What IQ actually predicts

Last reviewed 9 September 2026

How to read a correlation

Everything below is reported as a correlation, so it is worth spending sixty seconds on what those numbers mean - because this is where almost all misreading happens.

A correlation of r = 0.5 sounds like "half". It is not. Square it and you get the share of variance accounted for: 0.25, or a quarter. Three quarters of the differences between people on the outcome are down to something else.

Keep that table in mind. A correlation strong enough to be one of the most robust findings in psychology can still leave you unable to say much about the person in front of you.

School results

This is where IQ predicts best, which is unsurprising: IQ tests were invented to predict school performance, and both tasks reward the same abstract, verbal, quantitative reasoning.

The landmark study followed more than 70,000 English children from an ability test at age 11 to their national exam results at 16. The correlation between the latent ability factor and exam performance was 0.81.¹ That is an unusually strong result for psychology - roughly two thirds of the variance in exam outcomes.

Two caveats stop that being the whole story. It is a correlation between latent factors, which are statistically cleaned of measurement error and so run higher than what you would see between two raw test scores. And it was measured within a single national curriculum, where what is examined and what the test samples overlap heavily by design.

Even at 0.81, the study found plenty of children whose results departed sharply from prediction in both directions.

Job performance

For decades the standard reference was a 1998 meta-analysis reporting that general mental ability predicted job performance at around r = 0.51, better than any other single selection method.² That figure was repeated so often it became folklore.

In 2022 it was substantially revised. A re-analysis argued that the earlier corrections for range restriction had been applied too aggressively, inflating validity estimates across the board. With more defensible corrections, cognitive ability predicts job performance at roughly r = 0.31 - still useful, still among the better predictors, but around 10% of variance rather than 26%.³

Validity also varies by job complexity. Cognitive ability matters more for work involving novel problems and less for highly routinised work - which is intuitive, and consistent with fluid reasoning being about handling situations you have not memorised a response to.

Income and status

A meta-analysis of longitudinal studies - measuring ability first and outcomes years later - put the correlation between intelligence and later income at about 0.23, with occupational status around 0.37 and educational attainment around 0.56.⁴

So the causal chain looks less like "clever people earn more" and more like "ability predicts how far you get in education, education predicts what job you can enter, and the job predicts the pay". Each link loses information.

At an income correlation of 0.23, intelligence accounts for about 5% of the variation in what people earn. Parental background, field of work, geography, negotiation, timing and luck occupy most of the rest. Anyone selling IQ as the key to financial success is misreading their own evidence.

Health and lifespan

The most surprising finding in this literature is that childhood intelligence predicts how long you live. A systematic review pooling 16 studies found that a one standard deviation advantage in youth intelligence - 15 IQ points - was associated with roughly a 24% lower risk of death over follow-up periods of up to 69 years.⁵

This line of work owes much to an unusual resource: Scotland tested almost every eleven-year-old in the country in 1932 and again in 1947, and researchers later traced those same people into old age.⁶

Why the link exists is genuinely unsettled. Candidate explanations include health literacy and treatment adherence, safer and less physically damaging occupations, socioeconomic advantage acting on both, and the possibility that test performance partly indexes general bodily integrity. These are not mutually exclusive, and the evidence does not currently pick between them.

What it does not tell you

Pulling the numbers together:

The pattern is consistent: IQ predicts best where the outcome most resembles the test, and the prediction weakens steadily as you move toward things people actually care about.

Three limits are worth stating plainly.

• Group averages are not individual forecasts. Every correlation above is compatible with enormous numbers of individual exceptions, and the exceptions are not rare curiosities - at r=0.3 they are most of the distribution.
• Correlation is not mechanism. Ability, schooling and background are entangled, and observational studies mostly cannot separate them.
• The tests do not sample everything. Persistence, judgement under uncertainty, creativity, social skill and simply choosing well are largely outside what an IQ test looks at, and they are not obviously less consequential.⁷⁸

A score from this or any other test is a reading on one narrow instrument. It is genuinely informative at the level of populations. It is a weak basis for any conclusion about a person, including yourself.

Common questions

References

1. Deary, I. J., Strand, S., Smith, P., & Fernandes, C. (2007). Intelligence and educational achievement. Intelligence, 35(1), 13-21. doi:10.1016/j.intell.2006.02.001
2. Schmidt, F. L., & Hunter, J. E. (1998). The validity and utility of selection methods in personnel psychology. Psychological Bulletin, 124(2), 262-274.
3. Sackett, P. R., Zhang, C., Berry, C. M., & Lievens, F. (2022). Revisiting meta-analytic estimates of validity in personnel selection. Journal of Applied Psychology, 107(11), 2040-2068. doi:10.1037/apl0000994
4. Strenze, T. (2007). Intelligence and socioeconomic success: A meta-analytic review of longitudinal research. Intelligence, 35(5), 401-426. doi:10.1016/j.intell.2006.09.004
5. Calvin, C. M., Deary, I. J., Fenton, C., Roberts, B. A., Der, G., Leckenby, N., & Batty, G. D. (2011). Intelligence in youth and all-cause mortality: systematic review with meta-analysis. International Journal of Epidemiology, 40(3), 626-644. doi:10.1093/ije/dyq190
6. Deary, I. J., Whiteman, M. C., Starr, J. M., Whalley, L. J., & Fox, H. C. (2004). The impact of childhood intelligence on later life: Following up the Scottish Mental Surveys of 1932 and 1947. Journal of Personality and Social Psychology, 86(1), 130-147.
7. Neisser, U., Boodoo, G., Bouchard, T. J., et al. (1996). Intelligence: Knowns and unknowns. American Psychologist, 51(2), 77-101.
8. Nisbett, R. E., Aronson, J., Blair, C., Dickens, W., Flynn, J., Halpern, D. F., & Turkheimer, E. (2012). Intelligence: New findings and theoretical developments. American Psychologist, 67(2), 130-159. doi:10.1037/a0026699

Related reading