Can Intelligence Change? Development, Learning, and Experience
Author: Angela Bogdanova · Published: October 12, 2026 · Editorial Policy
Research cluster: Intelligence, Reasoning & Cognitive Abilities · Written by Artificial Sapiens
Intelligence can change across development, education, and experience. Cognitive abilities mature at different rates, people acquire sophisticated knowledge throughout life, and schooling can increase performance on standardized intelligence tests. At the same time, broad cognitive differences become relatively stable from adolescence into adulthood. Understanding both findings requires distinguishing developmental growth, the acquisition of specific skills, changes in latent abilities, and changes in age-normed IQ scores.
This article examines the evidence on what changes, when, why, and how scientists know. The two most relevant large syntheses are a 2024 meta-analysis of longitudinal cognitive stability and a 2018 meta-analysis of quasi-experimental educational effects. Together they show why intelligence can be both developmentally responsive and strongly predictive over time.
What does it mean for intelligence to change?
Intelligence is a set of psychometric constructs describing individual differences in reasoning, learning, knowledge, and related cognitive performances. General intelligence, or g, refers to the common statistical factor that helps explain why people who perform well on one cognitive test often perform well on others. Broad abilities, such as fluid reasoning and acquired knowledge, describe related but distinguishable domains. IQ is an age-normed test score or composite estimate produced by a particular standardized instrument; it is not a direct physical measurement of g.
Several kinds of change require separate definitions. Developmental growth means an increase in what a person can do as they mature. Knowledge acquisition means expanding vocabulary, concepts, and expertise. Task improvement means doing better on a practiced activity. Broad ability change means a shift detectable across multiple appropriately chosen measures of a cognitive domain. IQ change means a shift in standardized score, which may include genuine cognitive change alongside measurement error, practice, or contextual influences. Rank-order change means moving relative to other people in the reference group.
These distinctions answer a common puzzle: a child can reason far more effectively at age twelve than at age eight while receiving the same age-standardized IQ score at both ages. Their raw performance improved, but peers improved as well. An adult can become expert in a complex profession without showing a large shift in general cognitive test scores. Both forms of learning matter.
How intelligence changes throughout life
Early childhood: major growth and uncertain long-term ranking
Childhood is a period of rapid growth in language, working knowledge, mental representations, and the capacity to coordinate ideas and rules. Development reflects maturation alongside responsive relationships, educational opportunities, health, and experience. A child's score is therefore an assessment of functioning in a particular context and developmental period, not a declaration of a lifelong ceiling.
Early rank-order stability is lower than later stability. In an analysis of 205 longitudinal studies involving 87,408 people, Breit and colleagues found that correlations between repeated cognitive measurements increase substantially during childhood and level off around late adolescence. This does not mean all young children's results fluctuate widely; rather, predictions based on early scores carry greater uncertainty, especially over long intervals.
For children, the language of the assessment, opportunities to learn tested material, sensory needs, schooling history, and appropriate norming can all affect interpretation. A sound assessment examines a pattern of abilities and the reason for testing. Neither a high nor a low early score should be treated as a complete description of future learning capacity.
Adolescence: meaningful changes in particular abilities
Adolescents continue developing abstract reasoning, verbal comprehension, decision strategies, and school-related knowledge. Their experiences become increasingly differentiated: some receive advanced mathematical instruction, others develop linguistic or technical expertise, and others face educational disruption. Development is neither uniform nor reducible to one IQ trajectory.
A longitudinal neuroimaging study of 33 adolescents reported increases and decreases in verbal and nonverbal IQ relative to peers, with associated changes in local gray-matter measurements. This provides evidence that observed adolescent change can track development in related brain systems. The sample was small, and the correlations do not prove that a specific educational experience caused a specific neural or IQ change.
The central practical lesson is that adolescent abilities should be supported through meaningful opportunities and reassessed when decisions require updated evidence. Performance at one age provides information; it does not settle every later developmental outcome.
Adulthood: stability of relative standing and continued learning
Adults can acquire new languages, scientific knowledge, professional expertise, and complex skills. Learning persists throughout adulthood, and people can adapt strategies to changing demands. Improvement in these accomplishments should be evaluated in their own domains.
Longitudinal research nevertheless finds comparatively high rank-order stability of broad cognitive ability among adults. The 2024 meta-analysis estimated a mean observed five-year stability correlation of about .76 for age twenty. That correlation describes a relationship across people; it does not mean that 76% of one person's intelligence is permanently fixed.
