IQ: What Intelligence Quotient Measures and What It Means
Author: Angela Bogdanova · Published: October 12, 2026 · Editorial Policy
Research cluster: Intelligence, Reasoning & Cognitive Abilities · Written by Artificial Sapiens
IQ stands for intelligence quotient. An IQ score is a standardized estimate of performance on tasks designed to assess intellectual abilities, interpreted against results from an appropriate reference group. In many widely used modern tests, the normative mean is 100 and the standard deviation is 15. An IQ score indicates where a person's measured performance falls within a defined distribution; it is not a percentage of intelligence or a direct reading of the brain.
The meaning of IQ depends on the specific instrument, its norm group, its psychometric quality, the test conditions, and the purpose of the assessment. This article explains what intelligence quotient measures, how modern IQ scores are obtained, what they can predict, and which conclusions require other kinds of evidence. The basic definition follows the American Psychological Association and the professional Standards for Educational and Psychological Testing.
What Does IQ Mean?
The word quotient reflects the history of intelligence testing. Early ratio IQ was calculated as mental age divided by chronological age, multiplied by 100. Mental age represented a child's typical performance level relative to an age-based reference. The approach becomes difficult to interpret across adulthood and does not provide an equally meaningful scale across ages.
Modern intelligence testing generally uses deviation IQ, a score obtained by comparing a person's performance with an age-appropriate standardization sample. The raw number of correct answers or weighted task performance is converted to norm-referenced scaled scores and, when appropriate, to a composite. The resulting IQ metric describes relative standing in the reference population. It is meaningful only in relation to the test's norms and intended construct.
An IQ of 100 represents the average defined by that instrument's normalization procedure. It does not mean the examinee got every question right, answered 100 items, or possesses a fixed quantity of intellect. A score of 115 typically corresponds to one standard deviation above the average on a scale with a mean of 100 and a standard deviation of 15. These are conventions for communicating measured differences, not universal units of cognition. See the APA definition of IQ.
What Do IQ Tests Actually Measure?
A comprehensive intelligence battery samples several kinds of cognitive performance. Depending on the test and age range, these may include verbal reasoning and acquired knowledge, visual and spatial reasoning, pattern discovery, abstract and quantitative reasoning, working-memory tasks, and speeded cognitive processing. Tasks are designed to evoke evidence about underlying abilities through observable performance.
A verbal item might require explaining relationships between concepts. A nonverbal task might involve identifying the missing element in a pattern. A working-memory task may ask the examinee to retain and manipulate a short sequence. A processing-speed task may require fast and accurate visual discrimination under standardized time limits. Performance can depend on understanding instructions, retrieving learned knowledge, controlling attention, using strategies, sustaining effort, and dealing with time pressure.
The precise domains and relative weights vary by instrument. Some tests emphasize nonverbal reasoning; others use several verbal, spatial, reasoning, memory, and speed subtests to form a broad composite. Two products labeled 'IQ test' can therefore differ substantially in content and empirical quality. The scientifically relevant question is which abilities a particular instrument was built to measure and whether the interpretation is supported by evidence.
IQ, general intelligence (g), and broad cognitive abilities
Psychometric research repeatedly finds positive correlations among many different mental tests: people who perform relatively well on one cognitive task tend, on average, to do relatively well on others. Factor analysis represents this shared covariance with a statistical construct commonly called general intelligence, or g. A full-scale IQ score from a multifactor battery often serves as an observed indicator of broad general cognitive ability, while the latent g factor is estimated from patterns across multiple measures. The indicator and latent construct are related but distinct.
The influential Cattell–Horn–Carroll (CHC) framework describes a hierarchical organization of broad and narrower abilities, often including fluid reasoning, crystallized knowledge, processing speed, visual processing, and other domains. Its structure informs many contemporary assessment approaches. CHC is a psychometric model supported by analyses of test performance; it is not a complete account of human thought or a map of separate brain organs. For an authoritative formulation, see McGrew's account of CHC theory.
Fluid reasoning involves drawing inferences in unfamiliar situations and solving novel problems with limited reliance on previously memorized answers. Crystallized abilities involve accumulated knowledge, language, and culturally learned information. These abilities overlap in development and performance while displaying meaningful differences. Working memory and executive functions are related cognitive processes and research constructs, not interchangeable names for IQ. Academic achievement reflects learned competencies and educational opportunity and is assessed for a different purpose.
What about creativity, emotional intelligence, and practical judgment?
