IQ glossary

A

Abstract reasoning

Abstract reasoning is the ability to find rules in shapes, patterns and relationships that have no words or real-world meaning attached, and to apply those rules to a new case. Matrix puzzles, figure sequences and figure analogies are the classic abstract reasoning items.

Age norms

Age norms are the reference results a test publisher collects for each age group. They let an IQ test compare you only with people your own age, so a score of 100 means exactly average for your age, whether you are 20 or 70.

C

Cattell Culture Fair test

The Cattell Culture Fair test is a nonverbal intelligence test built from shapes and patterns to reduce the role of language and schooling. Its Cattell III scale uses a standard deviation of 24 rather than 15, so the same rank prints as a much larger or smaller number.

Ceiling effect

A ceiling effect happens when a test does not have enough hard items to tell high scorers apart. Many strong performers bunch up near the top possible score, so the test cannot say how far above that point they really are. A floor effect is the same problem at the bottom.

Confidence interval

A confidence interval is a range of scores that is likely to contain your true score, given the error in the test. A 95% interval of 110 to 120 means that, if the test were repeated many times, ranges built this way would capture the true score about 95% of the time.

Correlation

A correlation is a number from -1 to 1 that describes how closely two measures move together across a group. At 1 they rise together perfectly, at 0 there is no straight-line link, and at -1 one falls as the other rises. Correlations describe groups, not individuals.

Crystallized intelligence

Crystallized intelligence is the knowledge and skill you have built up through experience and education, such as vocabulary, facts and learned methods, along with the ability to use them. Raymond Cattell set it apart from fluid intelligence, reasoning with new problems, in 1963.

D

Deviation IQ

A deviation IQ is a score based on how far your result sits above or below the average of your age group, measured in standard deviations and then rescaled so the average is 100 and one standard deviation is 15 points. Every modern IQ test reports this kind of score.

Distractor

A distractor is a wrong answer option in a multiple-choice question, built to look plausible. Good distractors are near misses that catch a specific mistake, such as applying only one of two rules, so a puzzle rewards real reasoning rather than ruling out obviously wrong choices.

F

Fluid intelligence

Fluid intelligence is the ability to reason through problems you have not seen before, spotting patterns and relationships without leaning on learned facts. Raymond Cattell separated it from crystallized intelligence in 1963. Matrix puzzles and number series are classic tests of fluid reasoning.

G

General intelligence (g)

General intelligence, written g, is the shared ability that shows up across many different mental tests. People who do well on one kind of test, such as matrices, tend to do well on others, such as vocabulary or arithmetic. Charles Spearman proposed g in 1904 to explain that overlap.

H

Heritability

Heritability is the share of the differences between people in a trait, within a particular population and time, that is linked to genetic differences between them. It is a statement about variation in a group, not about how much of one person's trait comes from their genes.

I

Inductive reasoning

Inductive reasoning is working out a general rule from a few specific examples, then using that rule to predict or check a new case. Sequence, matrix and odd-one-out puzzles are inductive: you see examples of a pattern and must infer the rule that produced them.

Intelligence quotient (IQ)

An intelligence quotient, or IQ, is a standardized score from an intelligence test. It shows how your performance compares with people your own age, on a scale where the average is 100 and one standard deviation is 15 points. It describes rank, not a count of right answers.

Item difficulty

Item difficulty is how hard a single test question is. In item response theory it is a number on the same scale as ability: the ability level at which a person has an even, one in two chance of answering the item correctly. Higher values mean harder items.

Item response theory

Item response theory, or IRT, is a family of statistical models for test scoring. Instead of just counting right answers, it models how likely each person is to get each item right, based on their ability and the properties of the item, such as its difficulty.

M

Mean

The mean is the arithmetic average of a set of scores: add them all up and divide by how many there are. IQ tests are built so that the mean score of each age group in the reference sample is exactly 100, which is why an average IQ is 100 by definition.

Mensa

Mensa is a society open to people who score at or above the 98th percentile, the top 2%, on an approved, supervised standardized intelligence test. The qualifying number depends on the test scale: about 131 on a scale with a standard deviation of 15.

Mental rotation

Mental rotation is the ability to turn a shape in your mind and judge whether it matches another shape or is its mirror image. It is a core spatial reasoning skill, tested with flat figures turned on the page or with clusters of cubes turned in space.

Meta-analysis

A meta-analysis is a study of studies. It gathers the results of many separate studies on the same question and combines them into one overall estimate, usually giving more weight to larger and more precise studies. It smooths out the noise in any single sample.

N

Negative Flynn effect

The negative Flynn effect is a reported decline in average measured IQ among people born in recent decades, reversing the long rise seen through the 20th century. The best-known evidence comes from Norwegian conscript data for cohorts born after about 1975.

