People can change substantially while their relative standing compared with others becomes increasingly stable. Those are not contradictory facts.
A six-year-old who scores above most children of the same age is not carrying a fixed quantity of intelligence that will remain unchanged until old age. Their brain will develop, their knowledge will expand, their strategies will change and the tests appropriate for them will change too.
Yet something else can also be true: as people grow older, individual differences in cognitive ability become remarkably stable. People who perform relatively strongly compared with their peers at one point are increasingly likely to perform relatively strongly years later.
The apparent contradiction disappears once we separate stability of rank from stability of level.
Imagine a moving race
Suppose five children complete age-appropriate reasoning assessments at age 8 and again at age 16. All five can become capable of solving problems at 16 that none could solve at 8. Their absolute cognitive performance has changed enormously.
But if the child who ranked first at 8 tends still to be near the top at 16, and the child who ranked fifth tends still to be lower relative to this same peer group, their rank order is stable.
Psychologists often quantify this with a test–retest or longitudinal correlation. A high correlation does not mean scores are literally unchanged. It means relative differences are preserved to a substantial degree.
This distinction is crucial because standardized IQ scores are age-normed. A score near 100 at age 10 and again at age 20 does not mean the person solved the same problems at the same level. It means their performance was near the average of the relevant age-norm group at both times.
Stability grows through childhood
The newest large meta-analytic synthesis of longitudinal cognitive-ability studies reinforces a long-standing pattern: rank-order stability is lower early in childhood and rises with age, becoming high in adulthood. Longer intervals between measurements reduce stability, although the decline is not indefinite.
Why should childhood be less stable? Development is rapid, measurement is harder in very young children, and differences in language, attention, schooling and developmental timing can have large effects. As cognitive systems and environments become more established, relative differences become more predictable.
This does not imply that adulthood is cognitively frozen. It means that large reshufflings of people’s relative positions become less common.
What happens in older age?
Average trajectories differ across cognitive abilities. Some aspects of processing speed and fluid reasoning tend to peak earlier and decline across adulthood, while accumulated knowledge can remain stable or increase much longer before late-life decline. Individuals vary widely around these averages.
Again, average change and rank-order stability can coexist. Imagine an entire group slowing somewhat over ten years while the faster members of the group tend to remain faster than the slower members. The mean has changed; the ordering remains similar.
Can an individual’s IQ score change?
Yes. A score is an estimate, not a permanent label. Measurement error alone means repeated scores will not be identical. Education, health, neurological events, development, practice effects, language, testing conditions and other influences can contribute to change.
Additional schooling, for example, has causal effects on cognitive-test performance in quasi-experimental evidence. Severe illness or brain injury can alter cognitive functioning. Childhood interventions can produce changes, although the durability and breadth of transfer vary greatly by intervention.
What high population-level stability tells us is not that change is impossible. It tells us that large enduring changes in relative standing are less common than casual talk about “reinventing your IQ” suggests, especially after childhood.
What can one person’s change tell us?
Less than it first appears. Suppose a standardized score rises from 105 to 112. Before treating seven points as a real change in underlying ability, we would want to know the reliability of the instrument, its standard error of measurement, whether the same test was repeated, the interval, practice effects, the person’s age and the confidence intervals around both scores.
Psychological measurement is probabilistic. The meaningful question is not simply “did the number move?” but whether the observed difference is larger than we would reasonably expect from measurement and testing conditions.
Stable does not mean immutable
Intelligence is often described as highly stable because, compared with many psychological characteristics, relative differences in cognitive ability are indeed persistent. But stability is a description of longitudinal patterns, not a theory that nothing can change them.
A person can learn enormously while remaining at a similar percentile. A population can show age-related decline while preserving much of its rank order. An individual can experience meaningful change even when the population correlation is high.
Once those levels are separated, the lifespan picture becomes clearer: intelligence develops, different abilities follow different average trajectories, and individual differences become increasingly—but never perfectly—stable.
References
- Breuer, J., et al. (2024). The stability of cognitive abilities: A meta-analytic review of longitudinal studies. Psychological Bulletin. DOI
- Deary, I. J., Pattie, A., & Starr, J. M. (2013). The stability of intelligence from age 11 to age 90 years. Psychological Science, 24, 2361–2368. DOI
- Tucker-Drob, E. M., & Briley, D. A. (2014). Continuity of genetic and environmental influences on cognition across the life span. Psychological Bulletin, 140, 949–979. DOI
- Ritchie, S. J., & Tucker-Drob, E. M. (2018). How much does education improve intelligence? A meta-analysis. Psychological Science, 29, 1358–1369. —. DOI