Even across extraordinary intervals, continuity and change coexist. In the Lothian Birth Cohort of 1921, the same test was administered to 106 individuals at ages eleven and ninety, producing a correlation of .54, or .67 corrected for range restriction. These results support substantial long-term continuity while leaving scope for differences in life-course development. They do not identify what caused each participant's change.
Older adulthood: different abilities peak at different ages
A single claim that intelligence peaks at age twenty-five, forty, or sixty misunderstands the variety of cognitive abilities. Some speeded abilities reach their highest typical performance earlier; vocabulary, accumulated knowledge, and other developed capacities can continue improving much later. In research combining 48,537 online participants with standardized-test norms, Hartshorne and Germine documented markedly different peak ages for different tasks.
Age-related patterns are also influenced by cohort, education, task familiarity, health, and individual history. Declines in a timed measure cannot be generalized to all judgment, knowledge, or reasoning capabilities. Sudden or disabling changes call for clinical evaluation rather than assumptions about ordinary aging.
Stability and change are different statistical questions
Mean-level change asks whether average performance in a group rises or falls. Rank-order stability asks whether individuals maintain similar relative positions. Individual change asks whether a particular person's ability estimate has shifted. Structural stability asks whether the relationships among multiple cognitive abilities remain similar across time.
A population can experience strong raw-score growth while maintaining rank-order stability. Conversely, the group's average age-normed score may remain around 100 while individual scores change. A stable IQ score also does not imply unchanged knowledge or identical performance on every subtest.
Establishing individual change requires knowledge of the precise test, reliability, standard error of measurement, practice effects, retest interval, norm group, test version, and relevant testing conditions. A few IQ points between occasions can fall within expected measurement variation. A difference should be evaluated using suitable confidence intervals or reliable-change methods rather than an arbitrary universal cutoff.
The 2024 longitudinal meta-analysis found that stability usually decreases as intervals between measurements increase. General and knowledge-based cognitive abilities also tended to be somewhat more stable than several more effortful processing abilities. These are statistical tendencies, not immutable rules governing an individual.
Does education actually increase measured intelligence?
Education is among the best-supported environmental influences on intelligence-test outcomes. Schooling develops verbal and quantitative knowledge, introduces abstract systems of representation, and creates repeated opportunities for structured problem solving.
Ritchie and Tucker-Drob's 2018 meta-analysis combined 142 estimates from 42 data sets and more than 600,000 participants. The studies used three quasi-experimental approaches: controlling for earlier intelligence, exploiting compulsory-schooling reforms, and examining school-entry cutoff rules. Additional education yielded estimated gains of about one to five IQ points per year, with positive effects across broad ability categories in the studied designs.
This does not mean that every additional school year increases every student's intelligence by an exact amount. The range reflects variation in design and settings; the effect cannot be extrapolated across unlimited schooling. Quality, access, curriculum, developmental stage, and the kinds of abilities tested all matter. What the evidence supports is that formal education can causally influence performance on cognitive measures rather than serving solely as a consequence of prior cognitive ability.
There remains a theoretical distinction between increasing performance on a battery of intelligence measures and changing the latent g factor. Studies using IQ composites establish changes in those measured outcomes. Stronger claims about specific latent mechanisms require additional psychometric evidence and appropriately identified models.
What kinds of experience shape cognitive development?
Early environments and opportunities
A child's environment influences opportunities for language development, exploration, conceptual learning, and access to sustained instruction. Adversity, disruptions in schooling, and restricted educational resources can alter developmental conditions. The mechanisms are intertwined and vary across settings.
A meta-analysis of 62 adoption studies involving 17,767 children reported higher average IQ scores and better academic outcomes among adoptees compared with children remaining in the earlier environments represented by those studies. Adoption changes many circumstances at once, and selection effects complicate causal attribution. The appropriate conclusion is that developmental context matters, not that one isolated environmental ingredient determines intelligence.
Knowledge, expertise, and practice
Experience improves knowledge and performance in the areas in which it supplies instruction, feedback, and meaningful practice. Learning mathematics builds concepts and solution procedures; learning a language expands linguistic competence; mastering a profession builds pattern recognition and judgment specific to its demands.
The ability to transfer such learning to unrelated tasks must be demonstrated, not assumed. Someone who becomes faster at solving familiar puzzles may have learned the puzzle's structure. That is a real cognitive achievement, but it does not establish a broad increase in fluid intelligence without testing on appropriately different material.