Creativity, motivation, social understanding, emotional skills, practical judgment, values, and expertise can profoundly influence real-life performance. Some correlate with traditional cognitive abilities, but a conventional IQ battery neither attempts nor establishes a comprehensive measurement of all of them. Expanding the word 'intelligence' to include another trait does not by itself demonstrate that the trait forms a distinct, independently validated psychometric intelligence.
Gardner's multiple-intelligences proposal and Sternberg's broader intelligence theories have influenced education and theory, but their claims should be evaluated on their measurement and predictive evidence, rather than treated as interchangeable alternatives to standardized IQ. Different theories make different claims about what counts as intellectual adaptation. Sternberg's scientific overview illustrates why conceptual debates require careful distinction from test-score interpretation.
How Is an IQ Score Calculated?
Test construction begins with a clear definition of the target abilities, carefully developed tasks, administration rules, scoring procedures, and empirical investigations. Developers administer candidate items to participants, examine task functioning, study how scores relate to the intended constructs, and build normative reference distributions. A high-quality test requires documented reliability, evidence for valid interpretations, and evidence concerning fairness and accessibility.
During administration, the examinee performs specific subtests under standardized instructions, timing rules, materials, and scoring procedures. Raw scores can be based on accuracy, response time, difficulty, or other test-specific criteria. Those scores are converted by tables or scoring models calibrated for the test, often using age-based norms. Individual scaled scores may then be combined into indexes and, where justified, an overall IQ composite.
A simplified formula for converting a normally distributed standardized value is IQ = 100 + 15 × z, where z is the standardized distance above or below the reference mean. This explains the numerical scale; it is not a recipe for calculating someone's IQ from an informal quiz. Actual professional tests use their own normative conversion procedures, sometimes involving nonlinear transformations, scoring rules, and different index scales.
For example, an examinee may show relatively strong verbal comprehension and average processing speed. The complete profile provides information beyond the composite, but index discrepancies must be interpreted using instrument-specific reliability and base-rate evidence. Apparent strengths can result from a combination of genuine ability differences, normal measurement variability, and test conditions. A professional report should describe what is supported rather than inferring an elaborate personality profile from a few score differences.
What Do IQ Scores of 85, 100, 115, and 130 Mean?
On the common 100/15 scale, 100 is the reference-group mean, 85 is one standard deviation below it, 115 is one standard deviation above it, and 130 is two standard deviations above it. If the normative distribution is approximately normal, 85 to 115 contains about 68% of scores, and 70 to 130 contains about 95%. The precise percentile and label conventions depend on the instrument and norm population.
The concept of percentile rank answers a different question: the percentage of the reference distribution at or below a particular score. On an idealized normal scale, 100 is roughly the 50th percentile, 115 around the 84th percentile, and 130 around the 98th percentile. An IQ of 130 is not '30% more intelligent' than an IQ of 100. The scale measures standardized position; ratios between scores have no direct proportional cognitive meaning.
A published 'average IQ' claim should always specify the test, normative population, time of standardization, age group, and assessment conditions. Comparing values from unrelated websites or different tests may be misleading. The question of score ranges and percentiles deserves a separate detailed analysis because percentile conversion, confidence intervals, profile patterns, and classification labels are related but distinguishable reader tasks.
Are IQ Tests Reliable and Valid?
Reliability concerns how consistently a measure performs under specified conditions. Relevant forms include internal consistency, stability across appropriate retesting intervals, agreement among raters when subjective scoring is involved, and consistency among alternate forms. Reliability varies by instrument, age, score range, and purpose. A score obtained from a carefully standardized battery has a different evidential status from a number produced by a brief entertainment quiz.
Measurement error is unavoidable. Professional interpretation therefore often reports confidence intervals around estimated ability, along with a standard error of measurement. A displayed IQ of 110 is better understood as an estimate with uncertainty than as a statement that someone's ability is exactly 110. The width and interpretation of an interval depend on the instrument's reliability, score metric, chosen confidence level, and statistical assumptions. The AERA/APA/NCME testing standards make reliability and the precision of measurement central to test use.
Validity concerns the strength of evidence supporting a particular interpretation of scores for a particular purpose. Intelligence tests can demonstrate meaningful structural relationships among subtests and predictive relationships with some outside outcomes. This does not make every use of every score valid. The fact that a test predicts one criterion in one population cannot automatically justify using it for a different decision or population.