Normal distribution

A normal distribution is a symmetric, bell-shaped spread of values around a mean. Most values sit near the middle and fewer appear the farther you go out. IQ scores are scaled to follow a normal distribution, which is why they draw the familiar bell curve.

O

Odd one out

An odd-one-out puzzle shows several figures that all follow one hidden rule and one figure that breaks it. Your task is to find the figure that does not belong. The trick is to ignore the features that vary freely, such as size or tilt, and find the one that matters.

P

Percentile

A percentile is the share of people in a reference group who score below a given score. An IQ at the 90th percentile is higher than about 90% of people the same age. It describes rank, so it runs from just above 0 to just below 100, never past them.

Practice effect

The practice effect is the rise in test scores that comes from having taken the same or a similar test before. On cognitive ability tests, a retake raises scores by about a quarter of a standard deviation on average, roughly 4 IQ points, and more when the same form is reused.

Processing speed

Processing speed is how quickly a person can take in simple information and respond to it accurately, such as scanning a row of symbols for a match. It is measured on easy tasks, where nearly everyone would be correct given enough time, so speed is what separates people.

Psychometrics

Psychometrics is the field that designs, scores and evaluates psychological tests, including IQ tests, personality questionnaires and aptitude tests. It uses statistics to check whether a test is consistent, whether it measures what it claims, and how to turn answers into fair scores.

R

Rasch model

The Rasch model is a scoring method from item response theory. It says the chance of answering an item correctly depends on just two things: your ability and the item difficulty, both on the same scale. When ability equals difficulty, the chance is exactly one in two.

Raven's Progressive Matrices

Raven's Progressive Matrices is a nonverbal reasoning test made of matrix puzzles: a grid of figures with one missing, and a set of options to complete it. John C. Raven developed it in the 1930s and first published it in 1938. It is widely used as a measure of fluid reasoning.

Raw score

A raw score is the plain count of items you answered correctly on a test, before it is compared with anyone else. It is the starting point for an IQ, not the IQ itself. The same raw score can mean very different things on an easy test and a hard one.

Reliability

Reliability is how consistent a test is. A reliable test gives the same person a similar score if they take it again, and its parts agree with each other. It is usually reported as a coefficient from 0 to 1, where values closer to 1 mean less random error.

S

Standard deviation

Standard deviation is a measure of how spread out a set of scores is around its average. A small standard deviation means scores cluster near the mean; a large one means they spread widely. On most modern IQ tests one standard deviation equals 15 points.

Standard error of measurement

The standard error of measurement, or SEM, estimates how much a person's observed score would vary around their true score if they could take the test again and again without practice effects. A smaller SEM means a more precise score. It is used to build confidence intervals.

Standardization sample

A standardization sample, also called a norming sample, is the large group of people a test is given to before release. Their results define what average and above average mean on that test, separately for each age group. Every IQ score is a comparison with this group.

Stanford-Binet

The Stanford-Binet Intelligence Scales are a clinical intelligence test descended from Binet and Simon's 1905 test. Lewis Terman published the Stanford revision in 1916, and the fifth edition, SB5, by Gale Roid, appeared in 2003. Modern editions report deviation IQs with mean 100 and SD 15.

T

Theta

Theta is the ability estimate in item response theory. It sits on the same scale as item difficulty, with the average person near 0, and is estimated from the full pattern of right and wrong answers. On IQ Octopus, the estimated IQ equals 100 plus 15 times theta.

V

Validity

Validity is whether a test measures what it claims to measure and whether its scores support the way they are used. A test can be valid for one purpose, such as research on groups, and not for another, such as a clinical diagnosis or a hiring decision.

W

WAIS

The WAIS, or Wechsler Adult Intelligence Scale, is a clinical IQ test for adults, given by a trained examiner. It grew from the Wechsler-Bellevue scale of 1939. WAIS-IV was published in 2008 and WAIS-5 in 2024. It reports deviation IQs with mean 100 and SD 15.

Working memory

Working memory is the small mental workspace that holds a few pieces of information in mind while you use them. Keeping two rules in mind while checking a puzzle option, or carrying a number while doing mental arithmetic, both rely on working memory.

Z

Z-score

A z-score says how many standard deviations a score sits above or below the mean. A z of 0 is exactly average, +1 is one standard deviation above and -2 is two below. On the standard IQ scale, z equals the IQ minus 100, divided by 15.

Related

• No. 01QuestionsShort, sourced answers to the most asked IQ questions.
• No. 02GuidesPlain, sourced explanations of IQ and IQ tests.
• No. 03The IQ bell curveMean 100, standard deviation 15, and what a percentile really says.