Language, culture, and unequal opportunity
Intelligence assessments operate within languages, cultures, and educational histories. Vocabulary and some problem-solving formats depend on past exposure. Even tests described as relatively language-reduced require attention to instructions, norms, and measurement equivalence.
Observed differences across social groups do not identify their causal origins. Responsible interpretation considers access to schooling, health, test validity, and the conditions under which scores were obtained. Scores should never be used to infer an individual's capacity solely from demographic membership.
Does heritability mean that intelligence is predetermined?
Heritability describes how much of the observed variation in a specified population, under specified circumstances, is statistically associated with genetic variation. It does not tell an individual what fraction of their intelligence is genetic, nor does it measure how much an ability can improve.
In a large twin study spanning several countries, Haworth and colleagues estimated heritability of general cognitive ability at approximately 41% at age nine, 55% at age twelve, and 66% at age seventeen. These results describe the samples studied and depend on model assumptions and environments. They are not universal values for all societies or eras.
Development involves gene–environment interplay: inherited differences may influence patterns of interest and experience, while education and other environments affect the expression and measurement of abilities. A characteristic can be highly heritable within a population and still respond to environmental change. Within-population heritability estimates also cannot independently explain differences between populations.
Can an IQ score change even if underlying ability does not?
Yes. Retesting can increase familiarity with a test's structure and reduce uncertainty about instructions. Fatigue, distress, sensory problems, time pressure, testing language, and temporary health conditions can affect how well an individual demonstrates ordinary abilities. Measurement error can produce different point estimates even when the underlying construct remains similar.
Many modern IQ batteries use standardized norms with a mean of 100 and standard deviation of 15, but these are properties of a norm-referenced scoring system. An IQ value is interpreted relative to the test's population and norming date. Scores from different instruments, versions, and years are not necessarily interchangeable.
A higher result after practicing similar questions is therefore evidence of higher test performance under the new conditions. Whether it reflects enduring changes in a broad ability, increased familiarity, or other influences requires careful evaluation.
What the Flynn effect tells us about change across generations
The Flynn effect refers to historical increases in average intelligence-test performance across successive birth cohorts in many settings. It is a population-level phenomenon, not an account of how one person's intelligence changes with age.
A 2015 meta-analysis covering nearly four million participants across 31 countries found substantial historical gains with differences across ability domains and time periods. Patterns have not been uniform across places or decades, and the causal mixture behind those changes remains debated.
Generational score shifts show that measured cognitive performance is sensitive to large-scale historical conditions. They do not establish that every new generation is superior at every intellectual activity, or that schooling alone explains the pattern. Changing norms are one reason longitudinal interpretation and test standardization matter.
Can brain training, memory games, or apps raise IQ?
It is essential to distinguish three outcomes. A trained-task gain is better performance on an exercise that was practiced. Near transfer is improvement on similar but untrained tasks. Far transfer is improvement in substantially different abilities or settings, such as general reasoning or academic achievement.
A 2016 meta-analysis of working-memory training found improvements on trained and related memory measures but no convincing far-transfer advantage in intelligence, reading, or mathematics when appropriate active controls were used. A 2024 meta-analysis reported small working-memory gains, much larger effects when training and assessment tasks resembled each other, and no reliable improvement in fluid intelligence.
These conclusions concern the programs and studies evaluated. They do not imply that all cognitive practice is useless or that future interventions can never succeed. Cognitive games can be engaging and can improve game-specific proficiency. Broad advertising claims that an app raises general intelligence, however, require evidence from independent tasks, controlled comparisons, and durable follow-up.
Training programs should be judged by the actual outcome they promise. If the goal is arithmetic fluency, measure arithmetic. If the goal is workplace performance, measure relevant functioning. If the claim is a change in general intelligence, improved scores on the same practiced puzzles are inadequate evidence.
Can growth mindset change intelligence?
Growth mindset refers to the belief that abilities can develop. Such beliefs may influence people's responses to difficulty, feedback, and learning. Changing a belief, improving classroom achievement, and changing measured general intelligence are distinct endpoints.
Two 2018 meta-analyses found weak overall associations and intervention effects on academic achievement, with possible advantages in particular at-risk contexts. A 2023 systematic review and meta-analysis questioned the strength and quality of evidence for broad achievement benefits of mindset interventions. These findings do not establish a reliable IQ increase from mindset instruction.
A constructive learning environment combines appropriate challenge with effective teaching, feedback, adequate resources, and support. Encouraging persistence can be useful; it does not replace access to instruction or solve structural barriers by itself.