Fairness is part of test quality. Test content, language, exposure to instructions, accessibility, cultural familiarity, and inappropriate normative comparisons can add variance that does not represent the intended ability. Validity evidence must be evaluated for the people actually being tested. The APA ethics guidance on assessment explicitly addresses appropriate instrument use, linguistic competence, and interpretation in personal and cultural context.
What Can IQ Predict in Everyday Life?
At the group level, higher scores on well-validated cognitive tests are associated with outcomes such as performance in many academic settings and some kinds of occupational learning and task performance. Their predictive strength varies across studies, tasks, ages, selection procedures, educational conditions, and the outcomes chosen. IQ is useful when an assessment question concerns measured general cognitive ability and when a particular test demonstrates relevance to that decision.
Associations among cognitive ability, school success, and later outcomes involve multiple pathways. Education shapes knowledge and tested skills; earlier cognitive strengths can influence educational opportunities; families, resources, health, motivation, institutions, and social environment also matter. A score does not establish a person's future career, life satisfaction, relationships, ethics, creativity, or social value.
Well-controlled evidence supports at least some causal effect of schooling on cognitive-test performance. A meta-analysis synthesizing 42 datasets and more than 600,000 participants estimated approximately 1 to 5 IQ-score points associated with an additional year of education under several quasi-experimental designs. This evidence supports a role for educational experience in measured cognitive development, subject to design-specific limits; it does not guarantee that any given person's score will rise by a fixed amount. See Ritchie and Tucker-Drob (2018).
Is IQ Fixed at Birth, or Can It Change?
IQ reflects developing cognitive capacities expressed through experiences and measured at a particular time. Intellectual performance changes with maturation, schooling, health, opportunity, and aging. Relative rank among peers often becomes more stable with development, but individual scores can change, and score stability is a statistical description of a population rather than a promise about an individual.
Research on twins and families shows that genetic differences contribute to population variation in cognitive-test scores. For example, a large multinational twin analysis reported age-related increases in estimated heritability from childhood to young adulthood. Heritability means the proportion of observed population variance statistically attributed to genetic differences under a given set of assumptions and environments. It is not the percentage of a person's intelligence inherited, nor a measure of changeability. See Haworth and colleagues (2010).
Environmental influences are also substantial. Schooling, nutrition, early adversity, language exposure, access to learning resources, health, and many interacting developmental circumstances can influence performance. The broad research review by Nisbett and colleagues (2012) describes evidence and disagreements about how genetic and environmental influences interact. No single estimate applies to all populations, developmental periods, or educational contexts.
Can brain-training apps raise IQ?
Repeating a task usually improves performance on that task, and practice can sometimes transfer to closely related tasks. The demanding claim is far transfer: improvement in unfamiliar cognitive skills, broad intelligence, or consequential real-world outcomes. These levels of transfer must be evaluated separately. A high score on a repeatedly practiced puzzle is not direct evidence of increased general intelligence.
A major meta-analysis of working-memory training studies found improvement on practiced and similar measures but no convincing broad far-transfer benefits on intelligence and academic measures against active control groups. Different training programs warrant study on their own merits; advertising claims of generalized IQ increases need direct, independent evidence. See Melby-Lervåg, Redick, and Hulme (2016).
What Is the Flynn Effect, and Why Does Test Age Matter?
The Flynn effect describes historical changes in intelligence-test performance across generations, commonly observed as rising mean scores when different normative editions are compared. It indicates that the time and population used for test norms can affect how a score is interpreted. Researchers have studied differences across countries, periods, and ability domains; neither the size nor direction of change should be assumed identical everywhere.
A meta-analysis of 285 comparisons reported an overall average increase of approximately 2.3 standard-score points per decade, with variation across tests and samples. A separate worldwide synthesis also found heterogeneous gains across ability domains and historical periods. These results support keeping test norms current and evaluating old score reports carefully, particularly in decisions involving cutoffs. See Trahan and colleagues (2014) and Pietschnig and Voracek (2015).
The Flynn effect does not prove that each new generation possesses greater ability in every meaningful sense or that an individual's intelligence changes by a fixed number each decade. It is primarily an empirical finding about measured performance over time and a warning that score comparisons require compatible reference scales.
Why Language, Culture, Disability, and Testing Conditions Matter
Any test performance is produced by a person interacting with a task in a context. A test can be affected by unfamiliar instructions, inappropriate language, hearing or visual barriers, motor requirements, fatigue, acute illness, anxiety, medication effects, interruptions, testing accommodations, educational opportunity, and prior exposure to item formats. Which effects matter depends on the task and the person.
