The science of intelligence
Intelligence across the lifespan
Short answer
Intelligence is fairly stable in rank across life. In a Scottish group who took the same test at age 11 and again at 90, the two scores correlated .54, or .67 after a statistical correction. Separate abilities still change: processing speed peaks in the late teens and twenties, while vocabulary keeps rising into the sixties.
Want your own number? The test is 25 visual puzzles, about 12 minutes, no sign-up.
Two different questions: rank and level
"Does intelligence change with age?" can mean two things. One is rank: does someone who scores above their peers at 11 still score above their peers at 90? The other is level: can a person do more or less on a given task than they could before? The answers differ, and mixing them up causes most of the confusion. The IQ and age guide covers how age norms work; this page focuses on what long-term studies found.
Rank: the 80-year study
Scotland ran an unusual experiment. The Scottish Mental Surveys tested almost everyone born in 1921 and 1936 who was at school on the survey days, 89,498 children in 1932 and 70,805 in 1947 (Deary et al., 2004). Decades later, researchers found some of them again.
In the Lothian Birth Cohort 1921 (N = 106), people took the same intelligence test at age 11 and at age 90. Their scores correlated .54, or .67 after correcting for range restriction (Deary et al., 2013). Range restriction means a sample is less varied than the full population, which pulls a correlation down; the correction estimates what the correlation would be without it.
How strong is that? Squaring the correlation gives the share of variation in common: about 29% for .54 and about 45% for .67. Childhood scores carry real information about old age, yet most of the variation at 90 is not explained by the score at 11. That fits the 1994 mainstream statement that IQs gradually stabilize during childhood and generally change little thereafter (Gottfredson, 1997), while leaving room for many individual exceptions.
Try one · Pattern Recognition
Which tile comes next?
Tap the option you think is right, or show the answer.
Show the answer
The missing tile shows a very large solid arrow pointing up and to the left (option 3).
How to solve it
- Compare neighboring tiles and name what changes each time.
- From tile to tile, the arrow turns an eighth of a turn clockwise.
- From tile to tile, the arrow gets bigger.
- Everything else stays the same from tile to tile.
- The last tile shows a large solid arrow pointing left. Apply each change once more.
- So the missing tile shows a very large solid arrow pointing up and to the left.
That was 1 puzzle. The full test has 25.
Level: abilities peak at different ages
Within that stable rank, specific abilities follow their own paths. Processing speed and some reasoning measures peak in the late teens and twenties, while vocabulary and accumulated knowledge keep rising into the sixties (Hartshorne & Germine, 2015) (Salthouse, 2009). This lines up with the split between fluid and crystallized ability described in the fluid vs crystallized guide.
| Ability | Typical pattern in adulthood |
|---|---|
| Processing speed | Peaks in the late teens and twenties |
| Some reasoning measures | Peak in the late teens and twenties |
| Vocabulary and knowledge | Keep rising into the sixties |
Why speed matters so much
Timothy Salthouse's processing-speed theory proposes that increased age in adulthood is associated with slower execution of many mental operations, and that this slowing impairs cognitive functioning (Salthouse, 1996). It names two mechanisms: limited time, where operations cannot finish, and simultaneity, where early results are gone by the time later processing completes (Salthouse, 1996). A key piece of support is that age differences on many cognitive measures shrink a lot once speed is statistically controlled (Salthouse, 1996). The processing speed page goes further.
Childhood and adolescence
In childhood, the genetic side shifts too. The heritability of general cognitive ability rises from 41% at age 9 to 66% at age 17 (Haworth et al., 2010), as children increasingly shape their own experiences. Schooling also matters at every stage: education effects on cognitive ability persisted across the life span (Ritchie & Tucker-Drob, 2018).
What this means for a single score
Because IQ is age-normed, a typical decline in raw speed with age does not lower your IQ score by itself; the score is relative to same-age peers. For answers to the common follow-ups, see at what age does intelligence peak? and does IQ decline with age? Stability over decades also explains why childhood scores turn up in health research, covered in intelligence and health.
Sources
- Deary, I. J., Pattie, A., & Starr, J. M. (2013). The stability of intelligence from age 11 to age 90 years: The Lothian Birth Cohort of 1921. Psychological Science, 24(12), 2361-2368. doi.org/10.1177/0956797613486487
- 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. doi.org/10.1037/0022-3514.86.1.130
- Gottfredson, L. S. (1997). Mainstream science on intelligence: An editorial with 52 signatories, history, and bibliography. Intelligence, 24(1), 13-23. doi.org/10.1016/S0160-2896(97)90011-8
- Hartshorne, J. K., & Germine, L. T. (2015). Psychological Science, 26(4), 433-443. doi.org/10.1177/0956797614567339
- Salthouse, T. A. (2009). Neurobiology of Aging, 30(4), 507-514.
- Salthouse, T. A. (1996). The processing-speed theory of adult age differences in cognition. Psychological Review, 103(3), 403-428. doi.org/10.1037/0033-295X.103.3.403
- Ritchie, S. J., & Tucker-Drob, E. M. (2018). How much does education improve intelligence? A meta-analysis. Psychological Science, 29(8), 1358-1369. doi.org/10.1177/0956797618774253
- Haworth, C. M. A., et al. (2010). The heritability of general cognitive ability increases linearly from childhood to young adulthood. Molecular Psychiatry, 15(11), 1112-1120. doi.org/10.1038/mp.2009.55