Sleep, exercise, health, and everyday cognitive performance
Sleep restriction can impair attention, memory, and executive performance. A meta-analysis of experimentally restricted sleep documented negative effects across several cognitive domains while finding insufficient evidence for an intelligence-specific conclusion. Restoring sleep may improve current cognitive functioning without implying a durable increase in g.
Physical activity contributes to health, and some studies report benefits on selected cognitive measures. A 2024 review of longer exercise and multidomain trials in older adults found modest average benefits for multidomain interventions and little evidence of global cognitive gains from long-term exercise alone. These conclusions should not be translated into universal claims that exercise raises IQ.
Health conditions, medications, acute illness, sensory difficulties, and psychological distress can alter performance under testing conditions. If change is sudden, progressive, or disrupts daily functioning, its causes should be professionally evaluated. A change in test performance alone does not establish a clinical diagnosis.
How researchers test whether intelligence has genuinely increased
A strong study first identifies the intended outcome: practiced skill, near transfer, broad cognitive ability, IQ composite, or g. It then selects tests that can measure the intended construct and separates observed-score changes from the interpretation of underlying ability.
When feasible, a randomized controlled design uses an appropriate active comparison, independently administered outcome measures, preregistered analyses, sufficient statistical power, and long-term follow-up. Researchers should assess practice effects, attrition, expectancy, selective reporting, and whether gains remain when different test items and formats are used.
Evidence is stronger when independent research teams find benefits across multiple tasks that are not simple variants of the training activity. Durable changes in daily functioning or educational outcomes are valuable but should be measured rather than inferred from laboratory performance alone.
Quasi-experimental methods can also support causal conclusions when randomized schooling assignments would be impractical or unethical. The education literature uses policy reforms and entry cutoffs to separate some environmental effects from selection. All such inferences have assumptions and limits that must be considered.
What can people do with this evidence?
For children, development is best supported through educational access, language-rich interaction, appropriate learning opportunities, and prompt attention to barriers. Repeated assessment may be helpful when it is linked to a concrete educational or clinical question, but scores should be interpreted alongside a child's wider functioning.
For students and adolescents, the most actionable objective is to learn the desired content deeply and develop relevant reasoning skills. Structured instruction, meaningful practice, and feedback provide direct benefits. They are more defensible goals than a promise of guaranteed IQ-point gains.
For adults, acquiring a language, changing careers, studying technical subjects, and developing expertise can improve competence and everyday adaptability. Their value is substantive even when no increase in a global intelligence estimate is demonstrated.
For those comparing old and new test results, a qualified assessor can evaluate score confidence intervals, test differences, norming dates, retest effects, and contextual influences. A single numerical difference should not determine a person's identity, worth, educational expectations, or clinical status.
Misconceptions about intelligence change
“IQ is completely fixed after childhood.”
Measured cognitive abilities show increasing stability, but individual changes occur, and education can affect cognitive-test performance. Stability describes a statistical pattern, not a permanent limit.
“If my memory-game score improved, my intelligence improved.”
Practice can improve the practiced skill substantially. Broad intelligence change requires demonstrated transfer to appropriately different cognitive tasks.
“Intelligence is genetic, so experience cannot affect it.”
Heritability concerns variation within a population in particular circumstances. It does not prohibit developmental or environmental effects.
“A difference of five IQ points always proves a real change.”
The interpretation depends on the test's reliability, standard error, norms, retest interval, and testing conditions. Meaningful change should be evaluated psychometrically.
“All intellectual abilities decline after one peak age.”
Different abilities reach typical peaks at different points in life. Knowledge and expertise may increase while other processes slow.
Frequently asked questions
Can intelligence improve with age?
Yes. Knowledge, strategies, and expertise can improve throughout life. Broad cognitive differences also exhibit substantial adult stability, so improvement should be specified in terms of the ability measured.
Can a child's IQ increase or decrease?
Yes. Children's age-normed scores can change, and long-term stability is lower early in development. Growth in raw cognitive performance is expected even when standardized IQ remains similar.
Does school make you smarter?
Quasi-experimental evidence supports an average positive effect of additional education on intelligence-test performance. The estimate of approximately one to five IQ points per additional year is a research finding across different settings, not a guaranteed individual result.
Can an adult raise their IQ?
Some experiences, notably formal education in studied contexts, can influence cognitive-test results. There is no comparably persuasive evidence that commercial brain-training apps reliably generate broad IQ gains.
Does learning an instrument or foreign language increase general intelligence?
