What Is an IQ Test? Definition, History & What It Measures

What is an IQ test?

An IQ test is a standardized test that measures how well you perform on a set of reasoning and mental tasks compared with people your age. It turns that comparison into a single number, the intelligence quotient, on a scale where 100 is average and 15 points is one standard deviation.

“Standardized” carries most of the weight in that definition. It means three things:

• Same tasks, same rules. Everyone gets the same items, instructions and time limits.
• Normed. The test has been given to a large sample chosen to represent the population, so your score can be compared against theirs.
• Scored the same way. Raw answers are converted using fixed tables, by age.

Clinical IQ tests, such as the Wechsler Adult Intelligence Scale (WAIS), the Wechsler Intelligence Scale for Children (WISC) and the Stanford-Binet, are usually given one-to-one by a psychologist and take about one to two hours. They are used to help diagnose learning disabilities and intellectual disability, to assess cognitive changes after brain injury, and in school placement decisions.

Where did IQ tests come from?

IQ tests started as a practical tool for schools and changed several times on the way to their modern form.

Binet and Simon (1905)

In 1905, the French psychologists Alfred Binet and Théodore Simon published the first practical intelligence test. France had made schooling compulsory, and the education authorities wanted a fair way to identify children who needed extra support. Binet and Simon’s test used a series of tasks of increasing difficulty and expressed results as a “mental level”: the age at which a typical child could pass the same tasks.

Binet was cautious about what his test meant. He saw it as a way to find children who needed help, not as a measure of a fixed, inborn quantity.

Stern’s ratio IQ (1912)

In 1912 the German psychologist William Stern proposed dividing mental age by chronological age. A 10-year-old performing like a typical 12-year-old would get a ratio of 1.2. This was the original “intelligence quotient.”

Terman and the Stanford-Binet (1916)

Lewis Terman at Stanford University adapted Binet’s test for American children in 1916, creating the Stanford-Binet, and multiplied Stern’s ratio by 100. That gave the familiar formula:

IQ = (mental age ÷ chronological age) × 100

The ratio approach had a flaw: it falls apart for adults. Mental age stops rising in the way the formula assumes, so a 40-year-old’s “ratio” makes little sense.

This era also has a dark side worth stating plainly. Early American testing, including mass army testing in World War I, was used to support eugenic arguments, restrictive immigration policy and forced sterilization laws. Modern test publishers and psychologists have moved far from that, but the history explains much of the suspicion IQ testing still attracts.

Wechsler’s deviation IQ (1939)

David Wechsler fixed the adult problem with the Wechsler-Bellevue scale in 1939. Instead of a ratio, he compared each person with others of the same age and expressed the result as a deviation from that group’s average, setting the mean at 100 and the standard deviation at 15. The adult WAIS followed in 1955, and the children’s WISC in 1949.

The deviation IQ is what every major test uses today. The current Stanford-Binet (fifth edition) also uses a standard deviation of 15. For what the resulting numbers mean, see the IQ scale and ranges chart.

What does an IQ test measure?

An IQ test measures performance on a range of mental tasks, and the full-scale score mainly reflects what those tasks have in common. A modern clinical battery typically has subtests grouped into areas like these:

The g factor

In 1904, Charles Spearman noticed that children who did well on one kind of school test tended to do well on others. He proposed a general factor, g, behind that overlap. More than a century of data has confirmed the basic pattern, sometimes called the “positive manifold”: scores on very different mental tasks correlate positively.

Today, most researchers use hierarchical models, often the Cattell-Horn-Carroll framework, with g at the top and broader abilities (such as fluid reasoning, crystallized knowledge and processing speed) underneath. Full-scale IQ is mostly an estimate of g. The index scores describe the profile below it.

What g is biologically remains an open question. It’s a statistical regularity that’s very well replicated, not a single identified brain mechanism.

What IQ predicts

IQ scores are among the most reliable measures in psychology, and they predict a fair amount: school grades, years of education, and, more modestly, job performance and income. The predictions are for groups and averages. For any one person, plenty of other things (effort, opportunity, health, interests, luck) matter at least as much.

What doesn’t an IQ test measure?

An IQ test leaves out a lot of what people mean by “smart.” The main gaps:

• Rational thinking. Psychologist Keith Stanovich, in What Intelligence Tests Miss (2009), argued that IQ and rationality are different things. High-IQ people fall for confirmation bias, overconfidence and many other errors about as readily as anyone, especially when they aren’t prompted to check their reasoning.
• Judgment and decision-making. Choosing well under uncertainty, weighing trade-offs, and knowing what you don’t know aren’t on the test.
• Creativity. Tests reward finding the one correct answer; creative work often means finding a new question.
• Expertise and knowledge. Beyond general vocabulary and knowledge items, specialist skill is invisible.
• Motivation and persistence. Effort over years drives most achievement, and a 90-minute test can’t see it.
• Social and emotional skills. Reading people, managing conflict, and leading are outside the scope.

Alternative theories, such as Howard Gardner’s multiple intelligences (1983) and Robert Sternberg’s triarchic theory (1985), tried to widen the definition of intelligence. They’re popular in education, though they have weaker empirical support as measurement models than the g-based approach. The narrower point holds either way: IQ tests measure a specific, useful slice of cognition, not the whole mind.

Are online IQ tests the same thing?

No. The definition of an IQ test depends on norming, and most online tests skip it. They aren’t calibrated against a representative sample, the people who take them are self-selected, and conditions aren’t controlled. Some are built to flatter you so you’ll pay for a certificate.

That doesn’t make them useless. A well-made practice test can show you the kinds of items IQ tests use and help you get comfortable with them. The free Mentarium IQ practice test is built for exactly that: 30 original questions across number series, letter series, analogies, logic and patterns. It’s practice, not a clinical assessment. For what a real score means, including confidence intervals and retest effects, see IQ test scores explained.

The Mentarium take

An IQ test is a well-engineered instrument for one job: ranking how well you handle a particular set of abstract reasoning, memory and speed tasks, relative to your peers. It does that job reliably. The trouble starts when people treat the score as a verdict on their whole mind.

The things that decide how your life and work actually go, judgment, calibration, knowing when you’re wrong, are skills, and unlike IQ they respond to practice. That’s what the critical thinking skills and the mental models library are for. Take an IQ test if you’re curious. Then spend the effort where it moves something.