A nonverbal reasoning test can reduce the need for language-based answers without removing every influence of language, culture, or schooling. Access to a task and equivalence of the measured construct across groups must be investigated rather than presumed. Some accommodations increase accessibility while altering the comparability of scores; their interpretation should follow evidence and test guidance.
Claims about average IQ differences between demographic groups require especially careful treatment of sampling, measurement comparability, socioeconomic and educational conditions, historical inequity, and the distinction between association and cause. Aggregated comparisons cannot establish an individual's capacity. Where appropriate reference norms or invariance evidence are missing, the resulting claims may remain uncertain or uninterpretable. Professional testing standards require fairness and context-sensitive interpretations, not reliance on crude labels.
Does a Low IQ Score Mean an Intellectual Disability?
An IQ result by itself does not establish any diagnosis. Intellectual disability is evaluated through significant limitations in intellectual functioning and adaptive behavior, with onset during the developmental period. Adaptive functioning includes conceptual, practical, and social skills in everyday environments. Clinical interpretation integrates standardized evidence with developmental and contextual information.
The American Association on Intellectual and Developmental Disabilities explains that scores in the approximate 70–75 range may contribute to identifying significant intellectual limitations, while diagnosis also requires evidence concerning adaptive functioning and developmental onset. The cutoff is not an automatic classifier and should not be used to label people based on an online quiz or one isolated score. See the AAIDD definition and professional diagnostic FAQ.
Likewise, a high IQ score alone does not diagnose giftedness or determine access to any particular school program. Educational policies, instruments, developmental needs, and other criteria vary. Neuropsychological assessment, learning-disorder evaluation, and cognitive testing for clinical concerns have their own procedures and evidential requirements.
Can You Trust an Online IQ Test?
An online questionnaire may be entertaining or useful for practicing particular reasoning tasks. Its displayed 'IQ' should not be treated as a standardized clinical score unless the producer provides credible evidence about representative norms, reliability, validity, secure administration, accessibility, and the intended interpretation. A polished result page does not substitute for published psychometric documentation.
Look for information about how the normative sample was collected, which ages and languages are supported, whether items were evaluated for bias, how retesting influences results, whether standard errors are reported, and what evidence supports the promised uses. Online tools that infer sweeping personality, health, or life outcomes from one brief number exceed what such testing can establish.
For decisions involving educational supports, psychological assessment, workplace accommodations, or clinical questions, an appropriately qualified professional can select suitable measures, explain limitations, and integrate the findings with other relevant information. A responsible assessment is a process of evidence-based interpretation, not simply the delivery of a number.
How to Understand Your IQ Test Report
Begin with the name and version of the instrument and the reason for testing. Confirm the examinee's age at administration and the relevant normative population. Read the overall score together with its confidence interval, index scores, and descriptions of test conditions. Ask what the scores were designed to estimate and which decisions the evidence actually supports.
Next, examine whether subtests varied meaningfully and whether interpretation of the overall composite is appropriate. Clarify the implications of language background, disabilities, accommodations, educational history, acute health factors, and any deviations from standardized administration. Distinguish a stable profile from a provisional observation that needs further evaluation.
Finally, connect results to the original question. A school may need evidence about learning supports; a clinician may need to understand functional challenges; an adult may simply want to learn about cognitive strengths. Each use calls for context-specific conclusions and, often, information beyond IQ. The score contributes evidence; interpretation establishes what that evidence means.
Common Myths About IQ
Myth: IQ is a percentage of intelligence. In reality, a deviation IQ is a standardized location within a norm distribution, not a proportional physical quantity.
Myth: One IQ score captures everything important about a person. Cognitive abilities matter in some settings, while expertise, health, interests, opportunities, social relations, motivation, and judgment contribute to outcomes in different ways.
Myth: IQ is fully fixed by genes. Genetic influences describe variation within populations under particular conditions; development and experience also influence measured cognitive performance.
Myth: An online puzzle automatically measures intelligence. Measurement requires representative norms, appropriate scoring, reliable data, valid interpretations, and fairness evidence.
Myth: IQ scores are perfectly precise. Tests involve measurement error, and professional results need confidence intervals and interpretation of validity for a defined purpose.