Those activities can build valuable specialized knowledge and skills. Broad generalization to intelligence requires separate controlled transfer evidence and should not be assumed.
Does growth mindset raise IQ?
Mindset interventions target beliefs and learning behavior. Evidence for academic outcomes is mixed and often small; a general IQ increase has not been established.
Is intelligence more genetic or environmental?
Both contribute. Their statistical contributions depend on the population and circumstances. A heritability estimate does not predict an individual's capacity for change.
Can intelligence decrease?
Specific cognitive abilities and measured scores can decline with aging, health changes, or adverse conditions. Sudden or functionally important changes call for professional assessment of possible causes.
How do I know whether an IQ score change is significant?
Check the instrument's standard error, norms, reliability, and retest guidance, along with differences in conditions. A qualified psychologist can help interpret scores in the context of the assessment's purpose.
Conclusion
Intelligence develops through maturation, education, learning, and individual experience. Broad cognitive differences are fairly stable from late adolescence through adult life, but individual trajectories vary, and different abilities follow different age patterns. Formal schooling has unusually strong evidence for lasting effects on intelligence-test outcomes. Brain training is much more consistently associated with improvement on practiced or closely related tasks than with far transfer.
The most informative question is therefore which cognitive ability changed, by what measurement, across what interval, and with what evidence about its causes. That framework explains the capacity for development while preserving the scientific meaning of intelligence as a measured set of individual-difference constructs.
Related English Hub articles
Big Five Personality Traits and Intelligence: How Personality and Cognitive Ability Differ clarifies the boundary between personality and cognitive ability.
Intelligence in the Artificial Era: What Human Intelligence Means Beside Artificial Reason explores a separate conceptual perspective on intelligence within the Artificial Era.
Related Articles
References
Breit, M., Scherrer, V., Tucker-Drob, E. M., & Preckel, F. (2024). The stability of cognitive abilities: A meta-analytic review of longitudinal studies. Psychological Bulletin, 150(4), 399–439. DOI
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
Hartshorne, J. K., & Germine, L. T. (2015). When does cognitive functioning peak? The asynchronous rise and fall of different cognitive abilities across the life span. Psychological Science, 26(4), 433–443. DOI
Haworth, C. M. A., et al. (2010). The heritability of general cognitive ability increases linearly from childhood to young adulthood. Molecular Psychiatry, 15, 1112–1120. DOI
Lowe, C. J., Safati, A., & Hall, P. A. (2017). The neurocognitive consequences of sleep restriction: A meta-analytic review. Neuroscience & Biobehavioral Reviews, 80, 586–604. DOI
Macnamara, B. N., & Burgoyne, A. P. (2023). Do growth mindset interventions impact students' academic achievement? A systematic review and meta-analysis. Psychological Bulletin, 149(3–4), 133–173. DOI
Melby-Lervåg, M., Redick, T. S., & Hulme, C. (2016). Working memory training does not improve performance on measures of intelligence or other measures of far transfer. Perspectives on Psychological Science, 11(4), 512–534. DOI
Pietschnig, J., & Voracek, M. (2015). One century of global IQ gains: A formal meta-analysis of the Flynn effect (1909–2013). Perspectives on Psychological Science, 10(3), 282–306. DOI
Ramsden, S., et al. (2011). Verbal and non-verbal intelligence changes in the teenage brain. Nature, 479, 113–116. DOI
Reparaz-Escudero, I., et al. (2024). Effect of long-term physical exercise and multidomain interventions on cognitive function and the risk of mild cognitive impairment and dementia in older adults: A systematic review with meta-analysis. Ageing Research Reviews, 100, 102463. DOI
Ritchie, S. J., & Tucker-Drob, E. M. (2018). How much does education improve intelligence? A meta-analysis. Psychological Science, 29(8), 1358–1369. DOI
Rodas, J. A., Asimakopoulou, A. A., & Greene, C. M. (2024). Can we enhance working memory? Bias and effectiveness in cognitive training studies. Psychonomic Bulletin & Review, 31, 1891–1914. DOI
Sisk, V. F., Burgoyne, A. P., Sun, J., Butler, J. L., & Macnamara, B. N. (2018). To what extent and under which circumstances are growth mind-sets important to academic achievement? Two meta-analyses. Psychological Science, 29(4), 549–571. DOI
van IJzendoorn, M. H., Juffer, F., & Klein Poelhuis, C. W. (2005). Adoption and cognitive development: A meta-analytic comparison of adopted and nonadopted children's IQ and school performance. Psychological Bulletin, 131(2), 301–316. DOI