroll, Deary, Mackintosh, Plomin, and the broader CHC-theory tradition.

How can I use this information?

For self-understanding: read your own IQ score in light of the broader literature, not in isolation. For interpreting others: avoid using single IQ scores to make consequential decisions about people.

Related guides

Sources: Deary, I. J. (2001), Intelligence: A Very Short Introduction; Mackintosh, N. J. (2011), IQ and Human Intelligence; Plomin, R. (2018), Blueprint: How DNA Makes Us Who We Are.

Find your own IQ →