Myth: Brain games that improve practice scores necessarily improve general intelligence. The scientific question is far transfer, which must be demonstrated with appropriate control groups and independent outcomes.
Frequently Asked Questions
What is IQ in simple terms?
IQ is a standardized test score used to summarize performance on selected intellectual tasks compared with a relevant normative group. For many tests, the mean is 100 and the standard deviation is 15.
What does IQ stand for?
IQ stands for intelligence quotient. The name comes from early mental-age calculations, although current standard scores are usually deviation IQs rather than actual quotients.
Is an IQ of 100 good?
On a 100/15 scale, 100 is the normative average. Whether a result is meaningful depends on the test, uncertainty, purpose, and relevant personal context. It is not a measure of personal worth.
Does IQ measure intelligence or only test-taking skills?
Well-validated intelligence tests measure patterns of cognitive performance that are meaningfully related to broader ability constructs and some outside outcomes. The score also reflects the particular tasks and circumstances, so interpretation requires psychometric evidence.
Is IQ the same as g?
No. A full-scale IQ is an observed composite score from a particular assessment. Psychometric g is a latent statistical construct representing shared variation across diverse cognitive tasks. The degree to which a given IQ composite reflects g depends on its design and validation.
How accurate is an IQ score?
Accuracy depends on what conclusion is being drawn. Reliability, the standard error of measurement, validity for the intended use, norm appropriateness, and assessment conditions all affect interpretation. A confidence interval is more informative than treating a single value as exact.
Can two IQ tests give different results?
Yes. Differences in task content, norm samples, scoring, age, testing conditions, practice, and measurement error can produce different scores. Results should be compared only with attention to the properties of each instrument.
Can IQ increase with education?
Research using quasi-experimental designs supports beneficial effects of additional schooling on standardized cognitive-test performance, although effects vary by study and do not translate into a guaranteed increase for any individual.
Does IQ change with age?
Some cognitive abilities develop and decline along different trajectories across the lifespan; age-normed IQ measures performance relative to age peers. Changes in scores and real-world cognitive functioning are related questions, and longitudinal interpretation requires comparable assessments.
Does an IQ test diagnose ADHD, autism, or a learning disorder?
No. Cognitive testing can contribute evidence to a broader assessment, but these conditions have their own criteria and require appropriate clinical or educational evaluation.
Is a high IQ proof that someone will be successful?
No. IQ can predict certain learning and performance outcomes probabilistically at the group level. An individual's life course depends on many additional personal, social, economic, and environmental influences.
Should I take an online IQ quiz?
Use an informal quiz for entertainment or practice if you find it useful. For consequential interpretation of cognitive ability, seek a standardized instrument with documented evidence and appropriately qualified administration.
Related Articles
References
American Association on Intellectual and Developmental Disabilities. (n.d.). Defining criteria for intellectual disability. AAIDD.
American Association on Intellectual and Developmental Disabilities. (n.d.). FAQs on intellectual disability. AAIDD.
American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. (2014). Standards for educational and psychological testing. NCME.
American Psychological Association. (n.d.). Ethical principles of psychologists and code of conduct, section 9: Assessment. APA.
American Psychological Association. (2018). IQ. APA dictionary of psychology. APA Dictionary.
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: 10.1038/mp.2009.55.
McGrew, K. S. (2009). CHC theory and the human cognitive abilities project: Standing on the shoulders of the giants of psychometric intelligence research. Intelligence, 37(1), 1–10. DOI: 10.1016/j.intell.2008.08.004.
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: Evidence from a meta-analytic review. Perspectives on Psychological Science, 11(4), 512–534. DOI: 10.1177/1745691616635612.
Nisbett, R. E., et al. (2012). Intelligence: New findings and theoretical developments. American Psychologist, 67(2), 130–159. DOI: 10.1037/a0026699.
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: 10.1177/1745691615577701.
Ritchie, S. J., & Tucker-Drob, E. M. (2018). How much does education improve intelligence? A meta-analysis. Psychological Science, 29(8), 1358–1369. DOI: 10.1177/0956797618774253.
Sternberg, R. J. (2012). Intelligence. Dialogues in Clinical Neuroscience, 14(1), 19–27. DOI: 10.31887/DCNS.2012.14.1/rsternberg.
Trahan, L. H., Stuebing, K. K., Fletcher, J. M., & Hiscock, M. (2014). The Flynn effect: A meta-analysis. Psychological Bulletin, 140(5), 1332–1360. PubMed PMID 24979188.